{"chunk_id": "application_climate-pulse_extreme-events_q01__7dad8261f153", "report_id": "application_climate-pulse_extreme-events_q01", "dataset_id": "climate-pulse", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extreme-events_q01", "aspect_base": "extreme-events", "category": "Applications", "match_confidence": "unmatched", "section": "Visualisation of temperature during extreme weather events with Climate Pulse", "title": "Visualisation of temperature during extreme weather events with Climate Pulse", "chunk_index": 0, "token_count": 99, "text_raw": "Production date: 2025-06-30 (updated 2026-01-09).\n\nApplication version: 1.0.\n\nProduced by: Nicole Reynolds, Olivier Burggraaff (National Physical Laboratory).", "text_with_prefix": "EQC Quality Assessment: \"Visualisation of temperature during extreme weather events with Climate Pulse\"\nDataset: climate-pulse [CDS]\nAspect: extreme-events_q01 | Category: Applications\nSection: Visualisation of temperature during extreme weather events with Climate Pulse\n---\nProduction date: 2025-06-30 (updated 2026-01-09).\n\nApplication version: 1.0.\n\nProduced by: Nicole Reynolds, Olivier Burggraaff (National Physical Laboratory)."} {"chunk_id": "application_climate-pulse_extreme-events_q01__733288de2970", "report_id": "application_climate-pulse_extreme-events_q01", "dataset_id": "climate-pulse", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extreme-events_q01", "aspect_base": "extreme-events", "category": "Applications", "match_confidence": "unmatched", "section": "Visualisation of temperature during extreme weather events with Climate Pulse > Quality assessment question", "title": "Visualisation of temperature during extreme weather events with Climate Pulse", "chunk_index": 1, "token_count": 335, "text_raw": "* **Is the Climate Pulse application an appropriate tool for broad audiences to visualise temperature data relating to extreme weather and climate events, including El Niño & La Niña, heatwaves, and volcanic eruptions?**\n\nExtreme weather events – such as El Niño & La Niña, heatwaves, and volcanic eruptions – can have widespread impacts on human health, society, and the economy [[Bell+18](https://doi.org/10.1080/10962247.2017.1401017)]. The frequency and intensity of extreme weather events are increasing under climate change [[Kundzewicz+16](https://doi.org/10.1515/igbp-2016-0005)], making timely and accurate communication of these events and their impacts more critical than ever.\n\nThe [*Climate Pulse*](https://pulse.climate.copernicus.eu/) application provides near real-time visualisations of global surface air and sea surface temperature, along with an archive of past daily, monthly, and annual maps derived from the global ERA5 climate reanalysis dataset [[Hersbach+20](https://doi.org/10.1002/qj.3803)]. Climate Pulse is designed to help a broad audience clearly and effectively communicate global temperatures.\n\nThis assessment examines the suitability of the Climate Pulse application for visualising the temperature effects of extreme weather events.", "text_with_prefix": "EQC Quality Assessment: \"Visualisation of temperature during extreme weather events with Climate Pulse\"\nDataset: climate-pulse [CDS]\nAspect: extreme-events_q01 | Category: Applications\nSection: Visualisation of temperature during extreme weather events with Climate Pulse > Quality assessment question\n---\n* **Is the Climate Pulse application an appropriate tool for broad audiences to visualise temperature data relating to extreme weather and climate events, including El Niño & La Niña, heatwaves, and volcanic eruptions?**\n\nExtreme weather events – such as El Niño & La Niña, heatwaves, and volcanic eruptions – can have widespread impacts on human health, society, and the economy [[Bell+18](https://doi.org/10.1080/10962247.2017.1401017)]. The frequency and intensity of extreme weather events are increasing under climate change [[Kundzewicz+16](https://doi.org/10.1515/igbp-2016-0005)], making timely and accurate communication of these events and their impacts more critical than ever.\n\nThe [*Climate Pulse*](https://pulse.climate.copernicus.eu/) application provides near real-time visualisations of global surface air and sea surface temperature, along with an archive of past daily, monthly, and annual maps derived from the global ERA5 climate reanalysis dataset [[Hersbach+20](https://doi.org/10.1002/qj.3803)]. Climate Pulse is designed to help a broad audience clearly and effectively communicate global temperatures.\n\nThis assessment examines the suitability of the Climate Pulse application for visualising the temperature effects of extreme weather events."} {"chunk_id": "application_climate-pulse_extreme-events_q01__f13a31ea8f34", "report_id": "application_climate-pulse_extreme-events_q01", "dataset_id": "climate-pulse", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extreme-events_q01", "aspect_base": "extreme-events", "category": "Applications", "match_confidence": "unmatched", "section": "Visualisation of temperature during extreme weather events with Climate Pulse > Quality assessment statement", "title": "Visualisation of temperature during extreme weather events with Climate Pulse", "chunk_index": 2, "token_count": 296, "text_raw": "These are the key outcomes of this assessment\n\nThe Climate Pulse application:\n* Is suitable for visualising current and historical (from 1979) **El Niño and La Niña** events. The monthly maps are at a suitable spatial and temporal scale for the visualisation of global surface air and sea surface temperature anomalies. The global average temperature anomaly time series allows for long-term comparison and inspection of the impacts of ENSO events.\n* Is suitable for visualising **heatwaves** from January 2023. This is due to the availability of daily surface air temperature anomaly maps, which can be zoomed in on affected regions, thereby ensuring both spatial and temporal suitability. Prior to 2023, only monthly maps are available, which lack the temporal resolution needed to capture these typically short-term events.\n* Is not suitable for visualising the temperature impacts of **volcanic eruptions** due to it only displaying near-surface (2 m) air temperature data that do not capture the stratospheric effects of these events, along with its use of a 1991–2020 baseline, which makes cooling from historical eruptions like Pinatubo less visible.\n```", "text_with_prefix": "EQC Quality Assessment: \"Visualisation of temperature during extreme weather events with Climate Pulse\"\nDataset: climate-pulse [CDS]\nAspect: extreme-events_q01 | Category: Applications\nSection: Visualisation of temperature during extreme weather events with Climate Pulse > Quality assessment statement\n---\nThese are the key outcomes of this assessment\n\nThe Climate Pulse application:\n* Is suitable for visualising current and historical (from 1979) **El Niño and La Niña** events. The monthly maps are at a suitable spatial and temporal scale for the visualisation of global surface air and sea surface temperature anomalies. The global average temperature anomaly time series allows for long-term comparison and inspection of the impacts of ENSO events.\n* Is suitable for visualising **heatwaves** from January 2023. This is due to the availability of daily surface air temperature anomaly maps, which can be zoomed in on affected regions, thereby ensuring both spatial and temporal suitability. Prior to 2023, only monthly maps are available, which lack the temporal resolution needed to capture these typically short-term events.\n* Is not suitable for visualising the temperature impacts of **volcanic eruptions** due to it only displaying near-surface (2 m) air temperature data that do not capture the stratospheric effects of these events, along with its use of a 1991–2020 baseline, which makes cooling from historical eruptions like Pinatubo less visible.\n```"} {"chunk_id": "application_climate-pulse_extreme-events_q01__f7082f21bfa7", "report_id": "application_climate-pulse_extreme-events_q01", "dataset_id": "climate-pulse", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extreme-events_q01", "aspect_base": "extreme-events", "category": "Applications", "match_confidence": "unmatched", "section": "Visualisation of temperature during extreme weather events with Climate Pulse > Methodology", "title": "Visualisation of temperature during extreme weather events with Climate Pulse", "chunk_index": 3, "token_count": 755, "text_raw": "Climate Pulse is an interactive application designed for a broad audience to visualise global surface air temperature (at 2 m above the surface) and sea surface temperature (at 10 m below the surface).\nIt displays daily, monthly, and annual maps along with a global average time series (Figure {numref}`{number} `).\nThese are based on near real-time data (typical delay of about two days) from the ERA5 climate reanalysis [[Hersbach+20](https://doi.org/10.1002/qj.3803)],\nincluding preliminary data from ERA5T,\nregridded onto a regular 0.25° × 0.25° grid, approximately 30 km × 30 km.\nTable {numref}`{number} ` outlines the temporal availability of surface air temperature and sea surface temperature data.\n\nClimate Pulse displays both absolute temperature values and anomalies.\nThe anomalies are calculated as the difference between the current temperature and the 1991–2020 average,\nthe standard reference period used by the World Meteorological Organization (WMO).\nFor daily anomalies, this baseline is smoothed with a low-pass filter to remove short-term variability.\n\n:::{table} Temporal availability of each visualisation format for surface air temperature and sea surface temperature (start dates of available data).\n:widths: auto\n:align: left\n:name: application_climate-pulse_table_availability\n| | Surface Air Temperature | Sea Surface Temperature |\n|---------------|-------------------------| ------------------------|\n| Time Series | 1 January 1940 | 1 January 1979 |\n| Daily Map | 1 January 2023 | 1 January 2023 |\n| Monthly Map | January 1979 | January 1979 |\n| Annual Map | 1979 | 1979 |\n:::\n\nOutputs from Climate Pulse can be exported in three ways:\n1. Data download\n2. Time series and global image export (Figure {numref}`{number} `)\n3. Screenshot of the dashboard (Figure {numref}`{number} `)\n\nattachment:application_climate-pulse_example_exports.png\n---\nname: application_climate-pulse_example_exports\n---\nExample image exports from Climate Pulse. a) Time series of daily global surface air temperature average from 1940, b) Daily map of surface air temperature anomaly (7 January 2026).\n```\n\nattachment:application_climate-pulse_example_screenshots.png\n---\nname: application_climate-pulse_example_screenshots\n---\nScreenshots of the Climate Pulse dashboard. a) Air temperature absolute values, b) Air temperature anomalies, c) Sea temperature absolute values, d) Sea temperature anomalies.\n```\n\nApplications are quality assessed separately from their underpinning datasets. For information on quality attributes of the underlying datasets, the reader is referred to the quality assessment of the datasets, which can be found in their respective CDS catalogue entries or [on the EQC hub](../intro).\n\nThe analysis and results are organised in the following use cases, which are detailed in the sections below:\n\n**[](section-enso)**\n\n**[](section-heatwaves)**\n\n**[](section-volcanoes)**", "text_with_prefix": "EQC Quality Assessment: \"Visualisation of temperature during extreme weather events with Climate Pulse\"\nDataset: climate-pulse [CDS]\nAspect: extreme-events_q01 | Category: Applications\nSection: Visualisation of temperature during extreme weather events with Climate Pulse > Methodology\n---\nClimate Pulse is an interactive application designed for a broad audience to visualise global surface air temperature (at 2 m above the surface) and sea surface temperature (at 10 m below the surface).\nIt displays daily, monthly, and annual maps along with a global average time series (Figure {numref}`{number} `).\nThese are based on near real-time data (typical delay of about two days) from the ERA5 climate reanalysis [[Hersbach+20](https://doi.org/10.1002/qj.3803)],\nincluding preliminary data from ERA5T,\nregridded onto a regular 0.25° × 0.25° grid, approximately 30 km × 30 km.\nTable {numref}`{number} ` outlines the temporal availability of surface air temperature and sea surface temperature data.\n\nClimate Pulse displays both absolute temperature values and anomalies.\nThe anomalies are calculated as the difference between the current temperature and the 1991–2020 average,\nthe standard reference period used by the World Meteorological Organization (WMO).\nFor daily anomalies, this baseline is smoothed with a low-pass filter to remove short-term variability.\n\n:::{table} Temporal availability of each visualisation format for surface air temperature and sea surface temperature (start dates of available data).\n:widths: auto\n:align: left\n:name: application_climate-pulse_table_availability\n| | Surface Air Temperature | Sea Surface Temperature |\n|---------------|-------------------------| ------------------------|\n| Time Series | 1 January 1940 | 1 January 1979 |\n| Daily Map | 1 January 2023 | 1 January 2023 |\n| Monthly Map | January 1979 | January 1979 |\n| Annual Map | 1979 | 1979 |\n:::\n\nOutputs from Climate Pulse can be exported in three ways:\n1. Data download\n2. Time series and global image export (Figure {numref}`{number} `)\n3. Screenshot of the dashboard (Figure {numref}`{number} `)\n\nattachment:application_climate-pulse_example_exports.png\n---\nname: application_climate-pulse_example_exports\n---\nExample image exports from Climate Pulse. a) Time series of daily global surface air temperature average from 1940, b) Daily map of surface air temperature anomaly (7 January 2026).\n```\n\nattachment:application_climate-pulse_example_screenshots.png\n---\nname: application_climate-pulse_example_screenshots\n---\nScreenshots of the Climate Pulse dashboard. a) Air temperature absolute values, b) Air temperature anomalies, c) Sea temperature absolute values, d) Sea temperature anomalies.\n```\n\nApplications are quality assessed separately from their underpinning datasets. For information on quality attributes of the underlying datasets, the reader is referred to the quality assessment of the datasets, which can be found in their respective CDS catalogue entries or [on the EQC hub](../intro).\n\nThe analysis and results are organised in the following use cases, which are detailed in the sections below:\n\n**[](section-enso)**\n\n**[](section-heatwaves)**\n\n**[](section-volcanoes)**"} {"chunk_id": "application_climate-pulse_extreme-events_q01__f29a0afd53d8", "report_id": "application_climate-pulse_extreme-events_q01", "dataset_id": "climate-pulse", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extreme-events_q01", "aspect_base": "extreme-events", "category": "Applications", "match_confidence": "unmatched", "section": "Visualisation of temperature during extreme weather events with Climate Pulse > Analysis and results", "title": "Visualisation of temperature during extreme weather events with Climate Pulse", "chunk_index": 4, "token_count": 165, "text_raw": "This quality assessment evaluates the suitability of Climate Pulse for visualising extreme weather events for use by broad audiences.\nThree types of events were selected as test cases: El Niño & La Niña, heatwaves, and volcanic eruptions.\nClimate Pulse was assessed based on the availability of relevant data variables, granularity of the data, spatial and temporal coverage and resolution, and the timeliness of data updates.\nThe assessment is illustrated with practical examples of how these extreme weather events appear in Climate Pulse in real-world scenarios.\n\n(section-enso)=", "text_with_prefix": "EQC Quality Assessment: \"Visualisation of temperature during extreme weather events with Climate Pulse\"\nDataset: climate-pulse [CDS]\nAspect: extreme-events_q01 | Category: Applications\nSection: Visualisation of temperature during extreme weather events with Climate Pulse > Analysis and results\n---\nThis quality assessment evaluates the suitability of Climate Pulse for visualising extreme weather events for use by broad audiences.\nThree types of events were selected as test cases: El Niño & La Niña, heatwaves, and volcanic eruptions.\nClimate Pulse was assessed based on the availability of relevant data variables, granularity of the data, spatial and temporal coverage and resolution, and the timeliness of data updates.\nThe assessment is illustrated with practical examples of how these extreme weather events appear in Climate Pulse in real-world scenarios.\n\n(section-enso)="} {"chunk_id": "application_climate-pulse_extreme-events_q01__ec50fcaa36d7", "report_id": "application_climate-pulse_extreme-events_q01", "dataset_id": "climate-pulse", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extreme-events_q01", "aspect_base": "extreme-events", "category": "Applications", "match_confidence": "unmatched", "section": "Visualisation of temperature during extreme weather events with Climate Pulse > Analysis and results > 1. El Niño–Southern Oscillation (ENSO) > Introduction", "title": "Visualisation of temperature during extreme weather events with Climate Pulse", "chunk_index": 5, "token_count": 848, "text_raw": "El Niño (warm phase) and La Niña (cold phase) are opposite phases of the ENSO (El Niño–Southern Oscillation),\na recurring climate phenomenon in the equatorial Pacific Ocean.\nThey typically recur every 2 to 7 years and last 9 to 12 months\n[[NOAA+24](https://oceanservice.noaa.gov/facts/ninonina.html), [Wang+16](https://doi.org/10.1007/978-94-017-7499-4_4), [McPhaden+06](https://doi.org/10.1126/science.1132588)].\nENSO events are declared when regions in the central and eastern equatorial Pacific,\nso-called *Niño regions*,\nmeet certain temperature anomaly thresholds.\nCommonly used thresholds are +0.5 °C for an El Niño and –0.5 °C for a La Niña event,\nobserved for 3 months in a 5-month consecutive period\n[[Shen+23](https://doi.org/10.1007/s44195-023-00051-5), [McPhaden+06](https://doi.org/10.1126/science.1132588), [Philander+85](https://doi.org/10.1175/1520-0469%281985%29042%3C2652:ENALN%3E2.0.CO;2)].\n\nThe ENSO has widespread effects on global weather patterns (Figure {numref}`{number} `).\nWarming due to El Niño,\nalong with ongoing climate change,\nhas led recent events to break many temperature records.\nThe 2023–2024 El Niño event was widely reported in the media as driving nine straight months of record-breaking high temperatures [[AFP+24](https://timesofmalta.com/article/february-marks-9th-straight-month-recordsmashing-global-heat.1087990), [Rowlatt+23](https://www.bbc.co.uk/news/science-environment-66143682)]. \nhe media also focused on the severely negative impact of this event on human health,\nsuch as higher temperatures causing the spread of malaria into traditionally cooler regions of East Africa,\nand on the negative economic impacts historically associated with El Niño\n[[Mefo Newuh+24](https://www.dw.com/en/el-nino-climate-pattern-intensifies-in-east-africa/a-68932184), [Gerretsen+23](https://www.bbc.co.uk/future/article/20230525-what-will-an-el-nino-in-2023-mean-for-you)].\nThe 2020–2021 La Niña prompted the media to publish warnings of increased flood risk and increased storm power and frequency in North America;\nhowever,\nsome outlets focused on potential positive impacts,\nsuch as a reduction in Australia’s summer bushfires [[Johnson+20](https://www.dailymail.co.uk/news/article-8784073/How-La-Nina-weather-event-decade-bring-wet-summer.html), [McGrath+20](https://www.bbc.co.uk/news/science-environment-54725970)].\n\nattachment:application_climate-pulse_noaa_enso.png\n---\nname: application_climate-pulse_noaa_enso\n---\nGlobal distribution of climate impacts of El Niño and La Niña.\nSource: NOAA (National Oceanic and Atmospheric Administration).\nAvailable at: https://www.climate.gov/news-features/featured-images/global-impacts-el-ni%C3%B1o-and-la-ni%C3%B1a\n```", "text_with_prefix": "EQC Quality Assessment: \"Visualisation of temperature during extreme weather events with Climate Pulse\"\nDataset: climate-pulse [CDS]\nAspect: extreme-events_q01 | Category: Applications\nSection: Visualisation of temperature during extreme weather events with Climate Pulse > Analysis and results > 1. El Niño–Southern Oscillation (ENSO) > Introduction\n---\nEl Niño (warm phase) and La Niña (cold phase) are opposite phases of the ENSO (El Niño–Southern Oscillation),\na recurring climate phenomenon in the equatorial Pacific Ocean.\nThey typically recur every 2 to 7 years and last 9 to 12 months\n[[NOAA+24](https://oceanservice.noaa.gov/facts/ninonina.html), [Wang+16](https://doi.org/10.1007/978-94-017-7499-4_4), [McPhaden+06](https://doi.org/10.1126/science.1132588)].\nENSO events are declared when regions in the central and eastern equatorial Pacific,\nso-called *Niño regions*,\nmeet certain temperature anomaly thresholds.\nCommonly used thresholds are +0.5 °C for an El Niño and –0.5 °C for a La Niña event,\nobserved for 3 months in a 5-month consecutive period\n[[Shen+23](https://doi.org/10.1007/s44195-023-00051-5), [McPhaden+06](https://doi.org/10.1126/science.1132588), [Philander+85](https://doi.org/10.1175/1520-0469%281985%29042%3C2652:ENALN%3E2.0.CO;2)].\n\nThe ENSO has widespread effects on global weather patterns (Figure {numref}`{number} `).\nWarming due to El Niño,\nalong with ongoing climate change,\nhas led recent events to break many temperature records.\nThe 2023–2024 El Niño event was widely reported in the media as driving nine straight months of record-breaking high temperatures [[AFP+24](https://timesofmalta.com/article/february-marks-9th-straight-month-recordsmashing-global-heat.1087990), [Rowlatt+23](https://www.bbc.co.uk/news/science-environment-66143682)]. \nhe media also focused on the severely negative impact of this event on human health,\nsuch as higher temperatures causing the spread of malaria into traditionally cooler regions of East Africa,\nand on the negative economic impacts historically associated with El Niño\n[[Mefo Newuh+24](https://www.dw.com/en/el-nino-climate-pattern-intensifies-in-east-africa/a-68932184), [Gerretsen+23](https://www.bbc.co.uk/future/article/20230525-what-will-an-el-nino-in-2023-mean-for-you)].\nThe 2020–2021 La Niña prompted the media to publish warnings of increased flood risk and increased storm power and frequency in North America;\nhowever,\nsome outlets focused on potential positive impacts,\nsuch as a reduction in Australia’s summer bushfires [[Johnson+20](https://www.dailymail.co.uk/news/article-8784073/How-La-Nina-weather-event-decade-bring-wet-summer.html), [McGrath+20](https://www.bbc.co.uk/news/science-environment-54725970)].\n\nattachment:application_climate-pulse_noaa_enso.png\n---\nname: application_climate-pulse_noaa_enso\n---\nGlobal distribution of climate impacts of El Niño and La Niña.\nSource: NOAA (National Oceanic and Atmospheric Administration).\nAvailable at: https://www.climate.gov/news-features/featured-images/global-impacts-el-ni%C3%B1o-and-la-ni%C3%B1a\n```"} {"chunk_id": "application_climate-pulse_extreme-events_q01__c4655903e5f5", "report_id": "application_climate-pulse_extreme-events_q01", "dataset_id": "climate-pulse", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extreme-events_q01", "aspect_base": "extreme-events", "category": "Applications", "match_confidence": "unmatched", "section": "Visualisation of temperature during extreme weather events with Climate Pulse > Analysis and results > 1. El Niño–Southern Oscillation (ENSO) > Assessment of Climate Pulse", "title": "Visualisation of temperature during extreme weather events with Climate Pulse", "chunk_index": 6, "token_count": 528, "text_raw": "Climate Pulse is determined to be suitable for visualising the temperature impacts of El Niño and La Niña events for the following reasons:\n* **Variables:** Climate Pulse allows for the visualisation of temperature anomalies against WMO’s standard reference period, which is a key indicator of ENSO events. The application is suitable for visualising both increases (El Niño) and decreases (La Niña) in temperature anomalies.\n* **Data Granularity:** Climate Pulse’s monthly maps show anomalies on a scale from –7 °C to +7 °C, in increments of 0.5 to 1 °C (nonlinear colour scale) for surface air temperature and –3 °C to +3 °C, in increments of 0.1 to 0.5 °C for sea surface temperature. This scale is fine and wide enough to display the effects of ENSO events.\n* **Spatial Coverage and Resolution:** The spatial coverage (global) and resolution (0.25° × 0.25°) of the temperature anomaly maps are appropriate for visualising the effects of the ENSO on air and sea temperatures, notably including the central and eastern Pacific regions where ENSO patterns are most pronounced.\n* **Temporal Coverage and Resolution:** Climate Pulse shows near real-time and historical temperature data at daily, monthly, and annual time scales. For ENSO events, which typically develop and persist over several months, the monthly data are particularly suitable. The availability of monthly data from January 1979 allows users to look back at previous events, supporting comparisons across multiple ENSO cycles, and helping to contextualise current events within long-term trends.\n* **Timeliness:** Climate Pulse is typically updated with data for the latest month within three days, making it well-suited for timely reporting on current events and climate monitoring.\n* **Uncertainty:** Climate Pulse does not show uncertainty estimates in the time series or maps. [Uncertainties in ERA5 temperatures](https://confluence.ecmwf.int/x/ZSOgBg) are generally on the order of 0.5 °C. This is suitable for the use case of visualising ENSO events, which cause significantly larger temperature anomalies.", "text_with_prefix": "EQC Quality Assessment: \"Visualisation of temperature during extreme weather events with Climate Pulse\"\nDataset: climate-pulse [CDS]\nAspect: extreme-events_q01 | Category: Applications\nSection: Visualisation of temperature during extreme weather events with Climate Pulse > Analysis and results > 1. El Niño–Southern Oscillation (ENSO) > Assessment of Climate Pulse\n---\nClimate Pulse is determined to be suitable for visualising the temperature impacts of El Niño and La Niña events for the following reasons:\n* **Variables:** Climate Pulse allows for the visualisation of temperature anomalies against WMO’s standard reference period, which is a key indicator of ENSO events. The application is suitable for visualising both increases (El Niño) and decreases (La Niña) in temperature anomalies.\n* **Data Granularity:** Climate Pulse’s monthly maps show anomalies on a scale from –7 °C to +7 °C, in increments of 0.5 to 1 °C (nonlinear colour scale) for surface air temperature and –3 °C to +3 °C, in increments of 0.1 to 0.5 °C for sea surface temperature. This scale is fine and wide enough to display the effects of ENSO events.\n* **Spatial Coverage and Resolution:** The spatial coverage (global) and resolution (0.25° × 0.25°) of the temperature anomaly maps are appropriate for visualising the effects of the ENSO on air and sea temperatures, notably including the central and eastern Pacific regions where ENSO patterns are most pronounced.\n* **Temporal Coverage and Resolution:** Climate Pulse shows near real-time and historical temperature data at daily, monthly, and annual time scales. For ENSO events, which typically develop and persist over several months, the monthly data are particularly suitable. The availability of monthly data from January 1979 allows users to look back at previous events, supporting comparisons across multiple ENSO cycles, and helping to contextualise current events within long-term trends.\n* **Timeliness:** Climate Pulse is typically updated with data for the latest month within three days, making it well-suited for timely reporting on current events and climate monitoring.\n* **Uncertainty:** Climate Pulse does not show uncertainty estimates in the time series or maps. [Uncertainties in ERA5 temperatures](https://confluence.ecmwf.int/x/ZSOgBg) are generally on the order of 0.5 °C. This is suitable for the use case of visualising ENSO events, which cause significantly larger temperature anomalies."} {"chunk_id": "application_climate-pulse_extreme-events_q01__5c13442f4eb5", "report_id": "application_climate-pulse_extreme-events_q01", "dataset_id": "climate-pulse", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extreme-events_q01", "aspect_base": "extreme-events", "category": "Applications", "match_confidence": "unmatched", "section": "Visualisation of temperature during extreme weather events with Climate Pulse > Analysis and results > 1. El Niño–Southern Oscillation (ENSO) > Example: 2023–2024 El Niño event and 2020–2021 La Niña event", "title": "Visualisation of temperature during extreme weather events with Climate Pulse", "chunk_index": 7, "token_count": 1063, "text_raw": "The 2023–2024 El Niño event was officially declared by WMO on 4 July 2023 and dissipated in April 2024 [[WMO+24](https://wmo.int/news/media-centre/el-nino-forecast-swing-la-nina-later-year), [WMO+23](https://wmo.int/news/media-centre/world-meteorological-organization-declares-onset-of-el-nino-conditions)].\nThe 2020–2021 La Niña event emerged in August 2020 and dissipated in May 2021 [[Li+22](https://doi.org/10.1029/2021JD035546)].\nFigures {numref}`{number} ` and {numref}`{number} ` show global surface air and sea surface temperature anomalies during these periods,\nalong with a typical year without ENSO activity,\nvisualised using Climate Pulse.\nThe similarities with Figure {numref}`{number} ` are evident, particularly the warming (El Niño) and cooling (La Niña) over and in the central and eastern Pacific, and broader global temperature increases.\n\nClimate Pulse is shown to visualise ENSO events effectively by clearly depicting the associated temperature patterns.\nSome variations from the expected patterns exist,\nbut these are characteristic of the event.\n\nFigure {numref}`{number} ` shows the global temperatures for both surface air and sea surface as clearly higher for the El Niño period.\nThe general increase in global temperatures relative to the baseline period (1991–2020) may cause the amplitude of El Niño anomalies to appear exaggerated.\nThe opposite is true for La Niña, which may appear weaker than in reality.\nThis issue can be mitigated by using year-on-year comparisons.\n\nFigure {numref}`{number} ` shows the global average temperatures for 2019 (no ENSO activity), 2020, and 2021.\nThe 2020–2021 La Niña was part of a rare triple-dip La Niña,\nwhich occurred over three consecutive years from 2020;\nhence, comparison with succeeding years is not useful [[Shi+23](https://doi.org/10.1016/j.atmosres.2023.106937)].\nInstead, Figure {numref}`{number} ` compares temperature anomalies with 2019,\nshowing generally lower values in late 2020 and the first half of 2021,\ncorresponding with La Niña.\n\nIn conclusion, Climate Pulse is suitable for visualising El Niño and La Niña events through their temperature anomalies,\nmaking it a valuable tool for a broad audience,\nincluding journalists and communicators,\nseeking to illustrate the climate impacts of such events.\n\nattachment:application_climate-pulse_enso_air.png\n---\nname: application_climate-pulse_enso_air\n---\nImage exports of monthly surface air temperature anomaly for the 2020–2021 La Niña event (left), a typical year without ENSO activity (2019–2020; centre), and the 2023–2024 El Niño event (right).\n```\n\nattachment:application_climate-pulse_enso_sst.png\n---\nname: application_climate-pulse_enso_sst\n---\nImage exports of monthly sea surface temperature anomaly for the 2020–2021 La Niña event (left), a typical year without ENSO activity (2019–2020; centre), and the 2023–2024 El Niño event (right).\n```\n\nattachment:application_climate-pulse_timeseries_elnino.png\n---\nname: application_climate-pulse_timeseries_elnino\n---\nScreenshot of the Climate Pulse dashboard showing global surface air and sea surface temperature anomalies for 2023 and 2024.\nThe thick coloured lines marked El Niño have been added in this assessment to show when the events occurred; colours correspond to the year.\n```\n\nattachment:application_climate-pulse_timeseries_lanina.png\n---\nname: application_climate-pulse_timeseries_lanina\n---\nScreenshot of the Climate Pulse dashboard showing global surface air and sea surface temperature anomalies for 2019, 2020, and 2021.\nThe thick coloured lines marked La Niña have been added in this assessment to show when the events occurred; colours correspond to the year.\n```\n\n(section-heatwaves)=", "text_with_prefix": "EQC Quality Assessment: \"Visualisation of temperature during extreme weather events with Climate Pulse\"\nDataset: climate-pulse [CDS]\nAspect: extreme-events_q01 | Category: Applications\nSection: Visualisation of temperature during extreme weather events with Climate Pulse > Analysis and results > 1. El Niño–Southern Oscillation (ENSO) > Example: 2023–2024 El Niño event and 2020–2021 La Niña event\n---\nThe 2023–2024 El Niño event was officially declared by WMO on 4 July 2023 and dissipated in April 2024 [[WMO+24](https://wmo.int/news/media-centre/el-nino-forecast-swing-la-nina-later-year), [WMO+23](https://wmo.int/news/media-centre/world-meteorological-organization-declares-onset-of-el-nino-conditions)].\nThe 2020–2021 La Niña event emerged in August 2020 and dissipated in May 2021 [[Li+22](https://doi.org/10.1029/2021JD035546)].\nFigures {numref}`{number} ` and {numref}`{number} ` show global surface air and sea surface temperature anomalies during these periods,\nalong with a typical year without ENSO activity,\nvisualised using Climate Pulse.\nThe similarities with Figure {numref}`{number} ` are evident, particularly the warming (El Niño) and cooling (La Niña) over and in the central and eastern Pacific, and broader global temperature increases.\n\nClimate Pulse is shown to visualise ENSO events effectively by clearly depicting the associated temperature patterns.\nSome variations from the expected patterns exist,\nbut these are characteristic of the event.\n\nFigure {numref}`{number} ` shows the global temperatures for both surface air and sea surface as clearly higher for the El Niño period.\nThe general increase in global temperatures relative to the baseline period (1991–2020) may cause the amplitude of El Niño anomalies to appear exaggerated.\nThe opposite is true for La Niña, which may appear weaker than in reality.\nThis issue can be mitigated by using year-on-year comparisons.\n\nFigure {numref}`{number} ` shows the global average temperatures for 2019 (no ENSO activity), 2020, and 2021.\nThe 2020–2021 La Niña was part of a rare triple-dip La Niña,\nwhich occurred over three consecutive years from 2020;\nhence, comparison with succeeding years is not useful [[Shi+23](https://doi.org/10.1016/j.atmosres.2023.106937)].\nInstead, Figure {numref}`{number} ` compares temperature anomalies with 2019,\nshowing generally lower values in late 2020 and the first half of 2021,\ncorresponding with La Niña.\n\nIn conclusion, Climate Pulse is suitable for visualising El Niño and La Niña events through their temperature anomalies,\nmaking it a valuable tool for a broad audience,\nincluding journalists and communicators,\nseeking to illustrate the climate impacts of such events.\n\nattachment:application_climate-pulse_enso_air.png\n---\nname: application_climate-pulse_enso_air\n---\nImage exports of monthly surface air temperature anomaly for the 2020–2021 La Niña event (left), a typical year without ENSO activity (2019–2020; centre), and the 2023–2024 El Niño event (right).\n```\n\nattachment:application_climate-pulse_enso_sst.png\n---\nname: application_climate-pulse_enso_sst\n---\nImage exports of monthly sea surface temperature anomaly for the 2020–2021 La Niña event (left), a typical year without ENSO activity (2019–2020; centre), and the 2023–2024 El Niño event (right).\n```\n\nattachment:application_climate-pulse_timeseries_elnino.png\n---\nname: application_climate-pulse_timeseries_elnino\n---\nScreenshot of the Climate Pulse dashboard showing global surface air and sea surface temperature anomalies for 2023 and 2024.\nThe thick coloured lines marked El Niño have been added in this assessment to show when the events occurred; colours correspond to the year.\n```\n\nattachment:application_climate-pulse_timeseries_lanina.png\n---\nname: application_climate-pulse_timeseries_lanina\n---\nScreenshot of the Climate Pulse dashboard showing global surface air and sea surface temperature anomalies for 2019, 2020, and 2021.\nThe thick coloured lines marked La Niña have been added in this assessment to show when the events occurred; colours correspond to the year.\n```\n\n(section-heatwaves)="} {"chunk_id": "application_climate-pulse_extreme-events_q01__4b173d69bc53", "report_id": "application_climate-pulse_extreme-events_q01", "dataset_id": "climate-pulse", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extreme-events_q01", "aspect_base": "extreme-events", "category": "Applications", "match_confidence": "unmatched", "section": "Visualisation of temperature during extreme weather events with Climate Pulse > Analysis and results > 2. Heatwaves > Introduction", "title": "Visualisation of temperature during extreme weather events with Climate Pulse", "chunk_index": 8, "token_count": 616, "text_raw": "There is no universal definition of a [heatwave](https://climate.copernicus.eu/heatwaves-brief-introduction),\nwith different organisations and countries having established their own criteria.\nHowever, we can generally define them as a period of time, from a few days up to a few weeks,\nof temperatures significantly higher than the normal range for that location at that time of year\n[[Lhotka+22](https://doi.org/10.1029/2022EA002567), [Stefanon+12](https://doi.org/10.1088/1748-9326/7/1/014023)].\nFor example, the [European State of the Climate 2023](https://climate.copernicus.eu/heatwaves-brief-introduction) used the definition\n\"a period of at least three consecutive days when both the daily surface air temperature minima and maxima are higher than the highest 5% of values for the day in question during the 1991–2020 reference period\".\n\nOne of the main impacts of heatwaves is the danger they pose to human health.\nRoughly 489 000 heat-related deaths occur each year [[Zhao+21](https://doi.org/10.1016/S2542-5196%2821%2900081-4)],\nand high-intensity heatwave events are known to increase mortality rates.\nFor example,\nin Europe alone, in the summer of 2022,\nan estimated 61 672 heat-related excess deaths occurred [[Ballester+23](https://doi.org/10.1038/s41591-023-02419-z)].\nDue to these impacts, heatwaves often generate significant news coverage.\nThe June 2024 Eastern Mediterranean heatwave led to real-time reporting of its increasing death toll,\nas well as journalistic investigations into the overall impact and the drivers of this extreme event\n[[Bali+24](https://www.dw.com/en/fatal-heat-wave-sweeps-greece-claims-more-lives/a-69485803), [World Weather Attribution+24](https://www.worldweatherattribution.org/deadly-mediterranean-heatwave-would-not-have-occurred-without-human-induced-climate-change/)].\n\nUsers interested in how humans experience heat and cold waves,\nincluding feels-like temperatures and thermal stress indicators,\nmay want to use the [_Thermal Trace_ application](https://apps.climate.copernicus.eu/overview?app=thermal-trace),\nfor which there is an equivalent [quality assessment](./application_thermaltrace_extreme-events_q01) investigating its suitability for visualising extreme weather events.", "text_with_prefix": "EQC Quality Assessment: \"Visualisation of temperature during extreme weather events with Climate Pulse\"\nDataset: climate-pulse [CDS]\nAspect: extreme-events_q01 | Category: Applications\nSection: Visualisation of temperature during extreme weather events with Climate Pulse > Analysis and results > 2. Heatwaves > Introduction\n---\nThere is no universal definition of a [heatwave](https://climate.copernicus.eu/heatwaves-brief-introduction),\nwith different organisations and countries having established their own criteria.\nHowever, we can generally define them as a period of time, from a few days up to a few weeks,\nof temperatures significantly higher than the normal range for that location at that time of year\n[[Lhotka+22](https://doi.org/10.1029/2022EA002567), [Stefanon+12](https://doi.org/10.1088/1748-9326/7/1/014023)].\nFor example, the [European State of the Climate 2023](https://climate.copernicus.eu/heatwaves-brief-introduction) used the definition\n\"a period of at least three consecutive days when both the daily surface air temperature minima and maxima are higher than the highest 5% of values for the day in question during the 1991–2020 reference period\".\n\nOne of the main impacts of heatwaves is the danger they pose to human health.\nRoughly 489 000 heat-related deaths occur each year [[Zhao+21](https://doi.org/10.1016/S2542-5196%2821%2900081-4)],\nand high-intensity heatwave events are known to increase mortality rates.\nFor example,\nin Europe alone, in the summer of 2022,\nan estimated 61 672 heat-related excess deaths occurred [[Ballester+23](https://doi.org/10.1038/s41591-023-02419-z)].\nDue to these impacts, heatwaves often generate significant news coverage.\nThe June 2024 Eastern Mediterranean heatwave led to real-time reporting of its increasing death toll,\nas well as journalistic investigations into the overall impact and the drivers of this extreme event\n[[Bali+24](https://www.dw.com/en/fatal-heat-wave-sweeps-greece-claims-more-lives/a-69485803), [World Weather Attribution+24](https://www.worldweatherattribution.org/deadly-mediterranean-heatwave-would-not-have-occurred-without-human-induced-climate-change/)].\n\nUsers interested in how humans experience heat and cold waves,\nincluding feels-like temperatures and thermal stress indicators,\nmay want to use the [_Thermal Trace_ application](https://apps.climate.copernicus.eu/overview?app=thermal-trace),\nfor which there is an equivalent [quality assessment](./application_thermaltrace_extreme-events_q01) investigating its suitability for visualising extreme weather events."} {"chunk_id": "application_climate-pulse_extreme-events_q01__895079661a5c", "report_id": "application_climate-pulse_extreme-events_q01", "dataset_id": "climate-pulse", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extreme-events_q01", "aspect_base": "extreme-events", "category": "Applications", "match_confidence": "unmatched", "section": "Visualisation of temperature during extreme weather events with Climate Pulse > Analysis and results > 2. Heatwaves > Assessment of Climate Pulse", "title": "Visualisation of temperature during extreme weather events with Climate Pulse", "chunk_index": 9, "token_count": 539, "text_raw": "Climate Pulse is determined to be moderately suitable for visualising the temperature impacts of heatwaves, depending on the date and extent of the heatwave, for the following reasons:\n* **Variables:** Climate Pulse allows for the visualisation of temperature anomalies against WMO’s standard reference period, and significant regional warming anomalies are the key indicator of heatwaves.\n* **Data Granularity:** Climate Pulse’s daily maps have anomalies on a scale from –12 °C to +12 °C, with approximately 1 °C increments for surface air temperature. This is in line with the temperature anomalies experienced during a heatwave, with regional anomalies usually being around 3 to 8 °C of warming ([for Europe](https://www.ecmwf.int/en/about/media-centre/news/2023/european-climate-marked-heat-and-drought-2022-report)) [[Ma+20](https://doi.org/10.1029/2020GL087809)].\n* **Spatial Coverage and Resolution:** Climate Pulse uses a 0.25° × 0.25° grid, which is suitable for showing regional, national, and multi-country heatwave events. While exported maps are global by default, one can manually zoom in on the globe and take screenshots at different scales. The global time series is unsuitable for visualising heatwaves as they are regional, not global events.\n* **Temporal Coverage and Resolution:** Daily air temperature anomaly maps, matching the typical duration of heatwaves, are available in Climate Pulse from 1 January 2023 onwards. Before this date, only monthly and annual maps are available, limiting the usefulness of the application for visualising short historical heatwaves, which may be smoothed out and hence not visible over these time frames.\n* **Timeliness:** The near real-time updates of Climate Pulse, typically 2 days behind, allow for timely visualisation of the heatwaves currently taking place, supporting near-real-time weather monitoring and communication.\n* **Uncertainty:** Climate Pulse does not show uncertainty estimates in the time series or maps. The [ERA5 dataset has an uncertainty](https://confluence.ecmwf.int/x/ZSOgBg) of around 0.5 °C. This is suitable for the use case of visualising heatwave events, which cause significantly larger temperature anomalies.", "text_with_prefix": "EQC Quality Assessment: \"Visualisation of temperature during extreme weather events with Climate Pulse\"\nDataset: climate-pulse [CDS]\nAspect: extreme-events_q01 | Category: Applications\nSection: Visualisation of temperature during extreme weather events with Climate Pulse > Analysis and results > 2. Heatwaves > Assessment of Climate Pulse\n---\nClimate Pulse is determined to be moderately suitable for visualising the temperature impacts of heatwaves, depending on the date and extent of the heatwave, for the following reasons:\n* **Variables:** Climate Pulse allows for the visualisation of temperature anomalies against WMO’s standard reference period, and significant regional warming anomalies are the key indicator of heatwaves.\n* **Data Granularity:** Climate Pulse’s daily maps have anomalies on a scale from –12 °C to +12 °C, with approximately 1 °C increments for surface air temperature. This is in line with the temperature anomalies experienced during a heatwave, with regional anomalies usually being around 3 to 8 °C of warming ([for Europe](https://www.ecmwf.int/en/about/media-centre/news/2023/european-climate-marked-heat-and-drought-2022-report)) [[Ma+20](https://doi.org/10.1029/2020GL087809)].\n* **Spatial Coverage and Resolution:** Climate Pulse uses a 0.25° × 0.25° grid, which is suitable for showing regional, national, and multi-country heatwave events. While exported maps are global by default, one can manually zoom in on the globe and take screenshots at different scales. The global time series is unsuitable for visualising heatwaves as they are regional, not global events.\n* **Temporal Coverage and Resolution:** Daily air temperature anomaly maps, matching the typical duration of heatwaves, are available in Climate Pulse from 1 January 2023 onwards. Before this date, only monthly and annual maps are available, limiting the usefulness of the application for visualising short historical heatwaves, which may be smoothed out and hence not visible over these time frames.\n* **Timeliness:** The near real-time updates of Climate Pulse, typically 2 days behind, allow for timely visualisation of the heatwaves currently taking place, supporting near-real-time weather monitoring and communication.\n* **Uncertainty:** Climate Pulse does not show uncertainty estimates in the time series or maps. The [ERA5 dataset has an uncertainty](https://confluence.ecmwf.int/x/ZSOgBg) of around 0.5 °C. This is suitable for the use case of visualising heatwave events, which cause significantly larger temperature anomalies."} {"chunk_id": "application_climate-pulse_extreme-events_q01__42697b6b2f5b", "report_id": "application_climate-pulse_extreme-events_q01", "dataset_id": "climate-pulse", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extreme-events_q01", "aspect_base": "extreme-events", "category": "Applications", "match_confidence": "unmatched", "section": "Visualisation of temperature during extreme weather events with Climate Pulse > Analysis and results > 2. Heatwaves > Example: Southeastern Europe June 2024 Heatwave", "title": "Visualisation of temperature during extreme weather events with Climate Pulse", "chunk_index": 10, "token_count": 471, "text_raw": "Between 3 and 27 June 2024,\nseveral countries in Southeastern Europe and part of the Middle East were affected by a heatwave,\nwith the most extreme temperatures occurring between 11 and 14 June [[Dafis+24](https://www.climameter.org/20240611-14-june-eastern-mediterranean-heatwave)].\nFigure {numref}`{number} ` shows the image exports from Climate Pulse;\nthese are global maps that are useful to show where large-scale events are occurring,\nbut less useful for interpreting regional and national effects.\nFigure {numref}`{number} ` also displays the June 2024 average,\nwhere the heatwave event is visible due to its extended duration.\nThis visibility typically occurs only with longer heatwaves, not shorter ones.\nIn Figure {numref}`{number} `,\nthe region of interest has been magnified,\nmaking details in the location of the temperature anomalies more visible.\nThese magnified anomaly maps are more useful for visualising heatwave events with the Climate Pulse application,\nbut the image exports are still suitable.\n\nattachment:application_climate-pulse_heatwave_global.png\n---\nname: application_climate-pulse_heatwave_global\n---\nImage exports of the global surface air temperature anomaly over the 11 to 14 June 2024 peak heatwave period (top and middle) and averaged over June 2024 (bottom).\n```\n\nattachment:application_climate-pulse_heatwave_regional.png\n---\nname: application_climate-pulse_heatwave_regional\n---\nMagnified regions of interest of the global surface air temperature anomaly over the 11 to 14 June 2024 peak heatwave period. Note: these are screenshots of the Climate Pulse dashboard.\n```\n\n(section-volcanoes)=", "text_with_prefix": "EQC Quality Assessment: \"Visualisation of temperature during extreme weather events with Climate Pulse\"\nDataset: climate-pulse [CDS]\nAspect: extreme-events_q01 | Category: Applications\nSection: Visualisation of temperature during extreme weather events with Climate Pulse > Analysis and results > 2. Heatwaves > Example: Southeastern Europe June 2024 Heatwave\n---\nBetween 3 and 27 June 2024,\nseveral countries in Southeastern Europe and part of the Middle East were affected by a heatwave,\nwith the most extreme temperatures occurring between 11 and 14 June [[Dafis+24](https://www.climameter.org/20240611-14-june-eastern-mediterranean-heatwave)].\nFigure {numref}`{number} ` shows the image exports from Climate Pulse;\nthese are global maps that are useful to show where large-scale events are occurring,\nbut less useful for interpreting regional and national effects.\nFigure {numref}`{number} ` also displays the June 2024 average,\nwhere the heatwave event is visible due to its extended duration.\nThis visibility typically occurs only with longer heatwaves, not shorter ones.\nIn Figure {numref}`{number} `,\nthe region of interest has been magnified,\nmaking details in the location of the temperature anomalies more visible.\nThese magnified anomaly maps are more useful for visualising heatwave events with the Climate Pulse application,\nbut the image exports are still suitable.\n\nattachment:application_climate-pulse_heatwave_global.png\n---\nname: application_climate-pulse_heatwave_global\n---\nImage exports of the global surface air temperature anomaly over the 11 to 14 June 2024 peak heatwave period (top and middle) and averaged over June 2024 (bottom).\n```\n\nattachment:application_climate-pulse_heatwave_regional.png\n---\nname: application_climate-pulse_heatwave_regional\n---\nMagnified regions of interest of the global surface air temperature anomaly over the 11 to 14 June 2024 peak heatwave period. Note: these are screenshots of the Climate Pulse dashboard.\n```\n\n(section-volcanoes)="} {"chunk_id": "application_climate-pulse_extreme-events_q01__74ec2efd5566", "report_id": "application_climate-pulse_extreme-events_q01", "dataset_id": "climate-pulse", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extreme-events_q01", "aspect_base": "extreme-events", "category": "Applications", "match_confidence": "unmatched", "section": "Visualisation of temperature during extreme weather events with Climate Pulse > Analysis and results > 3. Volcanic Eruptions > Introduction", "title": "Visualisation of temperature during extreme weather events with Climate Pulse", "chunk_index": 11, "token_count": 704, "text_raw": "A [volcanic eruption](https://www.nps.gov/subjects/volcanoes/volcanic-eruptions.htm) is a sudden release of gas, ash, rock fragments, and/or molten lava through a vent onto the Earth's surface and/or into the atmosphere.\nBeyond their dramatic local impacts,\nlarge volcanic eruptions can influence global climate through the injection of sulfur dioxide (SO2) into the stratosphere,\nthe second major layer of Earth's atmosphere\n– above the troposphere –\nextending from approximately 11 to 50 km in altitude.\nOnce in the stratosphere,\nsulfur gases oxidise into sulfate aerosols, which reflect incoming solar radiation [[Cole-Dai+10](https://doi.org/10.1002/wcc.76)].\n\nIn the short term, volcanic eruptions can disrupt local tropospheric conditions, typically over a few days.\nThese effects include a reduction in the diurnal temperature cycle,\nwith cooling in the day due to reduced solar radiation and warming in the night from trapped infrared radiation.\nThe magnitude of this effect depends on the size and location of the eruption,\nbut can be around 15 °C [[Robock+00](https://doi.org/10.1029/1998RG000054)].\nOn a global scale, the long-term impact of stratospheric volcanic eruptions is tropospheric cooling,\nusually up to around 0.5 °C,\nlasting from a few months up to two years,\nagain dependent on the nature and scale of the eruption [[Robock+00](https://doi.org/10.1029/1998RG000054)].\n\nConversely,\nthe immediate stratospheric response to a volcanic eruption is localised heating.\nThis results from the absorption of infrared solar radiation at the top of the stratospheric layer and terrestrial radiation at its base,\nfollowing the injection of aerosols.\nHowever,\nthis effect only lasts a few weeks,\ndepending on the eruption magnitude and atmospheric circulation patterns,\nas the aerosols will mix through the stratosphere globally [[Robock+13](https://doi.org/10.1002/2013EO350001)].\nLarge eruptions may cause stratospheric temperature increases of a few degrees [[Boretti+24](https://doi.org/10.1016/j.jastp.2024.106187), [Robock+00](https://doi.org/10.1029/1998RG000054)].\nWhile the global signal is one of net surface cooling and net stratospheric heating,\nthe regional climate response to volcanic eruptions can be more complex.\n\nVolcanic eruptions are widely reported in the media due to their immediate and often severe impacts on local populations.\nHowever, media coverage also frequently highlights the broader implications of such events,\nincluding their effects on global surface temperatures and the insights they offer into climate processes and what can be learnt from volcanic eruptions for potential geoengineering strategies [[King+24](https://www.bbc.co.uk/news/articles/c98qp79gj4no)].", "text_with_prefix": "EQC Quality Assessment: \"Visualisation of temperature during extreme weather events with Climate Pulse\"\nDataset: climate-pulse [CDS]\nAspect: extreme-events_q01 | Category: Applications\nSection: Visualisation of temperature during extreme weather events with Climate Pulse > Analysis and results > 3. Volcanic Eruptions > Introduction\n---\nA [volcanic eruption](https://www.nps.gov/subjects/volcanoes/volcanic-eruptions.htm) is a sudden release of gas, ash, rock fragments, and/or molten lava through a vent onto the Earth's surface and/or into the atmosphere.\nBeyond their dramatic local impacts,\nlarge volcanic eruptions can influence global climate through the injection of sulfur dioxide (SO2) into the stratosphere,\nthe second major layer of Earth's atmosphere\n– above the troposphere –\nextending from approximately 11 to 50 km in altitude.\nOnce in the stratosphere,\nsulfur gases oxidise into sulfate aerosols, which reflect incoming solar radiation [[Cole-Dai+10](https://doi.org/10.1002/wcc.76)].\n\nIn the short term, volcanic eruptions can disrupt local tropospheric conditions, typically over a few days.\nThese effects include a reduction in the diurnal temperature cycle,\nwith cooling in the day due to reduced solar radiation and warming in the night from trapped infrared radiation.\nThe magnitude of this effect depends on the size and location of the eruption,\nbut can be around 15 °C [[Robock+00](https://doi.org/10.1029/1998RG000054)].\nOn a global scale, the long-term impact of stratospheric volcanic eruptions is tropospheric cooling,\nusually up to around 0.5 °C,\nlasting from a few months up to two years,\nagain dependent on the nature and scale of the eruption [[Robock+00](https://doi.org/10.1029/1998RG000054)].\n\nConversely,\nthe immediate stratospheric response to a volcanic eruption is localised heating.\nThis results from the absorption of infrared solar radiation at the top of the stratospheric layer and terrestrial radiation at its base,\nfollowing the injection of aerosols.\nHowever,\nthis effect only lasts a few weeks,\ndepending on the eruption magnitude and atmospheric circulation patterns,\nas the aerosols will mix through the stratosphere globally [[Robock+13](https://doi.org/10.1002/2013EO350001)].\nLarge eruptions may cause stratospheric temperature increases of a few degrees [[Boretti+24](https://doi.org/10.1016/j.jastp.2024.106187), [Robock+00](https://doi.org/10.1029/1998RG000054)].\nWhile the global signal is one of net surface cooling and net stratospheric heating,\nthe regional climate response to volcanic eruptions can be more complex.\n\nVolcanic eruptions are widely reported in the media due to their immediate and often severe impacts on local populations.\nHowever, media coverage also frequently highlights the broader implications of such events,\nincluding their effects on global surface temperatures and the insights they offer into climate processes and what can be learnt from volcanic eruptions for potential geoengineering strategies [[King+24](https://www.bbc.co.uk/news/articles/c98qp79gj4no)]."} {"chunk_id": "application_climate-pulse_extreme-events_q01__a638870bc03b", "report_id": "application_climate-pulse_extreme-events_q01", "dataset_id": "climate-pulse", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extreme-events_q01", "aspect_base": "extreme-events", "category": "Applications", "match_confidence": "unmatched", "section": "Visualisation of temperature during extreme weather events with Climate Pulse > Analysis and results > 3. Volcanic Eruptions > Assessment of Climate Pulse", "title": "Visualisation of temperature during extreme weather events with Climate Pulse", "chunk_index": 12, "token_count": 370, "text_raw": "Climate Pulse is determined not to be suitable for visualising the temperature impacts of volcanic eruptions, for the following reasons:\n* **Vertical resolution:** Climate Pulse only displays air temperature 2 m above the surface. Many of the effects of major volcanic eruptions take place in the stratosphere; these are not captured by near-surface temperature data.\n* **Confounding factors:** Climate Pulse uses a 1991–2020 baseline for the calculation of temperature anomalies, a period that includes warming from climate change. When analysing historical volcanic eruptions like Pinatubo (1991), this reference period dampens or distorts the apparent magnitude of cooling following the eruption. A baseline more representative of conditions at the time would more accurately reflect the cooling effect of an eruption, although it may be difficult to disentangle from other year-on-year differences such as ENSO events.\n* **Temporal resolution:** The annual and monthly maps in Climate Pulse cannot visualise the immediate, local surface air temperature impact of an eruption effectively, since the local effects only last a few days and are thus washed out in longer-term maps. However, with Climate Pulse now providing daily maps from January 2023, any major future volcanic eruptions should prompt an assessment of its suitability for visualising the associated localised, short-term surface cooling effects. The same is true for the sensitivity of the underpinning ERA5 dataset to the near-surface temperature impacts of sudden events like volcanic eruptions.", "text_with_prefix": "EQC Quality Assessment: \"Visualisation of temperature during extreme weather events with Climate Pulse\"\nDataset: climate-pulse [CDS]\nAspect: extreme-events_q01 | Category: Applications\nSection: Visualisation of temperature during extreme weather events with Climate Pulse > Analysis and results > 3. Volcanic Eruptions > Assessment of Climate Pulse\n---\nClimate Pulse is determined not to be suitable for visualising the temperature impacts of volcanic eruptions, for the following reasons:\n* **Vertical resolution:** Climate Pulse only displays air temperature 2 m above the surface. Many of the effects of major volcanic eruptions take place in the stratosphere; these are not captured by near-surface temperature data.\n* **Confounding factors:** Climate Pulse uses a 1991–2020 baseline for the calculation of temperature anomalies, a period that includes warming from climate change. When analysing historical volcanic eruptions like Pinatubo (1991), this reference period dampens or distorts the apparent magnitude of cooling following the eruption. A baseline more representative of conditions at the time would more accurately reflect the cooling effect of an eruption, although it may be difficult to disentangle from other year-on-year differences such as ENSO events.\n* **Temporal resolution:** The annual and monthly maps in Climate Pulse cannot visualise the immediate, local surface air temperature impact of an eruption effectively, since the local effects only last a few days and are thus washed out in longer-term maps. However, with Climate Pulse now providing daily maps from January 2023, any major future volcanic eruptions should prompt an assessment of its suitability for visualising the associated localised, short-term surface cooling effects. The same is true for the sensitivity of the underpinning ERA5 dataset to the near-surface temperature impacts of sudden events like volcanic eruptions."} {"chunk_id": "application_climate-pulse_extreme-events_q01__3ff71b932aa8", "report_id": "application_climate-pulse_extreme-events_q01", "dataset_id": "climate-pulse", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extreme-events_q01", "aspect_base": "extreme-events", "category": "Applications", "match_confidence": "unmatched", "section": "Visualisation of temperature during extreme weather events with Climate Pulse > Analysis and results > 3. Volcanic Eruptions > Example: Mount Pinatubo (June 1991)", "title": "Visualisation of temperature during extreme weather events with Climate Pulse", "chunk_index": 13, "token_count": 502, "text_raw": "The volcanic eruption of Mount Pinatubo in the Philippines occurred on 15 June 1991,\nreleasing approximately 15 million tons of SO2 into the stratosphere [[Boretti+24](https://doi.org/10.1016/j.jastp.2024.106187)],\nforming a global layer of aerosols in weeks.\nThese aerosols reflected incoming solar radiation,\nreducing global surface temperatures by an average of about 0.4 °C during 1992–1993 [[Boretti+24](https://doi.org/10.1016/j.jastp.2024.106187)],\nmaking it one of the most notable short-term climate impacts of a volcanic eruption in the 20th century.\nThe aerosol layer gradually dissipated over the following three years [[Minnis+93](https://doi.org/10.1126/science.259.5100.1411)].\nThere is no information available on the localised short-term effects of the eruption;\nhowever, due to the temporal availability of the data (monthly), the short-term effects would not be visible in Climate Pulse.\n\nFigure {numref}`{number} ` shows lower global average temperatures for 1992 and 1993 compared with 1991.\nThese years are known to have experienced cooling influenced by the volcanic eruption.\nHowever,\nfurther analysis is needed to distinguish the impact of the eruption from other contributing factors,\nsuch as ENSO events.\nTherefore, it is not possible to directly identify the temperature impact of the volcanic eruption based solely on Figure {numref}`{number} `.\n\nattachment:application_climate-pulse_pinatubo_timeseries.png\n---\nname: application_climate-pulse_pinatubo_timeseries\n---\nScreenshot of the Climate Pulse dashboard showing global average surface air temperature anomalies for 1990–1993. Mount Pinatubo erupted in June 1991.\n```", "text_with_prefix": "EQC Quality Assessment: \"Visualisation of temperature during extreme weather events with Climate Pulse\"\nDataset: climate-pulse [CDS]\nAspect: extreme-events_q01 | Category: Applications\nSection: Visualisation of temperature during extreme weather events with Climate Pulse > Analysis and results > 3. Volcanic Eruptions > Example: Mount Pinatubo (June 1991)\n---\nThe volcanic eruption of Mount Pinatubo in the Philippines occurred on 15 June 1991,\nreleasing approximately 15 million tons of SO2 into the stratosphere [[Boretti+24](https://doi.org/10.1016/j.jastp.2024.106187)],\nforming a global layer of aerosols in weeks.\nThese aerosols reflected incoming solar radiation,\nreducing global surface temperatures by an average of about 0.4 °C during 1992–1993 [[Boretti+24](https://doi.org/10.1016/j.jastp.2024.106187)],\nmaking it one of the most notable short-term climate impacts of a volcanic eruption in the 20th century.\nThe aerosol layer gradually dissipated over the following three years [[Minnis+93](https://doi.org/10.1126/science.259.5100.1411)].\nThere is no information available on the localised short-term effects of the eruption;\nhowever, due to the temporal availability of the data (monthly), the short-term effects would not be visible in Climate Pulse.\n\nFigure {numref}`{number} ` shows lower global average temperatures for 1992 and 1993 compared with 1991.\nThese years are known to have experienced cooling influenced by the volcanic eruption.\nHowever,\nfurther analysis is needed to distinguish the impact of the eruption from other contributing factors,\nsuch as ENSO events.\nTherefore, it is not possible to directly identify the temperature impact of the volcanic eruption based solely on Figure {numref}`{number} `.\n\nattachment:application_climate-pulse_pinatubo_timeseries.png\n---\nname: application_climate-pulse_pinatubo_timeseries\n---\nScreenshot of the Climate Pulse dashboard showing global average surface air temperature anomalies for 1990–1993. Mount Pinatubo erupted in June 1991.\n```"} {"chunk_id": "application_climate-pulse_extreme-events_q01__ef81008a2ecf", "report_id": "application_climate-pulse_extreme-events_q01", "dataset_id": "climate-pulse", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extreme-events_q01", "aspect_base": "extreme-events", "category": "Applications", "match_confidence": "unmatched", "section": "Visualisation of temperature during extreme weather events with Climate Pulse > Analysis and results > 3. Volcanic Eruptions > Example: 2022 Hunga Tonga–Hunga Haʻapai eruption", "title": "Visualisation of temperature during extreme weather events with Climate Pulse", "chunk_index": 14, "token_count": 690, "text_raw": "On 15 January 2022,\nthe [Hunga Tonga–Hunga Haʻapai volcano](https://www.ecmwf.int/en/about/media-centre/news/2022/hunga-tonga-eruption-seen-ecmwf) erupted explosively after weeks of activity,\ninjecting volcanic gases and materials into the stratosphere and generating atmospheric shock waves and tsunami waves across multiple ocean basins.\nUnlike the Mount Pinatubo eruption,\nthe Hunga eruption led to a cooling of global stratospheric temperatures by 0.5 to 1.0 °C from 2022 to mid-2023 due to the injection of large amounts of water vapour and minimal sulfur dioxide [[Randel+24](https://doi.org/10.1029/2024GL111500)].\nWhile this excess water vapour is believed to have contributed to global warming through enhanced radiative forcing,\nits precise impact remains uncertain due to the influence of other factors like ongoing climate change and ENSO events [[Gupta+25](https://doi.org/10.1038/s43247-025-02181-9)].\nWhether the eruption caused surface temperature warming or cooling is still being debated in the scientific literature [[Gupta+25](https://doi.org/10.1038/s43247-025-02181-9), [Jenkins+23](https://doi.org/10.1038/s41558-022-01568-2)].\n\nLocalised monthly maps (Figure {numref}`{number} `) do not show any surface air temperature effect of the eruption,\nas the local effects are very short-term;\ndaily or higher spatial resolution maps would be required to visualise such an effect.\nFigure {numref}`{number} ` shows the global average surface air temperature time series.\nAny potential warming or cooling effects of the eruption cannot be disentangled from the other contributing factors to global temperatures,\nsuch as ENSO events.\nHence, overall Climate Pulse is not suitable for visualising the air temperature impacts of this event.\n\nattachment:application_climate-pulse_tonga_map.png\n---\nname: application_climate-pulse_tonga_map\n---\nMagnified region of interest of the global surface air temperature anomaly for the months before (December 2021), during (January 2022), and after (February and March 2022) the eruption. Note: these are screenshots of the Climate Pulse dashboard.\n```\n\nattachment:application_climate-pulse_tonga_timeseries.png\n---\nname: application_climate-pulse_tonga_timeseries\n---\nScreenshot of the Climate Pulse dashboard of average surface air temperature anomaly, for 2021: the year before the eruption for comparison, 2022: the year of the eruption, 2023: the year after the eruption.\n```", "text_with_prefix": "EQC Quality Assessment: \"Visualisation of temperature during extreme weather events with Climate Pulse\"\nDataset: climate-pulse [CDS]\nAspect: extreme-events_q01 | Category: Applications\nSection: Visualisation of temperature during extreme weather events with Climate Pulse > Analysis and results > 3. Volcanic Eruptions > Example: 2022 Hunga Tonga–Hunga Haʻapai eruption\n---\nOn 15 January 2022,\nthe [Hunga Tonga–Hunga Haʻapai volcano](https://www.ecmwf.int/en/about/media-centre/news/2022/hunga-tonga-eruption-seen-ecmwf) erupted explosively after weeks of activity,\ninjecting volcanic gases and materials into the stratosphere and generating atmospheric shock waves and tsunami waves across multiple ocean basins.\nUnlike the Mount Pinatubo eruption,\nthe Hunga eruption led to a cooling of global stratospheric temperatures by 0.5 to 1.0 °C from 2022 to mid-2023 due to the injection of large amounts of water vapour and minimal sulfur dioxide [[Randel+24](https://doi.org/10.1029/2024GL111500)].\nWhile this excess water vapour is believed to have contributed to global warming through enhanced radiative forcing,\nits precise impact remains uncertain due to the influence of other factors like ongoing climate change and ENSO events [[Gupta+25](https://doi.org/10.1038/s43247-025-02181-9)].\nWhether the eruption caused surface temperature warming or cooling is still being debated in the scientific literature [[Gupta+25](https://doi.org/10.1038/s43247-025-02181-9), [Jenkins+23](https://doi.org/10.1038/s41558-022-01568-2)].\n\nLocalised monthly maps (Figure {numref}`{number} `) do not show any surface air temperature effect of the eruption,\nas the local effects are very short-term;\ndaily or higher spatial resolution maps would be required to visualise such an effect.\nFigure {numref}`{number} ` shows the global average surface air temperature time series.\nAny potential warming or cooling effects of the eruption cannot be disentangled from the other contributing factors to global temperatures,\nsuch as ENSO events.\nHence, overall Climate Pulse is not suitable for visualising the air temperature impacts of this event.\n\nattachment:application_climate-pulse_tonga_map.png\n---\nname: application_climate-pulse_tonga_map\n---\nMagnified region of interest of the global surface air temperature anomaly for the months before (December 2021), during (January 2022), and after (February and March 2022) the eruption. Note: these are screenshots of the Climate Pulse dashboard.\n```\n\nattachment:application_climate-pulse_tonga_timeseries.png\n---\nname: application_climate-pulse_tonga_timeseries\n---\nScreenshot of the Climate Pulse dashboard of average surface air temperature anomaly, for 2021: the year before the eruption for comparison, 2022: the year of the eruption, 2023: the year after the eruption.\n```"} {"chunk_id": "application_climate-pulse_extreme-events_q01__744872bff3d8", "report_id": "application_climate-pulse_extreme-events_q01", "dataset_id": "climate-pulse", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extreme-events_q01", "aspect_base": "extreme-events", "category": "Applications", "match_confidence": "unmatched", "section": "Visualisation of temperature during extreme weather events with Climate Pulse > ℹ️ If you want to know more > Key resources", "title": "Visualisation of temperature during extreme weather events with Climate Pulse", "chunk_index": 15, "token_count": 481, "text_raw": "More about the ERA5 reanalysis:\n* [The ERA5 global reanalysis](https://doi.org/10.1002/qj.3803)\n\nMore about temperature anomalies:\n* [WMO Climatological Normals](https://community.wmo.int/en/activity-areas/climate-services/climate-products-and-initiatives/wmo-climatological-normals)\n\nMore about ENSO:\n* [C3S NINO3.4 SST Forecast](https://climate.copernicus.eu/charts/packages/c3s_seasonal/products/c3s_seasonal_plume_mm?area=nino34&base_time=202507010000&type=plume)\n* [ENSO as an Integrating Concept in Earth Science](https://doi.org/10.1126/science.1132588)\n* [NOAA: \"What are El Niño and La Niña?\"](https://oceanservice.noaa.gov/facts/ninonina.html)\n\nMore about heatwaves:\n* [Heatwaves – a brief introduction](https://climate.copernicus.eu/heatwaves-brief-introduction)\n* [Heat Waves: Physical Understanding and Scientific Challenges](https://doi.org/10.1029/2022RG000780)\n* [World Health Organization: \"Heat and health\"](https://www.who.int/news-room/fact-sheets/detail/climate-change-heat-and-health)\n* [Thermal Trace application](https://apps.climate.copernicus.eu/overview?app=thermal-trace) for visualising feels-like temperatures\n * [](./application_thermaltrace_extreme-events_q01)\n\nMore about volcanic eruptions:\n* [The Hunga Tonga eruption as seen by ECMWF](https://www.ecmwf.int/en/about/media-centre/news/2022/hunga-tonga-eruption-seen-ecmwf)\n* [Volcanic eruptions and climate](https://doi.org/10.1029/1998RG000054)", "text_with_prefix": "EQC Quality Assessment: \"Visualisation of temperature during extreme weather events with Climate Pulse\"\nDataset: climate-pulse [CDS]\nAspect: extreme-events_q01 | Category: Applications\nSection: Visualisation of temperature during extreme weather events with Climate Pulse > ℹ️ If you want to know more > Key resources\n---\nMore about the ERA5 reanalysis:\n* [The ERA5 global reanalysis](https://doi.org/10.1002/qj.3803)\n\nMore about temperature anomalies:\n* [WMO Climatological Normals](https://community.wmo.int/en/activity-areas/climate-services/climate-products-and-initiatives/wmo-climatological-normals)\n\nMore about ENSO:\n* [C3S NINO3.4 SST Forecast](https://climate.copernicus.eu/charts/packages/c3s_seasonal/products/c3s_seasonal_plume_mm?area=nino34&base_time=202507010000&type=plume)\n* [ENSO as an Integrating Concept in Earth Science](https://doi.org/10.1126/science.1132588)\n* [NOAA: \"What are El Niño and La Niña?\"](https://oceanservice.noaa.gov/facts/ninonina.html)\n\nMore about heatwaves:\n* [Heatwaves – a brief introduction](https://climate.copernicus.eu/heatwaves-brief-introduction)\n* [Heat Waves: Physical Understanding and Scientific Challenges](https://doi.org/10.1029/2022RG000780)\n* [World Health Organization: \"Heat and health\"](https://www.who.int/news-room/fact-sheets/detail/climate-change-heat-and-health)\n* [Thermal Trace application](https://apps.climate.copernicus.eu/overview?app=thermal-trace) for visualising feels-like temperatures\n * [](./application_thermaltrace_extreme-events_q01)\n\nMore about volcanic eruptions:\n* [The Hunga Tonga eruption as seen by ECMWF](https://www.ecmwf.int/en/about/media-centre/news/2022/hunga-tonga-eruption-seen-ecmwf)\n* [Volcanic eruptions and climate](https://doi.org/10.1029/1998RG000054)"} {"chunk_id": "application_climate-pulse_extreme-events_q01__b1877f5b5deb", "report_id": "application_climate-pulse_extreme-events_q01", "dataset_id": "climate-pulse", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extreme-events_q01", "aspect_base": "extreme-events", "category": "Applications", "match_confidence": "unmatched", "section": "Visualisation of temperature during extreme weather events with Climate Pulse > ℹ️ If you want to know more > References", "title": "Visualisation of temperature during extreme weather events with Climate Pulse", "chunk_index": 16, "token_count": 1052, "text_raw": "[[AFP+24](https://timesofmalta.com/article/february-marks-9th-straight-month-recordsmashing-global-heat.1087990)] AFP, ‘February marks 9th straight month of record-smashing global heat’, Times of Malta, Mar. 07, 2024. Accessed: May 14, 2026. [Online]. Available: https://timesofmalta.com/article/february-marks-9th-straight-month-recordsmashing-global-heat.1087990\n\n[[Bali+24](https://www.dw.com/en/fatal-heat-wave-sweeps-greece-claims-more-lives/a-69485803)] K. Bali, ‘Heat wave in Greece claims more tourist lives’, DW, Jun. 26, 2024. Accessed: May 14, 2026. [Online]. Available: https://www.dw.com/en/fatal-heat-wave-sweeps-greece-claims-more-lives/a-69485803\n\n[[Ballester+23](https://doi.org/10.1038/s41591-023-02419-z)] J. Ballester et al., ‘Heat-related mortality in Europe during the summer of 2022’, Nature Medicine, vol. 29, no. 7, pp. 1857–1866, Jul. 2023, doi: 10.1038/s41591-023-02419-z.\n\n[[Bell+18](https://doi.org/10.1080/10962247.2017.1401017)] J. E. Bell et al., ‘Changes in extreme events and the potential impacts on human health’, Journal of the Air & Waste Management Association, vol. 68, no. 4, pp. 265–287, Apr. 2018, doi: 10.1080/10962247.2017.1401017.\n\n[[Boretti+24](https://doi.org/10.1016/j.jastp.2024.106187)] A. Boretti, ‘Reassessing the cooling that followed the 1991 volcanic eruption of Mt. Pinatubo’, Journal of Atmospheric and Solar-Terrestrial Physics, vol. 256, p. 106187, Mar. 2024, doi: 10.1016/j.jastp.2024.106187.\n\n[[Cole-Dai+10](https://doi.org/10.1002/wcc.76)] J. Cole‐Dai, ‘Volcanoes and climate’, WIREs Climate Change, vol. 1, no. 6, pp. 824–839, Nov. 2010, doi: 10.1002/wcc.76.\n\n[[Dafis+24](https://www.climameter.org/20240611-14-june-eastern-mediterranean-heatwave)] S. Dafis and D. Faranda, ‘The June 2024 Eastern Mediterranean Heatwave likely exacerbated by human-driven Climate Change and Natural Variability’, 2024, doi: 10.5281/zenodo.14103941.\n\n[[Gerretsen+23](https://www.bbc.co.uk/future/article/20230525-what-will-an-el-nino-in-2023-mean-for-you)] I. Gerretsen, ‘The high cost of an El Niño in 2023’, BBC, Jun. 08, 2023. Accessed: May 14, 2026. [Online]. Available: https://www.bbc.co.uk/future/article/20230525-what-will-an-el-nino-in-2023-mean-for-you\n\n[[Gupta+25](https://doi.org/10.1038/s43247-025-02181-9)] A. K. Gupta, T. Mittal, K. E. Fauria, R. Bennartz, and J. F. Kok, ‘The January 2022 Hunga eruption cooled the southern hemisphere in 2022 and 2023’, Communications Earth & Environment, vol. 6, no. 1, p. 240, Mar. 2025, doi: 10.1038/s43247-025-02181-9.\n\n[[Hersbach+20](https://doi.org/10.1002/qj.3803)] H. Hersbach et al., ‘The ERA5 global reanalysis’, Quarterly Journal of the Royal Meteorological Society, vol. 146, no. 730, pp. 1999–2049, Jul. 2020, doi: 10.1002/qj.3803.", "text_with_prefix": "EQC Quality Assessment: \"Visualisation of temperature during extreme weather events with Climate Pulse\"\nDataset: climate-pulse [CDS]\nAspect: extreme-events_q01 | Category: Applications\nSection: Visualisation of temperature during extreme weather events with Climate Pulse > ℹ️ If you want to know more > References\n---\n[[AFP+24](https://timesofmalta.com/article/february-marks-9th-straight-month-recordsmashing-global-heat.1087990)] AFP, ‘February marks 9th straight month of record-smashing global heat’, Times of Malta, Mar. 07, 2024. Accessed: May 14, 2026. [Online]. Available: https://timesofmalta.com/article/february-marks-9th-straight-month-recordsmashing-global-heat.1087990\n\n[[Bali+24](https://www.dw.com/en/fatal-heat-wave-sweeps-greece-claims-more-lives/a-69485803)] K. Bali, ‘Heat wave in Greece claims more tourist lives’, DW, Jun. 26, 2024. Accessed: May 14, 2026. [Online]. Available: https://www.dw.com/en/fatal-heat-wave-sweeps-greece-claims-more-lives/a-69485803\n\n[[Ballester+23](https://doi.org/10.1038/s41591-023-02419-z)] J. Ballester et al., ‘Heat-related mortality in Europe during the summer of 2022’, Nature Medicine, vol. 29, no. 7, pp. 1857–1866, Jul. 2023, doi: 10.1038/s41591-023-02419-z.\n\n[[Bell+18](https://doi.org/10.1080/10962247.2017.1401017)] J. E. Bell et al., ‘Changes in extreme events and the potential impacts on human health’, Journal of the Air & Waste Management Association, vol. 68, no. 4, pp. 265–287, Apr. 2018, doi: 10.1080/10962247.2017.1401017.\n\n[[Boretti+24](https://doi.org/10.1016/j.jastp.2024.106187)] A. Boretti, ‘Reassessing the cooling that followed the 1991 volcanic eruption of Mt. Pinatubo’, Journal of Atmospheric and Solar-Terrestrial Physics, vol. 256, p. 106187, Mar. 2024, doi: 10.1016/j.jastp.2024.106187.\n\n[[Cole-Dai+10](https://doi.org/10.1002/wcc.76)] J. Cole‐Dai, ‘Volcanoes and climate’, WIREs Climate Change, vol. 1, no. 6, pp. 824–839, Nov. 2010, doi: 10.1002/wcc.76.\n\n[[Dafis+24](https://www.climameter.org/20240611-14-june-eastern-mediterranean-heatwave)] S. Dafis and D. Faranda, ‘The June 2024 Eastern Mediterranean Heatwave likely exacerbated by human-driven Climate Change and Natural Variability’, 2024, doi: 10.5281/zenodo.14103941.\n\n[[Gerretsen+23](https://www.bbc.co.uk/future/article/20230525-what-will-an-el-nino-in-2023-mean-for-you)] I. Gerretsen, ‘The high cost of an El Niño in 2023’, BBC, Jun. 08, 2023. Accessed: May 14, 2026. [Online]. Available: https://www.bbc.co.uk/future/article/20230525-what-will-an-el-nino-in-2023-mean-for-you\n\n[[Gupta+25](https://doi.org/10.1038/s43247-025-02181-9)] A. K. Gupta, T. Mittal, K. E. Fauria, R. Bennartz, and J. F. Kok, ‘The January 2022 Hunga eruption cooled the southern hemisphere in 2022 and 2023’, Communications Earth & Environment, vol. 6, no. 1, p. 240, Mar. 2025, doi: 10.1038/s43247-025-02181-9.\n\n[[Hersbach+20](https://doi.org/10.1002/qj.3803)] H. Hersbach et al., ‘The ERA5 global reanalysis’, Quarterly Journal of the Royal Meteorological Society, vol. 146, no. 730, pp. 1999–2049, Jul. 2020, doi: 10.1002/qj.3803."} {"chunk_id": "application_climate-pulse_extreme-events_q01__dd96343ac997", "report_id": "application_climate-pulse_extreme-events_q01", "dataset_id": "climate-pulse", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extreme-events_q01", "aspect_base": "extreme-events", "category": "Applications", "match_confidence": "unmatched", "section": "Visualisation of temperature during extreme weather events with Climate Pulse > ℹ️ If you want to know more > References", "title": "Visualisation of temperature during extreme weather events with Climate Pulse", "chunk_index": 17, "token_count": 1024, "text_raw": "analysis’, Quarterly Journal of the Royal Meteorological Society, vol. 146, no. 730, pp. 1999–2049, Jul. 2020, doi: 10.1002/qj.3803.\n\n[[Jenkins+23](https://doi.org/10.1038/s41558-022-01568-2)] S. Jenkins, C. Smith, M. Allen, and R. Grainger, ‘Tonga eruption increases chance of temporary surface temperature anomaly above 1.5 °C’, Nature Climate Change, vol. 13, no. 2, pp. 127–129, Feb. 2023, doi: 10.1038/s41558-022-01568-2.\n\n[[Johnson+20](https://www.dailymail.co.uk/news/article-8784073/How-La-Nina-weather-event-decade-bring-wet-summer.html)] S. Johnson, ‘La Niña forms off Australia’s coast sparking warnings of dangerous weather and cyclones - here’s how the weather pattern will affect your city’, Mail Online, Sep. 29, 2020. Accessed: Jun. 16, 2025. [Online]. Available: https://www.dailymail.co.uk/news/article-8784073/How-La-Nina-weather-event-decade-bring-wet-summer.html\n\n[[King+24](https://www.bbc.co.uk/news/articles/c98qp79gj4no)] S. King, ‘Conspiracy theories swirl about geo-engineering, but could it help save the planet?’, BBC, Jul. 21, 2024. Accessed: Jun. 16, 2025. [Online]. Available: https://www.bbc.co.uk/news/articles/c98qp79gj4no\n\n[[Kundzewicz+16](https://doi.org/10.1515/igbp-2016-0005)] Z. W. Kundzewicz, ‘Extreme Weather Events and their Consequences’, Papers on Global Change IGBP, vol. 23, no. 1, pp. 59–69, Jan. 2016, doi: 10.1515/igbp-2016-0005.\n\n[[Li+22](https://doi.org/10.1029/2021JD035546)] X. Li, Z. Hu, Y. Tseng, Y. Liu, and P. Liang, ‘A Historical Perspective of the La Niña Event in 2020/2021’, JGR Atmospheres, vol. 127, no. 7, p. e2021JD035546, Apr. 2022, doi: 10.1029/2021JD035546.\n\n[[Lhotka+22](https://doi.org/10.1029/2022EA002567)] O. Lhotka and J. Kyselý, ‘The 2021 European Heat Wave in the Context of Past Major Heat Waves’, Earth and Space Science, vol. 9, no. 11, p. e2022EA002567, Nov. 2022, doi: 10.1029/2022EA002567.\n\n[[Ma+20](https://doi.org/10.1029/2020GL087809)] F. Ma, X. Yuan, Y. Jiao, and P. Ji, ‘Unprecedented Europe Heat in June–July 2019: Risk in the Historical and Future Context’, Geophysical Research Letters, vol. 47, no. 11, p. e2020GL087809, Jun. 2020, doi: 10.1029/2020GL087809.\n\n[[McGrath+20](https://www.bbc.co.uk/news/science-environment-54725970)] M. McGrath, ‘“Moderate to strong” La Niña weather event develops in the Pacific’, Oct. 29, 2020. [Online]. Available: https://www.bbc.co.uk/news/science-environment-54725970\n\n[[McPhaden+06](https://doi.org/10.1126/science.1132588)] M. J. McPhaden, S. E. Zebiak, and M. H. Glantz, ‘ENSO as an Integrating Concept in Earth Science’, Science, vol. 314, no. 5806, pp. 1740–1745, Dec. 2006, doi: 10.1126/science.1132588.", "text_with_prefix": "EQC Quality Assessment: \"Visualisation of temperature during extreme weather events with Climate Pulse\"\nDataset: climate-pulse [CDS]\nAspect: extreme-events_q01 | Category: Applications\nSection: Visualisation of temperature during extreme weather events with Climate Pulse > ℹ️ If you want to know more > References\n---\nanalysis’, Quarterly Journal of the Royal Meteorological Society, vol. 146, no. 730, pp. 1999–2049, Jul. 2020, doi: 10.1002/qj.3803.\n\n[[Jenkins+23](https://doi.org/10.1038/s41558-022-01568-2)] S. Jenkins, C. Smith, M. Allen, and R. Grainger, ‘Tonga eruption increases chance of temporary surface temperature anomaly above 1.5 °C’, Nature Climate Change, vol. 13, no. 2, pp. 127–129, Feb. 2023, doi: 10.1038/s41558-022-01568-2.\n\n[[Johnson+20](https://www.dailymail.co.uk/news/article-8784073/How-La-Nina-weather-event-decade-bring-wet-summer.html)] S. Johnson, ‘La Niña forms off Australia’s coast sparking warnings of dangerous weather and cyclones - here’s how the weather pattern will affect your city’, Mail Online, Sep. 29, 2020. Accessed: Jun. 16, 2025. [Online]. Available: https://www.dailymail.co.uk/news/article-8784073/How-La-Nina-weather-event-decade-bring-wet-summer.html\n\n[[King+24](https://www.bbc.co.uk/news/articles/c98qp79gj4no)] S. King, ‘Conspiracy theories swirl about geo-engineering, but could it help save the planet?’, BBC, Jul. 21, 2024. Accessed: Jun. 16, 2025. [Online]. Available: https://www.bbc.co.uk/news/articles/c98qp79gj4no\n\n[[Kundzewicz+16](https://doi.org/10.1515/igbp-2016-0005)] Z. W. Kundzewicz, ‘Extreme Weather Events and their Consequences’, Papers on Global Change IGBP, vol. 23, no. 1, pp. 59–69, Jan. 2016, doi: 10.1515/igbp-2016-0005.\n\n[[Li+22](https://doi.org/10.1029/2021JD035546)] X. Li, Z. Hu, Y. Tseng, Y. Liu, and P. Liang, ‘A Historical Perspective of the La Niña Event in 2020/2021’, JGR Atmospheres, vol. 127, no. 7, p. e2021JD035546, Apr. 2022, doi: 10.1029/2021JD035546.\n\n[[Lhotka+22](https://doi.org/10.1029/2022EA002567)] O. Lhotka and J. Kyselý, ‘The 2021 European Heat Wave in the Context of Past Major Heat Waves’, Earth and Space Science, vol. 9, no. 11, p. e2022EA002567, Nov. 2022, doi: 10.1029/2022EA002567.\n\n[[Ma+20](https://doi.org/10.1029/2020GL087809)] F. Ma, X. Yuan, Y. Jiao, and P. Ji, ‘Unprecedented Europe Heat in June–July 2019: Risk in the Historical and Future Context’, Geophysical Research Letters, vol. 47, no. 11, p. e2020GL087809, Jun. 2020, doi: 10.1029/2020GL087809.\n\n[[McGrath+20](https://www.bbc.co.uk/news/science-environment-54725970)] M. McGrath, ‘“Moderate to strong” La Niña weather event develops in the Pacific’, Oct. 29, 2020. [Online]. Available: https://www.bbc.co.uk/news/science-environment-54725970\n\n[[McPhaden+06](https://doi.org/10.1126/science.1132588)] M. J. McPhaden, S. E. Zebiak, and M. H. Glantz, ‘ENSO as an Integrating Concept in Earth Science’, Science, vol. 314, no. 5806, pp. 1740–1745, Dec. 2006, doi: 10.1126/science.1132588."} {"chunk_id": "application_climate-pulse_extreme-events_q01__4904400f5d76", "report_id": "application_climate-pulse_extreme-events_q01", "dataset_id": "climate-pulse", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extreme-events_q01", "aspect_base": "extreme-events", "category": "Applications", "match_confidence": "unmatched", "section": "Visualisation of temperature during extreme weather events with Climate Pulse > ℹ️ If you want to know more > References", "title": "Visualisation of temperature during extreme weather events with Climate Pulse", "chunk_index": 18, "token_count": 1092, "text_raw": "Earth Science’, Science, vol. 314, no. 5806, pp. 1740–1745, Dec. 2006, doi: 10.1126/science.1132588.\n\n[[Mefo Newuh+24](https://www.dw.com/en/el-nino-climate-pattern-intensifies-in-east-africa/a-68932184)] M. Mefo Newuh, ‘El Nino climate pattern intensifies in East Africa’, DW, Apr. 26, 2024. Accessed: Jul. 07, 2025. [Online]. Available: https://www.dw.com/en/el-nino-climate-pattern-intensifies-in-east-africa/a-68932184\n\n[[Minnis+93](https://doi.org/10.1126/science.259.5100.1411)] P. Minnis et al., ‘Radiative Climate Forcing by the Mount Pinatubo Eruption’, Science, vol. 259, no. 5100, pp. 1411–1415, Mar. 1993, doi: 10.1126/science.259.5100.1411.\n\n[[NOAA+24](https://oceanservice.noaa.gov/facts/ninonina.html)] NOAA, ‘What are El Niño and La Niña?’, National Ocean Service. Accessed: Jun. 11, 2025. [Online]. Available: https://oceanservice.noaa.gov/facts/ninonina.html\n\n[[Philander+85](https://doi.org/10.1175/1520-0469%281985%29042%3C2652:ENALN%3E2.0.CO;2)] S. G. H. Philander, ‘El Niño and La Niña’, Journal of the Atmospheric Sciences, vol. 42, no. 23, pp. 2652–2662, Dec. 1985, doi: 10.1175/1520-0469(1985)042<2652:ENALN>2.0.​CO;2.\n\n[[Randel+24](https://doi.org/10.1029/2024GL111500)] W. J. Randel, X. Wang, J. Starr, R. R. Garcia, and D. Kinnison, ‘Long‐Term Temperature Impacts of the Hunga Volcanic Eruption in the Stratosphere and Above’, Geophysical Research Letters, vol. 51, no. 21, p. e2024GL111500, Nov. 2024, doi: 10.1029/2024GL111500.\n\n[[Robock+00](https://doi.org/10.1029/1998RG000054)] A. Robock, ‘Volcanic eruptions and climate’, Reviews of Geophysics, vol. 38, no. 2, pp. 191–219, May 2000, doi: 10.1029/1998RG000054.\n\n[[Robock+13](https://doi.org/10.1002/2013EO350001)] A. Robock, ‘The Latest on Volcanic Eruptions and Climate’, EoS Transactions, vol. 94, no. 35, pp. 305–306, Aug. 2013, doi: 10.1002/2013EO350001.\n\n[[Rowlatt+23](https://www.bbc.co.uk/news/science-environment-66143682)] J. Rowlatt, ‘Excessive heat: Why this summer has been so hot’, BBC, Jul. 13, 2023. Accessed: Jun. 16, 2025. [Online]. Available: https://www.bbc.co.uk/news/science-environment-66143682\n\n[[Shen+23](https://doi.org/10.1007/s44195-023-00051-5)] M.-H. Shen and J.-Y. Yu, ‘Changes in El Niño characteristics and air–sea feedback mechanisms under progressive global warming’, Terrestrial, Atmospheric and Oceanic Sciences, vol. 34, no. 1, p. 19, Dec. 2023, doi: 10.1007/s44195-023-00051-5.\n\n[[Shi+23](https://doi.org/10.1016/j.atmosres.2023.106937)] L. Shi, R. Ding, S. Hu, X. Li, and J. Li, ‘Extratropical impacts on the 2020–2023 Triple-Dip La Niña event’, Atmospheric Research, vol. 294, p. 106937, Oct. 2023, doi: 10.1016/j.atmosres.2023.106937.", "text_with_prefix": "EQC Quality Assessment: \"Visualisation of temperature during extreme weather events with Climate Pulse\"\nDataset: climate-pulse [CDS]\nAspect: extreme-events_q01 | Category: Applications\nSection: Visualisation of temperature during extreme weather events with Climate Pulse > ℹ️ If you want to know more > References\n---\nEarth Science’, Science, vol. 314, no. 5806, pp. 1740–1745, Dec. 2006, doi: 10.1126/science.1132588.\n\n[[Mefo Newuh+24](https://www.dw.com/en/el-nino-climate-pattern-intensifies-in-east-africa/a-68932184)] M. Mefo Newuh, ‘El Nino climate pattern intensifies in East Africa’, DW, Apr. 26, 2024. Accessed: Jul. 07, 2025. [Online]. Available: https://www.dw.com/en/el-nino-climate-pattern-intensifies-in-east-africa/a-68932184\n\n[[Minnis+93](https://doi.org/10.1126/science.259.5100.1411)] P. Minnis et al., ‘Radiative Climate Forcing by the Mount Pinatubo Eruption’, Science, vol. 259, no. 5100, pp. 1411–1415, Mar. 1993, doi: 10.1126/science.259.5100.1411.\n\n[[NOAA+24](https://oceanservice.noaa.gov/facts/ninonina.html)] NOAA, ‘What are El Niño and La Niña?’, National Ocean Service. Accessed: Jun. 11, 2025. [Online]. Available: https://oceanservice.noaa.gov/facts/ninonina.html\n\n[[Philander+85](https://doi.org/10.1175/1520-0469%281985%29042%3C2652:ENALN%3E2.0.CO;2)] S. G. H. Philander, ‘El Niño and La Niña’, Journal of the Atmospheric Sciences, vol. 42, no. 23, pp. 2652–2662, Dec. 1985, doi: 10.1175/1520-0469(1985)042<2652:ENALN>2.0.​CO;2.\n\n[[Randel+24](https://doi.org/10.1029/2024GL111500)] W. J. Randel, X. Wang, J. Starr, R. R. Garcia, and D. Kinnison, ‘Long‐Term Temperature Impacts of the Hunga Volcanic Eruption in the Stratosphere and Above’, Geophysical Research Letters, vol. 51, no. 21, p. e2024GL111500, Nov. 2024, doi: 10.1029/2024GL111500.\n\n[[Robock+00](https://doi.org/10.1029/1998RG000054)] A. Robock, ‘Volcanic eruptions and climate’, Reviews of Geophysics, vol. 38, no. 2, pp. 191–219, May 2000, doi: 10.1029/1998RG000054.\n\n[[Robock+13](https://doi.org/10.1002/2013EO350001)] A. Robock, ‘The Latest on Volcanic Eruptions and Climate’, EoS Transactions, vol. 94, no. 35, pp. 305–306, Aug. 2013, doi: 10.1002/2013EO350001.\n\n[[Rowlatt+23](https://www.bbc.co.uk/news/science-environment-66143682)] J. Rowlatt, ‘Excessive heat: Why this summer has been so hot’, BBC, Jul. 13, 2023. Accessed: Jun. 16, 2025. [Online]. Available: https://www.bbc.co.uk/news/science-environment-66143682\n\n[[Shen+23](https://doi.org/10.1007/s44195-023-00051-5)] M.-H. Shen and J.-Y. Yu, ‘Changes in El Niño characteristics and air–sea feedback mechanisms under progressive global warming’, Terrestrial, Atmospheric and Oceanic Sciences, vol. 34, no. 1, p. 19, Dec. 2023, doi: 10.1007/s44195-023-00051-5.\n\n[[Shi+23](https://doi.org/10.1016/j.atmosres.2023.106937)] L. Shi, R. Ding, S. Hu, X. Li, and J. Li, ‘Extratropical impacts on the 2020–2023 Triple-Dip La Niña event’, Atmospheric Research, vol. 294, p. 106937, Oct. 2023, doi: 10.1016/j.atmosres.2023.106937."} {"chunk_id": "application_climate-pulse_extreme-events_q01__ae8c80f11880", "report_id": "application_climate-pulse_extreme-events_q01", "dataset_id": "climate-pulse", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extreme-events_q01", "aspect_base": "extreme-events", "category": "Applications", "match_confidence": "unmatched", "section": "Visualisation of temperature during extreme weather events with Climate Pulse > ℹ️ If you want to know more > References", "title": "Visualisation of temperature during extreme weather events with Climate Pulse", "chunk_index": 19, "token_count": 831, "text_raw": "–2023 Triple-Dip La Niña event’, Atmospheric Research, vol. 294, p. 106937, Oct. 2023, doi: 10.1016/j.atmosres.2023.106937.\n\n[[Stefanon+12](https://doi.org/10.1088/1748-9326/7/1/014023)] M. Stefanon, F. D’Andrea, and P. Drobinski, ‘Heatwave classification over Europe and the Mediterranean region’, Environmental Research Letters, vol. 7, no. 1, p. 014023, Mar. 2012, doi: 10.1088/1748-9326/7/1/014023.\n\n[[Wang+16](https://doi.org/10.1007/978-94-017-7499-4_4)] C. Wang, C. Deser, J.-Y. Yu, P. DiNezio, and A. Clement, ‘El Niño and Southern Oscillation (ENSO): A Review’, in Coral Reefs of the Eastern Tropical Pacific, vol. 8, P. W. Glynn, D. P. Manzello, and I. C. Enochs, Eds., in Coral Reefs of the World, vol. 8., Dordrecht: Springer Netherlands, 2017, pp. 85–106. doi: 10.1007/978-94-017-7499-4_4.\n\n[[WMO+23](https://wmo.int/news/media-centre/world-meteorological-organization-declares-onset-of-el-nino-conditions)] World Meteorological Organization (WMO), ‘World Meteorological Organization declares onset of El Niño conditions’, World Meteorological Organization. Accessed: Jun. 16, 2025. [Online]. Available: https://wmo.int/news/media-centre/world-meteorological-organization-declares-onset-of-el-nino-conditions\n\n[[WMO+24](https://wmo.int/news/media-centre/el-nino-forecast-swing-la-nina-later-year)] World Meteorological Organization (WMO), ‘El Niño is forecast to swing to La Niña later this year’, World Meteorological Organization. Accessed: Jun. 16, 2025. [Online]. Available: https://wmo.int/news/media-centre/el-nino-forecast-swing-la-nina-later-year\n\n[[World Weather Attribution+24](https://www.worldweatherattribution.org/deadly-mediterranean-heatwave-would-not-have-occurred-without-human-induced-climate-change/)] World Weather Attribution, ‘Deadly Mediterranean heatwave would not have occurred without human induced climate change’, World Weather Attribution, Jul. 31, 2024. Accessed: Jun. 16, 2025. [Online]. Available: https://www.worldweatherattribution.org/deadly-mediterranean-heatwave-would-not-have-occurred-without-human-induced-climate-change/\n\n[[Zhao+21](https://doi.org/10.1016/S2542-5196%2821%2900081-4)] Q. Zhao et al., ‘Global, regional, and national burden of mortality associated with non-optimal ambient temperatures from 2000 to 2019: a three-stage modelling study’, The Lancet Planetary Health, vol. 5, no. 7, pp. e415–e425, Jul. 2021, doi: 10.1016/S2542-5196(21)00081-4.", "text_with_prefix": "EQC Quality Assessment: \"Visualisation of temperature during extreme weather events with Climate Pulse\"\nDataset: climate-pulse [CDS]\nAspect: extreme-events_q01 | Category: Applications\nSection: Visualisation of temperature during extreme weather events with Climate Pulse > ℹ️ If you want to know more > References\n---\n–2023 Triple-Dip La Niña event’, Atmospheric Research, vol. 294, p. 106937, Oct. 2023, doi: 10.1016/j.atmosres.2023.106937.\n\n[[Stefanon+12](https://doi.org/10.1088/1748-9326/7/1/014023)] M. Stefanon, F. D’Andrea, and P. Drobinski, ‘Heatwave classification over Europe and the Mediterranean region’, Environmental Research Letters, vol. 7, no. 1, p. 014023, Mar. 2012, doi: 10.1088/1748-9326/7/1/014023.\n\n[[Wang+16](https://doi.org/10.1007/978-94-017-7499-4_4)] C. Wang, C. Deser, J.-Y. Yu, P. DiNezio, and A. Clement, ‘El Niño and Southern Oscillation (ENSO): A Review’, in Coral Reefs of the Eastern Tropical Pacific, vol. 8, P. W. Glynn, D. P. Manzello, and I. C. Enochs, Eds., in Coral Reefs of the World, vol. 8., Dordrecht: Springer Netherlands, 2017, pp. 85–106. doi: 10.1007/978-94-017-7499-4_4.\n\n[[WMO+23](https://wmo.int/news/media-centre/world-meteorological-organization-declares-onset-of-el-nino-conditions)] World Meteorological Organization (WMO), ‘World Meteorological Organization declares onset of El Niño conditions’, World Meteorological Organization. Accessed: Jun. 16, 2025. [Online]. Available: https://wmo.int/news/media-centre/world-meteorological-organization-declares-onset-of-el-nino-conditions\n\n[[WMO+24](https://wmo.int/news/media-centre/el-nino-forecast-swing-la-nina-later-year)] World Meteorological Organization (WMO), ‘El Niño is forecast to swing to La Niña later this year’, World Meteorological Organization. Accessed: Jun. 16, 2025. [Online]. Available: https://wmo.int/news/media-centre/el-nino-forecast-swing-la-nina-later-year\n\n[[World Weather Attribution+24](https://www.worldweatherattribution.org/deadly-mediterranean-heatwave-would-not-have-occurred-without-human-induced-climate-change/)] World Weather Attribution, ‘Deadly Mediterranean heatwave would not have occurred without human induced climate change’, World Weather Attribution, Jul. 31, 2024. Accessed: Jun. 16, 2025. [Online]. Available: https://www.worldweatherattribution.org/deadly-mediterranean-heatwave-would-not-have-occurred-without-human-induced-climate-change/\n\n[[Zhao+21](https://doi.org/10.1016/S2542-5196%2821%2900081-4)] Q. Zhao et al., ‘Global, regional, and national burden of mortality associated with non-optimal ambient temperatures from 2000 to 2019: a three-stage modelling study’, The Lancet Planetary Health, vol. 5, no. 7, pp. e415–e425, Jul. 2021, doi: 10.1016/S2542-5196(21)00081-4."} {"chunk_id": "application_gttm_uncertainty_q01__b7b9efea78d2", "report_id": "application_gttm_uncertainty_q01", "dataset_id": "gttm", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty_q01", "aspect_base": "uncertainty", "category": "Applications", "match_confidence": "unmatched", "section": "Assessing global warming with the C3S Global Temperature Trend Monitor > Quality assessment question", "title": "Assessing global warming with the C3S Global Temperature Trend Monitor", "chunk_index": 0, "token_count": 713, "text_raw": "* **Is the Global Temperature Trend Monitor application a suitable tool for calculating and visualising temperature trends?**\n\nGlobal warming, the steady increase in global average temperature over the past few decades [[Shen+22](https://doi.org/10.1007/s11442-022-1937-1), [Mudelsee+19](https://doi.org/10.1016/j.earscirev.2018.12.005)], is one of the most prominent aspects of climate change.\nMore hot, dry days lead to a stronger personal perception of climate change among the general public [[Marlon+21](https://doi.org/10.1016/j.gloenvcha.2021.102247)].\nAccordingly, changes in global and regional temperature are widely reported on in both scientific [[Copernicus Climate Change Service+24](https://doi.org/10.24381/bs9v-8c66), [IPCC+23](https://doi.org/10.59327/IPCC/AR6-9789291691647)] and popular media [[Frijters+25](https://www.volkskrant.nl/wetenschap/~b301f81f2/), [Poynting+25](https://www.bbc.com/news/articles/cd7575x8yq5o), [Horton+24](https://www.theguardian.com/environment/2024/nov/20/the-climate-crisis-in-charts-how-2024-has-set-unwanted-new-records)].\nThe Paris Agreement [[United Nations+15](https://treaties.un.org/doc/Treaties/2016/02/20160215%2006-03%20PM/Ch_XXVII-7-d.pdf)] set a goal of limiting average global warming to well below 2 °C, ideally 1.5 °C.\nThe actual rate of warming compared to this threshold is a crucial indicator of the progress of efforts to combat climate change.\n\nIn this assessment, the [*C3S Global Temperature Trend Monitor*](https://apps.climate.copernicus.eu/global-temperature-trend-monitor/) web application (Figure {numref}`{number} `) is examined for its ability to calculate and display trends in global temperature.\nThe uncertainty in the data and methods, and the ability of the application to display its uncertainty, are assessed.\n\nattachment:application_gttm_may2025.png\n---\nheight: 600px\nname: application_gttm_may2025\n---\nThe C3S Global Temperature Trend Monitor (GTTM) in May 2025. Orange dots show monthly global average temperature anomalies within (bright) and outside (faded) a user-specified 30-year window. A linear trend (red solid line) is fitted through the data and indicates the current average anomaly (orange dotted line; top text) and predicted time to reach 1.5 °C (red vertical line; top text), in this case April 2029. The orange envelope shows the IPCC climate projection uncertainty estimate.\n```", "text_with_prefix": "EQC Quality Assessment: \"Assessing global warming with the C3S Global Temperature Trend Monitor\"\nDataset: gttm [CDS]\nAspect: uncertainty_q01 | Category: Applications\nSection: Assessing global warming with the C3S Global Temperature Trend Monitor > Quality assessment question\n---\n* **Is the Global Temperature Trend Monitor application a suitable tool for calculating and visualising temperature trends?**\n\nGlobal warming, the steady increase in global average temperature over the past few decades [[Shen+22](https://doi.org/10.1007/s11442-022-1937-1), [Mudelsee+19](https://doi.org/10.1016/j.earscirev.2018.12.005)], is one of the most prominent aspects of climate change.\nMore hot, dry days lead to a stronger personal perception of climate change among the general public [[Marlon+21](https://doi.org/10.1016/j.gloenvcha.2021.102247)].\nAccordingly, changes in global and regional temperature are widely reported on in both scientific [[Copernicus Climate Change Service+24](https://doi.org/10.24381/bs9v-8c66), [IPCC+23](https://doi.org/10.59327/IPCC/AR6-9789291691647)] and popular media [[Frijters+25](https://www.volkskrant.nl/wetenschap/~b301f81f2/), [Poynting+25](https://www.bbc.com/news/articles/cd7575x8yq5o), [Horton+24](https://www.theguardian.com/environment/2024/nov/20/the-climate-crisis-in-charts-how-2024-has-set-unwanted-new-records)].\nThe Paris Agreement [[United Nations+15](https://treaties.un.org/doc/Treaties/2016/02/20160215%2006-03%20PM/Ch_XXVII-7-d.pdf)] set a goal of limiting average global warming to well below 2 °C, ideally 1.5 °C.\nThe actual rate of warming compared to this threshold is a crucial indicator of the progress of efforts to combat climate change.\n\nIn this assessment, the [*C3S Global Temperature Trend Monitor*](https://apps.climate.copernicus.eu/global-temperature-trend-monitor/) web application (Figure {numref}`{number} `) is examined for its ability to calculate and display trends in global temperature.\nThe uncertainty in the data and methods, and the ability of the application to display its uncertainty, are assessed.\n\nattachment:application_gttm_may2025.png\n---\nheight: 600px\nname: application_gttm_may2025\n---\nThe C3S Global Temperature Trend Monitor (GTTM) in May 2025. Orange dots show monthly global average temperature anomalies within (bright) and outside (faded) a user-specified 30-year window. A linear trend (red solid line) is fitted through the data and indicates the current average anomaly (orange dotted line; top text) and predicted time to reach 1.5 °C (red vertical line; top text), in this case April 2029. The orange envelope shows the IPCC climate projection uncertainty estimate.\n```"} {"chunk_id": "application_gttm_uncertainty_q01__42f21508940a", "report_id": "application_gttm_uncertainty_q01", "dataset_id": "gttm", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty_q01", "aspect_base": "uncertainty", "category": "Applications", "match_confidence": "unmatched", "section": "Assessing global warming with the C3S Global Temperature Trend Monitor > Quality assessment statement", "title": "Assessing global warming with the C3S Global Temperature Trend Monitor", "chunk_index": 1, "token_count": 203, "text_raw": "These are the key outcomes of this assessment\n\n* The Global Temperature Trend Monitor (GTTM) is a scientifically sound tool for monitoring global temperature trends over decade-long time scales, and yearly changes therein.\n* The uncertainty associated with the methods used in GTTM (anomaly baselines, least squares fitting) is sufficiently small compared to the uncertainty in the underpinning dataset and inherent to the climate, and the application provides suitable methods for visualising the expected level of uncertainty.\n* The provided source code makes it possible to adapt GTTM to different user needs, e.g. regional rather than global trends, albeit with some additional considerations such as the calculation of anomalies relative to a local baseline.\n```", "text_with_prefix": "EQC Quality Assessment: \"Assessing global warming with the C3S Global Temperature Trend Monitor\"\nDataset: gttm [CDS]\nAspect: uncertainty_q01 | Category: Applications\nSection: Assessing global warming with the C3S Global Temperature Trend Monitor > Quality assessment statement\n---\nThese are the key outcomes of this assessment\n\n* The Global Temperature Trend Monitor (GTTM) is a scientifically sound tool for monitoring global temperature trends over decade-long time scales, and yearly changes therein.\n* The uncertainty associated with the methods used in GTTM (anomaly baselines, least squares fitting) is sufficiently small compared to the uncertainty in the underpinning dataset and inherent to the climate, and the application provides suitable methods for visualising the expected level of uncertainty.\n* The provided source code makes it possible to adapt GTTM to different user needs, e.g. regional rather than global trends, albeit with some additional considerations such as the calculation of anomalies relative to a local baseline.\n```"} {"chunk_id": "application_gttm_uncertainty_q01__acf3facebbd4", "report_id": "application_gttm_uncertainty_q01", "dataset_id": "gttm", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty_q01", "aspect_base": "uncertainty", "category": "Applications", "match_confidence": "unmatched", "section": "Assessing global warming with the C3S Global Temperature Trend Monitor > Methodology", "title": "Assessing global warming with the C3S Global Temperature Trend Monitor", "chunk_index": 2, "token_count": 452, "text_raw": "The [*C3S Global Temperature Trend Monitor*](https://apps.climate.copernicus.eu/global-temperature-trend-monitor/) (*GTTM*) is a C3S web application displaying monthly global average temperature anomalies.\nTemperatures from the ERA5 reanalysis [[Hersbach+20](https://doi.org/10.1002/qj.3803)] are compared to the pre-industrial (1850–1900) average to calculate temperature anomalies.\nA 30-year linear trend is fitted to these anomalies and extrapolated to estimate when global warming will reach 1.5 °C.\nThe start and end of the 30-year range can be modified by the user to explore how the predicted date of reaching 1.5 °C has changed over time.\n\nGTTM is based on two datasets:\n* [_ERA5 monthly averaged data on single levels from 1940 to present_](https://doi.org/10.24381/cds.f17050d7) [[ERA5 monthly dataset](https://doi.org/10.24381/cds.f17050d7)].\n* [_Essential climate variables for assessment of climate variability from 1979 to present_](https://doi.org/10.24381/7470b643) [[ECVs for climate variability dataset](https://doi.org/10.24381/7470b643)].\n\nApplications are quality assessed separately from their underpinning datasets. For information on quality attributes of the underlying datasets, the reader is referred to the quality assessment of the datasets, which can be found in their respective CDS catalogue entries or [on the EQC hub](../intro).\n\nThe analysis and results are organised in the following steps, which are detailed in the sections below:\n\n**[](section-data)**\n\n**[](section-anomalies)**\n\n**[](section-trends)**\n\n**[](section-visualisation)**", "text_with_prefix": "EQC Quality Assessment: \"Assessing global warming with the C3S Global Temperature Trend Monitor\"\nDataset: gttm [CDS]\nAspect: uncertainty_q01 | Category: Applications\nSection: Assessing global warming with the C3S Global Temperature Trend Monitor > Methodology\n---\nThe [*C3S Global Temperature Trend Monitor*](https://apps.climate.copernicus.eu/global-temperature-trend-monitor/) (*GTTM*) is a C3S web application displaying monthly global average temperature anomalies.\nTemperatures from the ERA5 reanalysis [[Hersbach+20](https://doi.org/10.1002/qj.3803)] are compared to the pre-industrial (1850–1900) average to calculate temperature anomalies.\nA 30-year linear trend is fitted to these anomalies and extrapolated to estimate when global warming will reach 1.5 °C.\nThe start and end of the 30-year range can be modified by the user to explore how the predicted date of reaching 1.5 °C has changed over time.\n\nGTTM is based on two datasets:\n* [_ERA5 monthly averaged data on single levels from 1940 to present_](https://doi.org/10.24381/cds.f17050d7) [[ERA5 monthly dataset](https://doi.org/10.24381/cds.f17050d7)].\n* [_Essential climate variables for assessment of climate variability from 1979 to present_](https://doi.org/10.24381/7470b643) [[ECVs for climate variability dataset](https://doi.org/10.24381/7470b643)].\n\nApplications are quality assessed separately from their underpinning datasets. For information on quality attributes of the underlying datasets, the reader is referred to the quality assessment of the datasets, which can be found in their respective CDS catalogue entries or [on the EQC hub](../intro).\n\nThe analysis and results are organised in the following steps, which are detailed in the sections below:\n\n**[](section-data)**\n\n**[](section-anomalies)**\n\n**[](section-trends)**\n\n**[](section-visualisation)**"} {"chunk_id": "application_gttm_uncertainty_q01__53e0260257a3", "report_id": "application_gttm_uncertainty_q01", "dataset_id": "gttm", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty_q01", "aspect_base": "uncertainty", "category": "Applications", "match_confidence": "unmatched", "section": "Assessing global warming with the C3S Global Temperature Trend Monitor > Analysis and results > 1. Underpinning dataset", "title": "Assessing global warming with the C3S Global Temperature Trend Monitor", "chunk_index": 3, "token_count": 390, "text_raw": "GTTM is based on ERA5 data, both directly through its monthly averaged data [[ERA5 monthly dataset](https://doi.org/10.24381/cds.f17050d7)] and indirectly through a dataset of essential climate variables (ECVs) derived from ERA5, ERA5-Land, and ERA-Interim [[ECVs for climate variability dataset](https://doi.org/10.24381/7470b643)].\nNumerous validation studies have shown that ERA5 provides accurate and reliable estimates of surface air temperature, with strong agreement against in-situ measurements and other datasets, albeit with regional variations [[Cavalleri+24](https://doi.org/10.1002/joc.8475), [Liu+24](https://doi.org/10.1016/j.ecolind.2024.112481), [Yilmaz+23](https://doi.org/10.1016/j.scitotenv.2022.159182)].\n\nUncertainties on individual ERA5 temperature data are typically 0.1–0.5 °C [[Copernicus Climate Change Service+25a](https://confluence.ecmwf.int/display/CKB/ERA5%3A+uncertainty+estimation)].\nSome level of spatial and temporal covariance is to be expected, both physically and due to the methods of data production.\nWhile an exact value is difficult to provide, it can be fairly assumed that the uncertainty in global averages is smaller than that in individual data.\n\n(section-anomalies)=", "text_with_prefix": "EQC Quality Assessment: \"Assessing global warming with the C3S Global Temperature Trend Monitor\"\nDataset: gttm [CDS]\nAspect: uncertainty_q01 | Category: Applications\nSection: Assessing global warming with the C3S Global Temperature Trend Monitor > Analysis and results > 1. Underpinning dataset\n---\nGTTM is based on ERA5 data, both directly through its monthly averaged data [[ERA5 monthly dataset](https://doi.org/10.24381/cds.f17050d7)] and indirectly through a dataset of essential climate variables (ECVs) derived from ERA5, ERA5-Land, and ERA-Interim [[ECVs for climate variability dataset](https://doi.org/10.24381/7470b643)].\nNumerous validation studies have shown that ERA5 provides accurate and reliable estimates of surface air temperature, with strong agreement against in-situ measurements and other datasets, albeit with regional variations [[Cavalleri+24](https://doi.org/10.1002/joc.8475), [Liu+24](https://doi.org/10.1016/j.ecolind.2024.112481), [Yilmaz+23](https://doi.org/10.1016/j.scitotenv.2022.159182)].\n\nUncertainties on individual ERA5 temperature data are typically 0.1–0.5 °C [[Copernicus Climate Change Service+25a](https://confluence.ecmwf.int/display/CKB/ERA5%3A+uncertainty+estimation)].\nSome level of spatial and temporal covariance is to be expected, both physically and due to the methods of data production.\nWhile an exact value is difficult to provide, it can be fairly assumed that the uncertainty in global averages is smaller than that in individual data.\n\n(section-anomalies)="} {"chunk_id": "application_gttm_uncertainty_q01__f33d4274e686", "report_id": "application_gttm_uncertainty_q01", "dataset_id": "gttm", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty_q01", "aspect_base": "uncertainty", "category": "Applications", "match_confidence": "unmatched", "section": "Assessing global warming with the C3S Global Temperature Trend Monitor > Analysis and results > 2. Temperature anomalies", "title": "Assessing global warming with the C3S Global Temperature Trend Monitor", "chunk_index": 4, "token_count": 979, "text_raw": "GTTM calculates monthly temperature anomalies by subtracting the 1991–2020 climatology (ECV dataset) from ERA5 monthly temperatures.\nIn general, ERA5 temperature anomalies agree well with those from other sources [[Hersbach+20](https://doi.org/10.1002/qj.3803)].\n\nThe 1991–2020 baseline anomalies are converted to anomalies against the 1850–1900 pre-industrial average using a set of monthly offsets based on three datasets (Berkeley Earth, HadCRUT5, NOAAGlobalTemp; see table below) spanning 1850–2020.\nThese monthly offsets were calculated as global averages: averaged first by grid square, then by hemisphere, then overall.\nThis method is also used in the C3S Climate Bulletin and is explained in detail in [[Copernicus Climate Change Service+25b](https://climate.copernicus.eu/climate-bulletin-about-data-and-analysis)].\n\n| Month | Berkeley Earth | HadCRUT5 | NOAAGlobalTemp | Average | Applied Offset |\n|:-----------|---------------:|---------:|---------------:|--------:|---------------:|\n|Jan | 0.99 | 0.97 | 0.83 | 0.93 | 0.96 |\n|Feb | 1.03 | 1.00 | 0.84 | 0.96 | 0.96 |\n|Mar | 1.04 | 1.03 | 0.87 | 0.98 | 0.95 |\n|Apr | 0.98 | 0.97 | 0.84 | 0.93 | 0.91 |\n|May | 0.90 | 0.88 | 0.76 | 0.85 | 0.87 |\n|Jun | 0.86 | 0.83 | 0.76 | 0.82 | 0.83 |\n|Jul | 0.81 | 0.77 | 0.71 | 0.77 | 0.80 |\n|Aug | 0.88 | 0.80 | 0.73 | 0.80 | 0.80 |\n|Sep | 0.93 | 0.80 | 0.76 | 0.83 | 0.81 |\n|Oct | 0.92 | 0.87 | 0.81 | 0.87 | 0.85 |\n|Nov | 0.95 | 0.97 | 0.82 | 0.92 | 0.89 |\n|Dec | 0.97 | 0.93 | 0.80 | 0.90 | 0.93 |\n| | | | | | |\n|Annual Mean | 0.94 | 0.90 | 0.80 | 0.88 | 0.88 |\n\nThe applied offset differs from the three source datasets by –0.13 to +0.12 °C or –14% to +15% (Figure {numref}`{number} `).\nThis uncertainty in the baseline correction propagates into the anomalies used in GTTM on two scales:\n1. Yearly: Since the same offset is applied to every year, there is an uncertainty in the overall offset of the data but not in year-on-year differences. This uncertainty affects the offset of the fitted line but not its slope.\n2. Monthly: Seasonal differences in the offsets and in the uncertainties therein cause an uncertainty in the month-on-month differences within each year. Over multiple years, this averages out and should not significantly affect the fitted line.\n\nattachment:application_gttm_offsets.png\n---\nheight: 387px\nname: application_gttm_offsets\n---\nOffsets applied to adjust temperature anomalies from a 1991–2020 baseline to an 1850–1900 (pre-industrial) baseline.\n```", "text_with_prefix": "EQC Quality Assessment: \"Assessing global warming with the C3S Global Temperature Trend Monitor\"\nDataset: gttm [CDS]\nAspect: uncertainty_q01 | Category: Applications\nSection: Assessing global warming with the C3S Global Temperature Trend Monitor > Analysis and results > 2. Temperature anomalies\n---\nGTTM calculates monthly temperature anomalies by subtracting the 1991–2020 climatology (ECV dataset) from ERA5 monthly temperatures.\nIn general, ERA5 temperature anomalies agree well with those from other sources [[Hersbach+20](https://doi.org/10.1002/qj.3803)].\n\nThe 1991–2020 baseline anomalies are converted to anomalies against the 1850–1900 pre-industrial average using a set of monthly offsets based on three datasets (Berkeley Earth, HadCRUT5, NOAAGlobalTemp; see table below) spanning 1850–2020.\nThese monthly offsets were calculated as global averages: averaged first by grid square, then by hemisphere, then overall.\nThis method is also used in the C3S Climate Bulletin and is explained in detail in [[Copernicus Climate Change Service+25b](https://climate.copernicus.eu/climate-bulletin-about-data-and-analysis)].\n\n| Month | Berkeley Earth | HadCRUT5 | NOAAGlobalTemp | Average | Applied Offset |\n|:-----------|---------------:|---------:|---------------:|--------:|---------------:|\n|Jan | 0.99 | 0.97 | 0.83 | 0.93 | 0.96 |\n|Feb | 1.03 | 1.00 | 0.84 | 0.96 | 0.96 |\n|Mar | 1.04 | 1.03 | 0.87 | 0.98 | 0.95 |\n|Apr | 0.98 | 0.97 | 0.84 | 0.93 | 0.91 |\n|May | 0.90 | 0.88 | 0.76 | 0.85 | 0.87 |\n|Jun | 0.86 | 0.83 | 0.76 | 0.82 | 0.83 |\n|Jul | 0.81 | 0.77 | 0.71 | 0.77 | 0.80 |\n|Aug | 0.88 | 0.80 | 0.73 | 0.80 | 0.80 |\n|Sep | 0.93 | 0.80 | 0.76 | 0.83 | 0.81 |\n|Oct | 0.92 | 0.87 | 0.81 | 0.87 | 0.85 |\n|Nov | 0.95 | 0.97 | 0.82 | 0.92 | 0.89 |\n|Dec | 0.97 | 0.93 | 0.80 | 0.90 | 0.93 |\n| | | | | | |\n|Annual Mean | 0.94 | 0.90 | 0.80 | 0.88 | 0.88 |\n\nThe applied offset differs from the three source datasets by –0.13 to +0.12 °C or –14% to +15% (Figure {numref}`{number} `).\nThis uncertainty in the baseline correction propagates into the anomalies used in GTTM on two scales:\n1. Yearly: Since the same offset is applied to every year, there is an uncertainty in the overall offset of the data but not in year-on-year differences. This uncertainty affects the offset of the fitted line but not its slope.\n2. Monthly: Seasonal differences in the offsets and in the uncertainties therein cause an uncertainty in the month-on-month differences within each year. Over multiple years, this averages out and should not significantly affect the fitted line.\n\nattachment:application_gttm_offsets.png\n---\nheight: 387px\nname: application_gttm_offsets\n---\nOffsets applied to adjust temperature anomalies from a 1991–2020 baseline to an 1850–1900 (pre-industrial) baseline.\n```"} {"chunk_id": "application_gttm_uncertainty_q01__42521d5efb72", "report_id": "application_gttm_uncertainty_q01", "dataset_id": "gttm", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty_q01", "aspect_base": "uncertainty", "category": "Applications", "match_confidence": "unmatched", "section": "Assessing global warming with the C3S Global Temperature Trend Monitor > Analysis and results > 2. Temperature anomalies", "title": "Assessing global warming with the C3S Global Temperature Trend Monitor", "chunk_index": 5, "token_count": 386, "text_raw": ": 387px\nname: application_gttm_offsets\n---\nOffsets applied to adjust temperature anomalies from a 1991–2020 baseline to an 1850–1900 (pre-industrial) baseline.\n```\n\nThe uncertainty on the global average temperature anomaly offset resulting from this variability can be roughly estimated as the standard deviation of the three annual means, 0.06 °C.\nThis propagates into an uncertainty of 2–4 years on the estimated date of reaching 1.5 °C, depending on the slope and date range.\nSince the slope of the fitted line is not affected, there is no additional uncertainty on comparisons between 30-year fitting ranges.\nFor early starting dates, an uncertainty of 2–4 years is small compared to the year-on-year variations and to the total duration.\nIt is significant for recent years – for example, in January 2025, the estimated time left until 1.5 °C was <5 years.\n\nUsers intending to modify the application for a specific region should be aware that the global average offset might not be directly applicable, especially in regions with particularly strong (e.g. the Arctic [[Rantanen+22](https://doi.org/10.1038/s43247-022-00498-3)]) or weak (e.g. the North Atlantic [[Sgubin+17](https://doi.org/10.1038/ncomms14375)]) warming.\nIn these cases, it can be worthwhile to derive region-specific offsets from the source datasets.\n\n(section-trends)=", "text_with_prefix": "EQC Quality Assessment: \"Assessing global warming with the C3S Global Temperature Trend Monitor\"\nDataset: gttm [CDS]\nAspect: uncertainty_q01 | Category: Applications\nSection: Assessing global warming with the C3S Global Temperature Trend Monitor > Analysis and results > 2. Temperature anomalies\n---\n: 387px\nname: application_gttm_offsets\n---\nOffsets applied to adjust temperature anomalies from a 1991–2020 baseline to an 1850–1900 (pre-industrial) baseline.\n```\n\nThe uncertainty on the global average temperature anomaly offset resulting from this variability can be roughly estimated as the standard deviation of the three annual means, 0.06 °C.\nThis propagates into an uncertainty of 2–4 years on the estimated date of reaching 1.5 °C, depending on the slope and date range.\nSince the slope of the fitted line is not affected, there is no additional uncertainty on comparisons between 30-year fitting ranges.\nFor early starting dates, an uncertainty of 2–4 years is small compared to the year-on-year variations and to the total duration.\nIt is significant for recent years – for example, in January 2025, the estimated time left until 1.5 °C was <5 years.\n\nUsers intending to modify the application for a specific region should be aware that the global average offset might not be directly applicable, especially in regions with particularly strong (e.g. the Arctic [[Rantanen+22](https://doi.org/10.1038/s43247-022-00498-3)]) or weak (e.g. the North Atlantic [[Sgubin+17](https://doi.org/10.1038/ncomms14375)]) warming.\nIn these cases, it can be worthwhile to derive region-specific offsets from the source datasets.\n\n(section-trends)="} {"chunk_id": "application_gttm_uncertainty_q01__812fd6e600aa", "report_id": "application_gttm_uncertainty_q01", "dataset_id": "gttm", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty_q01", "aspect_base": "uncertainty", "category": "Applications", "match_confidence": "unmatched", "section": "Assessing global warming with the C3S Global Temperature Trend Monitor > Analysis and results > 3. Trend estimation", "title": "Assessing global warming with the C3S Global Temperature Trend Monitor", "chunk_index": 6, "token_count": 891, "text_raw": "GTTM uses a least squares linear fit (implemented through [`numpy.polyfit`](https://numpy.org/doc/stable/reference/generated/numpy.polyfit.html)) to estimate and extrapolate a 30-year trend.\nWhile past and projected temperature curves are typically more complicated than a straight line [[O’Neill+16](https://doi.org/10.5194/gmd-9-3461-2016)], a simple linear fit can adequately describe trends since the 1970s [[Mudelsee+19](https://doi.org/10.1016/j.earscirev.2018.12.005)], especially for illustrative rather than predictive purposes.\nThe 30-year range corresponds to the standard length of time used by the World Meteorological Organization (WMO) for climatology [[WMO](https://community.wmo.int/en/activity-areas/climate-services/climate-products-and-initiatives/wmo-climatological-normals)].\n\nAs an example, for the period of January 1995 – January 2025, the best-fitting parameters are a slope $m$ = 0.00214 °C/month (0.0257 °C/year) and an offset $c$ = 0.604 °C, meaning the 1.5 °C threshold would be reached $(1.5\\,\\text{°C} \\, – \\, c)/m$ = 418.6 months from January 1995, or November–December 2029. The GTTM interface rounds this result down, in this example to November 2029.\n\nThe `numpy.polyfit` routine can provide an uncertainty estimate in its outputs, representing the fit uncertainty due to the spread in the data but not the uncertainty in the data themselves. The corresponding estimated uncertainties are $\\sigma_m$ = 8.2 × 10–5 °C/month (3.8%) and $\\sigma_c$ = 0.017 °C (2.8%), with a correlation coefficient of $r_{m,c}$ = –0.87 (i.e. strongly negatively correlated). The corresponding uncertainty $\\sigma_t$ in the month of reaching 1.5 °C $t$ can be estimated using a Taylor expansion, the correlated equivalent of the sum-of-squares method:\n\n$$ \\sigma_t^2 = \\mathbf{J} \\mathbf{\\Sigma} \\mathbf{J}^\\top $$\n\nWhere $\\mathbf{\\Sigma}$ is the covariance matrix and $\\mathbf{J}$ is the Jacobian matrix $\\mathbf{J} = \\left[\\begin{matrix}\\frac{\\partial t}{\\partial m}&\\frac{\\partial t}{\\partial c}\\\\\\end{matrix}\\right]$.\n\nThe resulting uncertainty is 10 months, or February 2029 – October 2030 in this example.\n\nThe choice of a least squares fit for the regression line, rather than an alternative, does not significantly affect the result.\nFor example, a Theil–Sen (median-of-slopes) estimator, implemented using [scikit-learn](https://scikit-learn.org/stable/modules/generated/sklearn.linear_model.TheilSenRegressor.html), provides an estimated slope $m$ = 0.00208 °C/month and offset $c$ = 0.594 °C for the same data.\nThe equivalent date of reaching 1.5 °C would be April 2031.\nThe scikit-learn method does not provide an uncertainty estimate on these parameters, and a manual calculation is outside the scope of this assessment.\nHowever, comparing the difference between the two linear estimators (16 months) to the uncertainty in one (10 months), the unknown uncertainty in the other, the uncertainty resulting from the anomaly calculation (see above), as well as the uncertainty in ERA5 data in general, it can be concluded that the difference is not significant and that the least squares fit is suitable.\n\n(section-visualisation)=", "text_with_prefix": "EQC Quality Assessment: \"Assessing global warming with the C3S Global Temperature Trend Monitor\"\nDataset: gttm [CDS]\nAspect: uncertainty_q01 | Category: Applications\nSection: Assessing global warming with the C3S Global Temperature Trend Monitor > Analysis and results > 3. Trend estimation\n---\nGTTM uses a least squares linear fit (implemented through [`numpy.polyfit`](https://numpy.org/doc/stable/reference/generated/numpy.polyfit.html)) to estimate and extrapolate a 30-year trend.\nWhile past and projected temperature curves are typically more complicated than a straight line [[O’Neill+16](https://doi.org/10.5194/gmd-9-3461-2016)], a simple linear fit can adequately describe trends since the 1970s [[Mudelsee+19](https://doi.org/10.1016/j.earscirev.2018.12.005)], especially for illustrative rather than predictive purposes.\nThe 30-year range corresponds to the standard length of time used by the World Meteorological Organization (WMO) for climatology [[WMO](https://community.wmo.int/en/activity-areas/climate-services/climate-products-and-initiatives/wmo-climatological-normals)].\n\nAs an example, for the period of January 1995 – January 2025, the best-fitting parameters are a slope $m$ = 0.00214 °C/month (0.0257 °C/year) and an offset $c$ = 0.604 °C, meaning the 1.5 °C threshold would be reached $(1.5\\,\\text{°C} \\, – \\, c)/m$ = 418.6 months from January 1995, or November–December 2029. The GTTM interface rounds this result down, in this example to November 2029.\n\nThe `numpy.polyfit` routine can provide an uncertainty estimate in its outputs, representing the fit uncertainty due to the spread in the data but not the uncertainty in the data themselves. The corresponding estimated uncertainties are $\\sigma_m$ = 8.2 × 10–5 °C/month (3.8%) and $\\sigma_c$ = 0.017 °C (2.8%), with a correlation coefficient of $r_{m,c}$ = –0.87 (i.e. strongly negatively correlated). The corresponding uncertainty $\\sigma_t$ in the month of reaching 1.5 °C $t$ can be estimated using a Taylor expansion, the correlated equivalent of the sum-of-squares method:\n\n$$ \\sigma_t^2 = \\mathbf{J} \\mathbf{\\Sigma} \\mathbf{J}^\\top $$\n\nWhere $\\mathbf{\\Sigma}$ is the covariance matrix and $\\mathbf{J}$ is the Jacobian matrix $\\mathbf{J} = \\left[\\begin{matrix}\\frac{\\partial t}{\\partial m}&\\frac{\\partial t}{\\partial c}\\\\\\end{matrix}\\right]$.\n\nThe resulting uncertainty is 10 months, or February 2029 – October 2030 in this example.\n\nThe choice of a least squares fit for the regression line, rather than an alternative, does not significantly affect the result.\nFor example, a Theil–Sen (median-of-slopes) estimator, implemented using [scikit-learn](https://scikit-learn.org/stable/modules/generated/sklearn.linear_model.TheilSenRegressor.html), provides an estimated slope $m$ = 0.00208 °C/month and offset $c$ = 0.594 °C for the same data.\nThe equivalent date of reaching 1.5 °C would be April 2031.\nThe scikit-learn method does not provide an uncertainty estimate on these parameters, and a manual calculation is outside the scope of this assessment.\nHowever, comparing the difference between the two linear estimators (16 months) to the uncertainty in one (10 months), the unknown uncertainty in the other, the uncertainty resulting from the anomaly calculation (see above), as well as the uncertainty in ERA5 data in general, it can be concluded that the difference is not significant and that the least squares fit is suitable.\n\n(section-visualisation)="} {"chunk_id": "application_gttm_uncertainty_q01__fb43997cc6b2", "report_id": "application_gttm_uncertainty_q01", "dataset_id": "gttm", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty_q01", "aspect_base": "uncertainty", "category": "Applications", "match_confidence": "unmatched", "section": "Assessing global warming with the C3S Global Temperature Trend Monitor > Analysis and results > 4. Visualisation of uncertainty", "title": "Assessing global warming with the C3S Global Temperature Trend Monitor", "chunk_index": 7, "token_count": 261, "text_raw": "While GTTM does not explicitly provide uncertainty estimates on the underpinning data, its fitted trend line, or the estimated date of reaching 1.5 °C, it does provide users with some tools to intuit these uncertainties.\nFirst, sliding between different 30-year ranges and seeing the change in estimated date indicates the level of uncertainty due to natural variability (e.g. particularly hot or cold months) and actual changes in the rate of global warming.\nSecond, there is an overlay showing the typical uncertainty in IPCC projections of climate projections.\n\nThe documentation extensively explains that the purpose of GTTM is to *illustrate* climate change to a broad audience, not to *predict* it.\nComparing the uncertainty in different aspects of the application (anomaly calculation, trend estimation) to the uncertainty in the underpinning dataset and the physical variability in global temperature (natural and man-made), we can conclude that GTTM is fit for this purpose.", "text_with_prefix": "EQC Quality Assessment: \"Assessing global warming with the C3S Global Temperature Trend Monitor\"\nDataset: gttm [CDS]\nAspect: uncertainty_q01 | Category: Applications\nSection: Assessing global warming with the C3S Global Temperature Trend Monitor > Analysis and results > 4. Visualisation of uncertainty\n---\nWhile GTTM does not explicitly provide uncertainty estimates on the underpinning data, its fitted trend line, or the estimated date of reaching 1.5 °C, it does provide users with some tools to intuit these uncertainties.\nFirst, sliding between different 30-year ranges and seeing the change in estimated date indicates the level of uncertainty due to natural variability (e.g. particularly hot or cold months) and actual changes in the rate of global warming.\nSecond, there is an overlay showing the typical uncertainty in IPCC projections of climate projections.\n\nThe documentation extensively explains that the purpose of GTTM is to *illustrate* climate change to a broad audience, not to *predict* it.\nComparing the uncertainty in different aspects of the application (anomaly calculation, trend estimation) to the uncertainty in the underpinning dataset and the physical variability in global temperature (natural and man-made), we can conclude that GTTM is fit for this purpose."} {"chunk_id": "application_gttm_uncertainty_q01__14f9d3a79e25", "report_id": "application_gttm_uncertainty_q01", "dataset_id": "gttm", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty_q01", "aspect_base": "uncertainty", "category": "Applications", "match_confidence": "unmatched", "section": "Assessing global warming with the C3S Global Temperature Trend Monitor > ℹ️ If you want to know more > Key resources", "title": "Assessing global warming with the C3S Global Temperature Trend Monitor", "chunk_index": 8, "token_count": 428, "text_raw": "More about the ERA5 reanalysis:\n* [The ERA5 global reanalysis](https://doi.org/10.1002/qj.3803)\n* [ERA5: uncertainty estimation](https://confluence.ecmwf.int/display/CKB/ERA5%3A+uncertainty+estimation)\n\nMore about temperature anomalies:\n* [Climate Bulletin – About the data and analysis](https://climate.copernicus.eu/climate-bulletin-about-data-and-analysis)\n\nMore about climate trend estimation:\n* [Trend analysis of climate time series: A review of methods](https://doi.org/10.1016/j.earscirev.2018.12.005)\n* [Estimating trends and the current climate mean in a changing climate](https://doi.org/10.1016/j.cliser.2023.100428)\n\nMore about the 1.5 °C and 2 °C targets:\n* [Paris Agreement](https://treaties.un.org/doc/Treaties/2016/02/20160215%2006-03%20PM/Ch_XXVII-7-d.pdf)\n* [A history of the 1.5 °C target](https://doi.org/10.1002/wcc.824)\n* [Realizing the impacts of a 1.5 °C warmer world](https://doi.org/10.1038/nclimate3055)\n\nCode libraries used:\n* [NumPy polyfit](https://numpy.org/doc/stable/reference/generated/numpy.polyfit.html)\n* [scikit-learn Theil–Sen estimator](https://scikit-learn.org/stable/modules/generated/sklearn.linear_model.TheilSenRegressor.html)", "text_with_prefix": "EQC Quality Assessment: \"Assessing global warming with the C3S Global Temperature Trend Monitor\"\nDataset: gttm [CDS]\nAspect: uncertainty_q01 | Category: Applications\nSection: Assessing global warming with the C3S Global Temperature Trend Monitor > ℹ️ If you want to know more > Key resources\n---\nMore about the ERA5 reanalysis:\n* [The ERA5 global reanalysis](https://doi.org/10.1002/qj.3803)\n* [ERA5: uncertainty estimation](https://confluence.ecmwf.int/display/CKB/ERA5%3A+uncertainty+estimation)\n\nMore about temperature anomalies:\n* [Climate Bulletin – About the data and analysis](https://climate.copernicus.eu/climate-bulletin-about-data-and-analysis)\n\nMore about climate trend estimation:\n* [Trend analysis of climate time series: A review of methods](https://doi.org/10.1016/j.earscirev.2018.12.005)\n* [Estimating trends and the current climate mean in a changing climate](https://doi.org/10.1016/j.cliser.2023.100428)\n\nMore about the 1.5 °C and 2 °C targets:\n* [Paris Agreement](https://treaties.un.org/doc/Treaties/2016/02/20160215%2006-03%20PM/Ch_XXVII-7-d.pdf)\n* [A history of the 1.5 °C target](https://doi.org/10.1002/wcc.824)\n* [Realizing the impacts of a 1.5 °C warmer world](https://doi.org/10.1038/nclimate3055)\n\nCode libraries used:\n* [NumPy polyfit](https://numpy.org/doc/stable/reference/generated/numpy.polyfit.html)\n* [scikit-learn Theil–Sen estimator](https://scikit-learn.org/stable/modules/generated/sklearn.linear_model.TheilSenRegressor.html)"} {"chunk_id": "application_gttm_uncertainty_q01__8261e183ed58", "report_id": "application_gttm_uncertainty_q01", "dataset_id": "gttm", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty_q01", "aspect_base": "uncertainty", "category": "Applications", "match_confidence": "unmatched", "section": "Assessing global warming with the C3S Global Temperature Trend Monitor > ℹ️ If you want to know more > References", "title": "Assessing global warming with the C3S Global Temperature Trend Monitor", "chunk_index": 9, "token_count": 1025, "text_raw": "[[Cavalleri+24](https://doi.org/10.1002/joc.8475)] F. Cavalleri et al., ‘Inter-comparison and validation of high-resolution surface air temperature reanalysis fields over Italy’, International Journal of Climatology, vol. 44, no. 8, pp. 2681–2700, 2024, doi: 10.1002/joc.8475.\n\n[[Cointe+23](https://doi.org/10.1002/wcc.824)] B. Cointe and H. Guillemot, ‘A history of the 1.5°C target’, WIREs Climate Change, vol. 14, no. 3, p. e824, 2023, doi: 10.1002/wcc.824.\n\n[[Copernicus Climate Change Service+24](https://doi.org/10.24381/bs9v-8c66)] Copernicus Climate Change Service (C3S), ‘European State of the Climate 2023’, Copernicus Climate Change Service (C3S), Apr. 2024. doi: 10.24381/bs9v-8c66.\n\n[[Copernicus Climate Change Service+25a](https://confluence.ecmwf.int/display/CKB/ERA5%3A+uncertainty+estimation)] Copernicus Climate Change Service (C3S), ‘ERA5: uncertainty estimation’, Copernicus Knowledge Base. Accessed: May 21, 2025. [Online]. Available: https://confluence.ecmwf.int/display/CKB/ERA5%3A+uncertainty+estimation\n\n[[Copernicus Climate Change Service+25b](https://climate.copernicus.eu/climate-bulletin-about-data-and-analysis)] Copernicus Climate Change Service (C3S), ‘Climate Bulletin – About the data and analysis’, Copernicus Climate Change Service (C3S). Accessed: May 13, 2025. [Online]. Available: https://climate.copernicus.eu/climate-bulletin-about-data-and-analysis\n\n[[ECVs for climate variability dataset](https://doi.org/10.24381/7470b643)] H. Hersbach et al., ‘Essential climate variables for assessment of climate variability from 1979 to present’. Copernicus Climate Change Service (C3S) Climate Data Store (CDS), Apr. 18, 2019. doi: 10.24381/7470b643.\n\n[[ERA5 monthly dataset](https://doi.org/10.24381/cds.f17050d7)] H. Hersbach et al., ‘ERA5 monthly averaged data on single levels from 1940 to present’. Copernicus Climate Change Service (C3S) Climate Data Store (CDS), Apr. 18, 2019. doi: 10.24381/cds.f17050d7.\n\n[[Frijters+25](https://www.volkskrant.nl/wetenschap/~b301f81f2/)] S. Frijters, ‘366 warme dagen: het temperatuurrecord van 2024 in cijfers’, De Volkskrant, Jan. 10, 2025. Accessed: May 09, 2025. [Online]. Available: https://www.volkskrant.nl/wetenschap/~b301f81f2/\n\n[[Hersbach+20](https://doi.org/10.1002/qj.3803)] H. Hersbach et al., ‘The ERA5 global reanalysis’, Quarterly Journal of the Royal Meteorological Society, vol. 146, no. 730, pp. 1999–2049, May 2020, doi: 10.1002/qj.3803.\n\n[[Horton+24](https://www.theguardian.com/environment/2024/nov/20/the-climate-crisis-in-charts-how-2024-has-set-unwanted-new-records)] H. Horton, L. Swan, A. L. González Paz, and H. Symons, ‘The climate crisis in charts: how 2024 has set unwanted new records’, The Guardian, Nov. 20, 2024. Accessed: May 09, 2025. [Online]. Available: https://www.theguardian.com/environment/2024/nov/20/the-climate-crisis-in-charts-how-2024-has-set-unwanted-new-records", "text_with_prefix": "EQC Quality Assessment: \"Assessing global warming with the C3S Global Temperature Trend Monitor\"\nDataset: gttm [CDS]\nAspect: uncertainty_q01 | Category: Applications\nSection: Assessing global warming with the C3S Global Temperature Trend Monitor > ℹ️ If you want to know more > References\n---\n[[Cavalleri+24](https://doi.org/10.1002/joc.8475)] F. Cavalleri et al., ‘Inter-comparison and validation of high-resolution surface air temperature reanalysis fields over Italy’, International Journal of Climatology, vol. 44, no. 8, pp. 2681–2700, 2024, doi: 10.1002/joc.8475.\n\n[[Cointe+23](https://doi.org/10.1002/wcc.824)] B. Cointe and H. Guillemot, ‘A history of the 1.5°C target’, WIREs Climate Change, vol. 14, no. 3, p. e824, 2023, doi: 10.1002/wcc.824.\n\n[[Copernicus Climate Change Service+24](https://doi.org/10.24381/bs9v-8c66)] Copernicus Climate Change Service (C3S), ‘European State of the Climate 2023’, Copernicus Climate Change Service (C3S), Apr. 2024. doi: 10.24381/bs9v-8c66.\n\n[[Copernicus Climate Change Service+25a](https://confluence.ecmwf.int/display/CKB/ERA5%3A+uncertainty+estimation)] Copernicus Climate Change Service (C3S), ‘ERA5: uncertainty estimation’, Copernicus Knowledge Base. Accessed: May 21, 2025. [Online]. Available: https://confluence.ecmwf.int/display/CKB/ERA5%3A+uncertainty+estimation\n\n[[Copernicus Climate Change Service+25b](https://climate.copernicus.eu/climate-bulletin-about-data-and-analysis)] Copernicus Climate Change Service (C3S), ‘Climate Bulletin – About the data and analysis’, Copernicus Climate Change Service (C3S). Accessed: May 13, 2025. [Online]. Available: https://climate.copernicus.eu/climate-bulletin-about-data-and-analysis\n\n[[ECVs for climate variability dataset](https://doi.org/10.24381/7470b643)] H. Hersbach et al., ‘Essential climate variables for assessment of climate variability from 1979 to present’. Copernicus Climate Change Service (C3S) Climate Data Store (CDS), Apr. 18, 2019. doi: 10.24381/7470b643.\n\n[[ERA5 monthly dataset](https://doi.org/10.24381/cds.f17050d7)] H. Hersbach et al., ‘ERA5 monthly averaged data on single levels from 1940 to present’. Copernicus Climate Change Service (C3S) Climate Data Store (CDS), Apr. 18, 2019. doi: 10.24381/cds.f17050d7.\n\n[[Frijters+25](https://www.volkskrant.nl/wetenschap/~b301f81f2/)] S. Frijters, ‘366 warme dagen: het temperatuurrecord van 2024 in cijfers’, De Volkskrant, Jan. 10, 2025. Accessed: May 09, 2025. [Online]. Available: https://www.volkskrant.nl/wetenschap/~b301f81f2/\n\n[[Hersbach+20](https://doi.org/10.1002/qj.3803)] H. Hersbach et al., ‘The ERA5 global reanalysis’, Quarterly Journal of the Royal Meteorological Society, vol. 146, no. 730, pp. 1999–2049, May 2020, doi: 10.1002/qj.3803.\n\n[[Horton+24](https://www.theguardian.com/environment/2024/nov/20/the-climate-crisis-in-charts-how-2024-has-set-unwanted-new-records)] H. Horton, L. Swan, A. L. González Paz, and H. Symons, ‘The climate crisis in charts: how 2024 has set unwanted new records’, The Guardian, Nov. 20, 2024. Accessed: May 09, 2025. [Online]. Available: https://www.theguardian.com/environment/2024/nov/20/the-climate-crisis-in-charts-how-2024-has-set-unwanted-new-records"} {"chunk_id": "application_gttm_uncertainty_q01__a9107d15659a", "report_id": "application_gttm_uncertainty_q01", "dataset_id": "gttm", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty_q01", "aspect_base": "uncertainty", "category": "Applications", "match_confidence": "unmatched", "section": "Assessing global warming with the C3S Global Temperature Trend Monitor > ℹ️ If you want to know more > References", "title": "Assessing global warming with the C3S Global Temperature Trend Monitor", "chunk_index": 10, "token_count": 1112, "text_raw": "09, 2025. [Online]. Available: https://www.theguardian.com/environment/2024/nov/20/the-climate-crisis-in-charts-how-2024-has-set-unwanted-new-records\n\n[[IPCC+23](https://doi.org/10.59327/IPCC/AR6-9789291691647)] IPCC, ‘Climate Change 2023: Synthesis Report. Contribution of Working Groups I, II and III to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change’, Intergovernmental Panel on Climate Change (IPCC), Geneva, Switzerland, Jul. 2023. doi: 10.59327/IPCC/AR6-9789291691647.\n\n[[Liu+24](https://doi.org/10.1016/j.ecolind.2024.112481)] R. Liu et al., ‘Global-scale ERA5 product precipitation and temperature evaluation’, Ecological Indicators, vol. 166, p. 112481, Sep. 2024, doi: 10.1016/j.ecolind.2024.112481.\n\n[[Marlon+21](https://doi.org/10.1016/j.gloenvcha.2021.102247)] J. R. Marlon et al., ‘Hot dry days increase perceived experience with global warming’, Global Environmental Change, vol. 68, p. 102247, May 2021, doi: 10.1016/j.gloenvcha.2021.102247.\n\n[[Mitchell+16](https://doi.org/10.1038/nclimate3055)] D. Mitchell, R. James, P. M. Forster, R. A. Betts, H. Shiogama, and M. Allen, ‘Realizing the impacts of a 1.5 °C warmer world’, Nature Climate Change, vol. 6, no. 8, pp. 735–737, Aug. 2016, doi: 10.1038/nclimate3055.\n\n[[Mudelsee+19](https://doi.org/10.1016/j.earscirev.2018.12.005)] M. Mudelsee, ‘Trend analysis of climate time series: A review of methods’, Earth-Science Reviews, vol. 190, pp. 310–322, Mar. 2019, doi: 10.1016/j.earscirev.2018.12.005.\n\n[[O’Neill+16](https://doi.org/10.5194/gmd-9-3461-2016)] B. C. O’Neill et al., ‘The Scenario Model Intercomparison Project (ScenarioMIP) for CMIP6’, Geoscientific Model Development, vol. 9, pp. 3461–3482, Sep. 2016, doi: 10.5194/gmd-9-3461-2016.\n\n[[Poynting+25](https://www.bbc.com/news/articles/cd7575x8yq5o)] M. Poynting, E. Rivault, and B. Dale, ‘World’s hottest year: 2024 first to pass 1.5C warming limit’, BBC News, Jan. 10, 2025. Accessed: May 09, 2025. [Online]. Available: https://www.bbc.com/news/articles/cd7575x8yq5o\n\n[[Rantanen+22](https://doi.org/10.1038/s43247-022-00498-3)] M. Rantanen et al., ‘The Arctic has warmed nearly four times faster than the globe since 1979’, Communications Earth & Environment, vol. 3, p. 168, Aug. 2022, doi: 10.1038/s43247-022-00498-3.\n\n[[Scherrer+24](https://doi.org/10.1016/j.cliser.2023.100428)] S. C. Scherrer, C. de Valk, M. Begert, S. Gubler, S. Kotlarski, and M. Croci-Maspoli, ‘Estimating trends and the current climate mean in a changing climate’, Climate Services, vol. 33, p. 100428, Jan. 2024, doi: 10.1016/j.cliser.2023.100428.\n\n[[Sgubin+17](https://doi.org/10.1038/ncomms14375)] G. Sgubin, D. Swingedouw, S. Drijfhout, Y. Mary, and A. Bennabi, ‘Abrupt cooling over the North Atlantic in modern climate models’, Nature Communications, vol. 8, p. 14375, Feb. 2017, doi: 10.1038/ncomms14375.", "text_with_prefix": "EQC Quality Assessment: \"Assessing global warming with the C3S Global Temperature Trend Monitor\"\nDataset: gttm [CDS]\nAspect: uncertainty_q01 | Category: Applications\nSection: Assessing global warming with the C3S Global Temperature Trend Monitor > ℹ️ If you want to know more > References\n---\n09, 2025. [Online]. Available: https://www.theguardian.com/environment/2024/nov/20/the-climate-crisis-in-charts-how-2024-has-set-unwanted-new-records\n\n[[IPCC+23](https://doi.org/10.59327/IPCC/AR6-9789291691647)] IPCC, ‘Climate Change 2023: Synthesis Report. Contribution of Working Groups I, II and III to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change’, Intergovernmental Panel on Climate Change (IPCC), Geneva, Switzerland, Jul. 2023. doi: 10.59327/IPCC/AR6-9789291691647.\n\n[[Liu+24](https://doi.org/10.1016/j.ecolind.2024.112481)] R. Liu et al., ‘Global-scale ERA5 product precipitation and temperature evaluation’, Ecological Indicators, vol. 166, p. 112481, Sep. 2024, doi: 10.1016/j.ecolind.2024.112481.\n\n[[Marlon+21](https://doi.org/10.1016/j.gloenvcha.2021.102247)] J. R. Marlon et al., ‘Hot dry days increase perceived experience with global warming’, Global Environmental Change, vol. 68, p. 102247, May 2021, doi: 10.1016/j.gloenvcha.2021.102247.\n\n[[Mitchell+16](https://doi.org/10.1038/nclimate3055)] D. Mitchell, R. James, P. M. Forster, R. A. Betts, H. Shiogama, and M. Allen, ‘Realizing the impacts of a 1.5 °C warmer world’, Nature Climate Change, vol. 6, no. 8, pp. 735–737, Aug. 2016, doi: 10.1038/nclimate3055.\n\n[[Mudelsee+19](https://doi.org/10.1016/j.earscirev.2018.12.005)] M. Mudelsee, ‘Trend analysis of climate time series: A review of methods’, Earth-Science Reviews, vol. 190, pp. 310–322, Mar. 2019, doi: 10.1016/j.earscirev.2018.12.005.\n\n[[O’Neill+16](https://doi.org/10.5194/gmd-9-3461-2016)] B. C. O’Neill et al., ‘The Scenario Model Intercomparison Project (ScenarioMIP) for CMIP6’, Geoscientific Model Development, vol. 9, pp. 3461–3482, Sep. 2016, doi: 10.5194/gmd-9-3461-2016.\n\n[[Poynting+25](https://www.bbc.com/news/articles/cd7575x8yq5o)] M. Poynting, E. Rivault, and B. Dale, ‘World’s hottest year: 2024 first to pass 1.5C warming limit’, BBC News, Jan. 10, 2025. Accessed: May 09, 2025. [Online]. Available: https://www.bbc.com/news/articles/cd7575x8yq5o\n\n[[Rantanen+22](https://doi.org/10.1038/s43247-022-00498-3)] M. Rantanen et al., ‘The Arctic has warmed nearly four times faster than the globe since 1979’, Communications Earth & Environment, vol. 3, p. 168, Aug. 2022, doi: 10.1038/s43247-022-00498-3.\n\n[[Scherrer+24](https://doi.org/10.1016/j.cliser.2023.100428)] S. C. Scherrer, C. de Valk, M. Begert, S. Gubler, S. Kotlarski, and M. Croci-Maspoli, ‘Estimating trends and the current climate mean in a changing climate’, Climate Services, vol. 33, p. 100428, Jan. 2024, doi: 10.1016/j.cliser.2023.100428.\n\n[[Sgubin+17](https://doi.org/10.1038/ncomms14375)] G. Sgubin, D. Swingedouw, S. Drijfhout, Y. Mary, and A. Bennabi, ‘Abrupt cooling over the North Atlantic in modern climate models’, Nature Communications, vol. 8, p. 14375, Feb. 2017, doi: 10.1038/ncomms14375."} {"chunk_id": "application_gttm_uncertainty_q01__8567b6159a0a", "report_id": "application_gttm_uncertainty_q01", "dataset_id": "gttm", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty_q01", "aspect_base": "uncertainty", "category": "Applications", "match_confidence": "unmatched", "section": "Assessing global warming with the C3S Global Temperature Trend Monitor > ℹ️ If you want to know more > References", "title": "Assessing global warming with the C3S Global Temperature Trend Monitor", "chunk_index": 11, "token_count": 543, "text_raw": "Bennabi, ‘Abrupt cooling over the North Atlantic in modern climate models’, Nature Communications, vol. 8, p. 14375, Feb. 2017, doi: 10.1038/ncomms14375.\n\n[[Shen+22](https://doi.org/10.1007/s11442-022-1937-1)] B. Shen, S. Song, L. Zhang, Z. Wang, C. Ren, and Y. Li, ‘Temperature trends in some major countries from the 1980s to 2019’, Journal of Geographical Sciences, vol. 32, no. 1, pp. 79–100, Jan. 2022, doi: 10.1007/s11442-022-1937-1.\n\n[[United Nations+15](https://treaties.un.org/doc/Treaties/2016/02/20160215%2006-03%20PM/Ch_XXVII-7-d.pdf)] United Nations, ‘Paris Agreement’, United Nations, Paris, France, Dec. 2015. Accessed: May 12, 2025. [Online]. Available: https://treaties.un.org/doc/Treaties/2016/02/20160215%2006-03%20PM/Ch_XXVII-7-d.pdf\n\n[[Yilmaz+23](https://doi.org/10.1016/j.scitotenv.2022.159182)] M. Yilmaz, ‘Accuracy assessment of temperature trends from ERA5 and ERA5-Land’, Science of The Total Environment, vol. 856, p. 159182, Jan. 2023, doi: 10.1016/j.scitotenv.2022.159182.\n\n[[WMO](https://community.wmo.int/en/activity-areas/climate-services/climate-products-and-initiatives/wmo-climatological-normals)] World Meteorological Organization, ‘WMO Climatological Normals’, World Meteorological Organization. Accessed: Jun. 20, 2025. [Online]. Available: https://community.wmo.int/en/activity-areas/climate-services/climate-products-and-initiatives/wmo-climatological-normals", "text_with_prefix": "EQC Quality Assessment: \"Assessing global warming with the C3S Global Temperature Trend Monitor\"\nDataset: gttm [CDS]\nAspect: uncertainty_q01 | Category: Applications\nSection: Assessing global warming with the C3S Global Temperature Trend Monitor > ℹ️ If you want to know more > References\n---\nBennabi, ‘Abrupt cooling over the North Atlantic in modern climate models’, Nature Communications, vol. 8, p. 14375, Feb. 2017, doi: 10.1038/ncomms14375.\n\n[[Shen+22](https://doi.org/10.1007/s11442-022-1937-1)] B. Shen, S. Song, L. Zhang, Z. Wang, C. Ren, and Y. Li, ‘Temperature trends in some major countries from the 1980s to 2019’, Journal of Geographical Sciences, vol. 32, no. 1, pp. 79–100, Jan. 2022, doi: 10.1007/s11442-022-1937-1.\n\n[[United Nations+15](https://treaties.un.org/doc/Treaties/2016/02/20160215%2006-03%20PM/Ch_XXVII-7-d.pdf)] United Nations, ‘Paris Agreement’, United Nations, Paris, France, Dec. 2015. Accessed: May 12, 2025. [Online]. Available: https://treaties.un.org/doc/Treaties/2016/02/20160215%2006-03%20PM/Ch_XXVII-7-d.pdf\n\n[[Yilmaz+23](https://doi.org/10.1016/j.scitotenv.2022.159182)] M. Yilmaz, ‘Accuracy assessment of temperature trends from ERA5 and ERA5-Land’, Science of The Total Environment, vol. 856, p. 159182, Jan. 2023, doi: 10.1016/j.scitotenv.2022.159182.\n\n[[WMO](https://community.wmo.int/en/activity-areas/climate-services/climate-products-and-initiatives/wmo-climatological-normals)] World Meteorological Organization, ‘WMO Climatological Normals’, World Meteorological Organization. Accessed: Jun. 20, 2025. [Online]. Available: https://community.wmo.int/en/activity-areas/climate-services/climate-products-and-initiatives/wmo-climatological-normals"} {"chunk_id": "application_thermaltrace_extreme-events_q01__174c00dda495", "report_id": "application_thermaltrace_extreme-events_q01", "dataset_id": "thermaltrace", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extreme-events_q01", "aspect_base": "extreme-events", "category": "Applications", "match_confidence": "unmatched", "section": "Visualisation of heat and cold stress with Thermal Trace", "title": "Visualisation of heat and cold stress with Thermal Trace", "chunk_index": 0, "token_count": 83, "text_raw": "Production date: 2026-06-15.\n\nProduced by: Olivier Burggraaff (National Physical Laboratory).", "text_with_prefix": "EQC Quality Assessment: \"Visualisation of heat and cold stress with Thermal Trace\"\nDataset: thermaltrace [CDS]\nAspect: extreme-events_q01 | Category: Applications\nSection: Visualisation of heat and cold stress with Thermal Trace\n---\nProduction date: 2026-06-15.\n\nProduced by: Olivier Burggraaff (National Physical Laboratory)."} {"chunk_id": "application_thermaltrace_extreme-events_q01__398d90b0a862", "report_id": "application_thermaltrace_extreme-events_q01", "dataset_id": "thermaltrace", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extreme-events_q01", "aspect_base": "extreme-events", "category": "Applications", "match_confidence": "unmatched", "section": "Visualisation of heat and cold stress with Thermal Trace > Quality assessment question", "title": "Visualisation of heat and cold stress with Thermal Trace", "chunk_index": 1, "token_count": 914, "text_raw": "* **Is the Thermal Trace application an appropriate tool for broad audiences to visualise heat and cold stress relating to extreme temperature events, including heatwaves and cold snaps?**\n* **Is the Thermal Trace application an appropriate tool for broad audiences to visualise long-term trends in heat and cold stress, such as those relating to climate change?**\n\nHeat stress and cold stress occur when the human body faces extreme temperatures outside its comfort zone.\nThese thermal extremes cause widespread, severe impacts on human health, society, and regional economies [[Bell+18](https://doi.org/10.1080/10962247.2017.1401017)].\nSpecifically, global data show that cold temperatures cause about 4 594 000 deaths annually,\nwhile extreme heat contributes to another 489 000 deaths [[Zhao+21](https://doi.org/10.1016/S2542-5196%2821%2900081-4)].\nAs climate change accelerates,\nthese dangerous meteorological events are growing increasingly frequent and intense.\nConsequently, there is greater desire for decision-makers, urban planners, and the general public to be able to visualise and track heat and cold stress-health trends over time,\nto understand and plan for their widespread impacts on human health, society, and the economy.\n\nThe [_Thermal Trace_](https://apps.climate.copernicus.eu/overview?app=thermal-trace) application provides near real-time visualisations of heat and cold stress indicators as well as daily minimum/maximum feels-like temperature based on the ERA5 reanalysis [[Menary+25](https://www.ecmwf.int/en/newsletter/185/news/thermal-trace-health-related-weather-and-climate-monitoring)].\nThe application is focused on the human experience of temperature and weather rather than solely the physical air temperature.\nFor physical surface air and sea surface temperatures,\nusers may want to use the [_Climate Pulse_ application](https://apps.climate.copernicus.eu/overview?app=climate-pulse),\nfor which there is an equivalent [quality assessment](./application_climate-pulse_extreme-events_q01) investigating its suitability for visualising extreme weather events.\n\n_Feels-like temperature_\n(or _apparent temperature_)\nexpresses how humans _experience_ temperature by including additional variables relating to vapour pressure, wind speed, and irradiation [[Steadman+84](https://doi.org/10.1175/1520-0450%281984%29023%3C1674:AUSOAT%3E2.0.CO;2)].\n\nThe _Universal Thermal Climate Index_ (_UTCI_) is an international standard feels-like temperature indicator\nthat quantifies the thermal stress experienced by a representative human in given meteorological circumstances.\nUTCI is expressed quantitatively as an equivalent temperature (in K or °C) which acts as a feels-like temperature\n– e.g. a UTCI of 30 °C under any circumstances causes stress equivalent to an air temperature of 30 °C under standard circumstances [[Jendritzky+12](https://doi.org/10.1007/s00484-011-0513-7), [Błażejczyk+10](https://doi.org/10.2478/mgrsd-2010-0009)].\nIt is important to note that UTCI assumes an \"average person\" [[Fiala+12](https://doi.org/10.1007/s00484-011-0424-7)]\nwearing clothing typical for given weather conditions [[Havenith+12](https://doi.org/10.1007/s00484-011-0451-4)]\nand may be less representative especially for vulnerable groups, such as children, the elderly, and people with reduced thermoregulation [[Tousi+24](https://doi.org/10.3390/urbansci8040193)].\n\nThis assessment examines the suitability of the Thermal Trace application for visualising heat and cold stress,\nboth during extreme events like heatwaves and cold snaps and over longer periods of time,\nfor a broad (scientific and non-scientific) audience.", "text_with_prefix": "EQC Quality Assessment: \"Visualisation of heat and cold stress with Thermal Trace\"\nDataset: thermaltrace [CDS]\nAspect: extreme-events_q01 | Category: Applications\nSection: Visualisation of heat and cold stress with Thermal Trace > Quality assessment question\n---\n* **Is the Thermal Trace application an appropriate tool for broad audiences to visualise heat and cold stress relating to extreme temperature events, including heatwaves and cold snaps?**\n* **Is the Thermal Trace application an appropriate tool for broad audiences to visualise long-term trends in heat and cold stress, such as those relating to climate change?**\n\nHeat stress and cold stress occur when the human body faces extreme temperatures outside its comfort zone.\nThese thermal extremes cause widespread, severe impacts on human health, society, and regional economies [[Bell+18](https://doi.org/10.1080/10962247.2017.1401017)].\nSpecifically, global data show that cold temperatures cause about 4 594 000 deaths annually,\nwhile extreme heat contributes to another 489 000 deaths [[Zhao+21](https://doi.org/10.1016/S2542-5196%2821%2900081-4)].\nAs climate change accelerates,\nthese dangerous meteorological events are growing increasingly frequent and intense.\nConsequently, there is greater desire for decision-makers, urban planners, and the general public to be able to visualise and track heat and cold stress-health trends over time,\nto understand and plan for their widespread impacts on human health, society, and the economy.\n\nThe [_Thermal Trace_](https://apps.climate.copernicus.eu/overview?app=thermal-trace) application provides near real-time visualisations of heat and cold stress indicators as well as daily minimum/maximum feels-like temperature based on the ERA5 reanalysis [[Menary+25](https://www.ecmwf.int/en/newsletter/185/news/thermal-trace-health-related-weather-and-climate-monitoring)].\nThe application is focused on the human experience of temperature and weather rather than solely the physical air temperature.\nFor physical surface air and sea surface temperatures,\nusers may want to use the [_Climate Pulse_ application](https://apps.climate.copernicus.eu/overview?app=climate-pulse),\nfor which there is an equivalent [quality assessment](./application_climate-pulse_extreme-events_q01) investigating its suitability for visualising extreme weather events.\n\n_Feels-like temperature_\n(or _apparent temperature_)\nexpresses how humans _experience_ temperature by including additional variables relating to vapour pressure, wind speed, and irradiation [[Steadman+84](https://doi.org/10.1175/1520-0450%281984%29023%3C1674:AUSOAT%3E2.0.CO;2)].\n\nThe _Universal Thermal Climate Index_ (_UTCI_) is an international standard feels-like temperature indicator\nthat quantifies the thermal stress experienced by a representative human in given meteorological circumstances.\nUTCI is expressed quantitatively as an equivalent temperature (in K or °C) which acts as a feels-like temperature\n– e.g. a UTCI of 30 °C under any circumstances causes stress equivalent to an air temperature of 30 °C under standard circumstances [[Jendritzky+12](https://doi.org/10.1007/s00484-011-0513-7), [Błażejczyk+10](https://doi.org/10.2478/mgrsd-2010-0009)].\nIt is important to note that UTCI assumes an \"average person\" [[Fiala+12](https://doi.org/10.1007/s00484-011-0424-7)]\nwearing clothing typical for given weather conditions [[Havenith+12](https://doi.org/10.1007/s00484-011-0451-4)]\nand may be less representative especially for vulnerable groups, such as children, the elderly, and people with reduced thermoregulation [[Tousi+24](https://doi.org/10.3390/urbansci8040193)].\n\nThis assessment examines the suitability of the Thermal Trace application for visualising heat and cold stress,\nboth during extreme events like heatwaves and cold snaps and over longer periods of time,\nfor a broad (scientific and non-scientific) audience."} {"chunk_id": "application_thermaltrace_extreme-events_q01__a06803543f03", "report_id": "application_thermaltrace_extreme-events_q01", "dataset_id": "thermaltrace", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extreme-events_q01", "aspect_base": "extreme-events", "category": "Applications", "match_confidence": "unmatched", "section": "Visualisation of heat and cold stress with Thermal Trace > Quality assessment statement", "title": "Visualisation of heat and cold stress with Thermal Trace", "chunk_index": 2, "token_count": 293, "text_raw": "These are the key outcomes of this assessment\n\n* The Thermal Trace application is a scientifically sound and convenient tool for visualising feels-like temperature (UTCI) and thermal stress indicators. Its visuals can be used directly (via screenshot or image export) or it can be used as a jumping-off point for data download and quantitative analysis or custom visualisation.\n* Thermal Trace is suitable for visualising extreme events, like heatwaves and cold snaps, both recently and decades in the past. Its daily to yearly maps and hourly to yearly time series provide sufficient coverage and resolution to see the spatial extent and temporal evolution of an event. When using the application, it is important to note the known biases and uncertainties in the underpinning ERA5 data, particularly pre-1979 and in areas where ERA5 has fewer input observations.\n* Thermal Trace is suitable for visualising long-term trends in thermal stress, reproducing known trends in different locations. Its yearly summary statistics time series are particularly useful for visualising the number of days with certain heat/cold stress levels and tropical nights, important variables for health and climate policy. Users should be aware of known biases and temporal inconsistencies in ERA5.\n```", "text_with_prefix": "EQC Quality Assessment: \"Visualisation of heat and cold stress with Thermal Trace\"\nDataset: thermaltrace [CDS]\nAspect: extreme-events_q01 | Category: Applications\nSection: Visualisation of heat and cold stress with Thermal Trace > Quality assessment statement\n---\nThese are the key outcomes of this assessment\n\n* The Thermal Trace application is a scientifically sound and convenient tool for visualising feels-like temperature (UTCI) and thermal stress indicators. Its visuals can be used directly (via screenshot or image export) or it can be used as a jumping-off point for data download and quantitative analysis or custom visualisation.\n* Thermal Trace is suitable for visualising extreme events, like heatwaves and cold snaps, both recently and decades in the past. Its daily to yearly maps and hourly to yearly time series provide sufficient coverage and resolution to see the spatial extent and temporal evolution of an event. When using the application, it is important to note the known biases and uncertainties in the underpinning ERA5 data, particularly pre-1979 and in areas where ERA5 has fewer input observations.\n* Thermal Trace is suitable for visualising long-term trends in thermal stress, reproducing known trends in different locations. Its yearly summary statistics time series are particularly useful for visualising the number of days with certain heat/cold stress levels and tropical nights, important variables for health and climate policy. Users should be aware of known biases and temporal inconsistencies in ERA5.\n```"} {"chunk_id": "application_thermaltrace_extreme-events_q01__ace7b4266acf", "report_id": "application_thermaltrace_extreme-events_q01", "dataset_id": "thermaltrace", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extreme-events_q01", "aspect_base": "extreme-events", "category": "Applications", "match_confidence": "unmatched", "section": "Visualisation of heat and cold stress with Thermal Trace > Methodology", "title": "Visualisation of heat and cold stress with Thermal Trace", "chunk_index": 3, "token_count": 945, "text_raw": "Thermal Trace is an interactive application designed for a broad audience to visualise global feels-like temperature and UTCI.\nIt displays daily, monthly, seasonal, and yearly maps along with time series, yearly statistics, and climate stripes\n(Figures {numref}`{number} `–{numref}`{number} `).\n\nThermal Trace is based on the [_Thermal comfort indices derived from ERA5 reanalysis_](https://doi.org/10.24381/cds.553b7518) ([derived-utci-historical](https://doi.org/10.24381/cds.553b7518); ERA5-HEAT) dataset [[Di Napoli+21a](https://doi.org/10.1002/gdj3.102)],\nwhich is derived from the ERA5 reanalysis [[Hersbach+20](https://doi.org/10.1002/qj.3803)].\nData are available from 1940 up to near real-time (typical delay of about two days),\nincluding preliminary data from ERA5T,\non a regular 0.25° × 0.25° grid, approximately 30 km × 30 km.\nFor some variables,\ndata can be displayed not only as absolute values but also as anomalies relative to the 1991–2020 average,\nthe standard reference period used by the World Meteorological Organization (WMO).\n\nWithin the map and time series views,\nUTCI values are displayed both quantitatively and categorically.\nThe stress categories and colours used in Thermal Trace are as follows [[Di Napoli+19](https://doi.org/10.1175/JAMC-D-18-0246.1)]:\n\n\n\n\n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n
Stress categoryUTCI range [°C]
Extreme cold stressLower than –40
Very strong cold stress–40 to –27
Strong cold stress–27 to –13
Moderate cold stress–13 to 0
Slight cold stress0 to 9
None9 to 26
Moderate heat stress26 to 32
Strong heat stress32 to 38
Very strong heat stress38 to 46
Extreme heat stressHigher than 46
\n
", "text_with_prefix": "EQC Quality Assessment: \"Visualisation of heat and cold stress with Thermal Trace\"\nDataset: thermaltrace [CDS]\nAspect: extreme-events_q01 | Category: Applications\nSection: Visualisation of heat and cold stress with Thermal Trace > Methodology\n---\nThermal Trace is an interactive application designed for a broad audience to visualise global feels-like temperature and UTCI.\nIt displays daily, monthly, seasonal, and yearly maps along with time series, yearly statistics, and climate stripes\n(Figures {numref}`{number} `–{numref}`{number} `).\n\nThermal Trace is based on the [_Thermal comfort indices derived from ERA5 reanalysis_](https://doi.org/10.24381/cds.553b7518) ([derived-utci-historical](https://doi.org/10.24381/cds.553b7518); ERA5-HEAT) dataset [[Di Napoli+21a](https://doi.org/10.1002/gdj3.102)],\nwhich is derived from the ERA5 reanalysis [[Hersbach+20](https://doi.org/10.1002/qj.3803)].\nData are available from 1940 up to near real-time (typical delay of about two days),\nincluding preliminary data from ERA5T,\non a regular 0.25° × 0.25° grid, approximately 30 km × 30 km.\nFor some variables,\ndata can be displayed not only as absolute values but also as anomalies relative to the 1991–2020 average,\nthe standard reference period used by the World Meteorological Organization (WMO).\n\nWithin the map and time series views,\nUTCI values are displayed both quantitatively and categorically.\nThe stress categories and colours used in Thermal Trace are as follows [[Di Napoli+19](https://doi.org/10.1175/JAMC-D-18-0246.1)]:\n\n\n\n\n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n
Stress categoryUTCI range [°C]
Extreme cold stressLower than –40
Very strong cold stress–40 to –27
Strong cold stress–27 to –13
Moderate cold stress–13 to 0
Slight cold stress0 to 9
None9 to 26
Moderate heat stress26 to 32
Strong heat stress32 to 38
Very strong heat stress38 to 46
Extreme heat stressHigher than 46
\n
"} {"chunk_id": "application_thermaltrace_extreme-events_q01__3656969cc7d7", "report_id": "application_thermaltrace_extreme-events_q01", "dataset_id": "thermaltrace", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extreme-events_q01", "aspect_base": "extreme-events", "category": "Applications", "match_confidence": "unmatched", "section": "Visualisation of heat and cold stress with Thermal Trace > Methodology", "title": "Visualisation of heat and cold stress with Thermal Trace", "chunk_index": 4, "token_count": 948, "text_raw": "=\"background-color:#993333; color:white;\">\n Extreme heat stress\n Higher than 46\n \n \n\n
\n\nOutputs from Thermal Trace can be exported in four ways:\n1. Screenshot of the map view (Figure {numref}`{number} `)\n2. Time series exports (Figure {numref}`{number} `)\n3. Time series data download\n4. Linking directly to a specific view in the application by copying the current URL, e.g. [daily peak heat stress in Reading, UK, on 20 August 2025](https://thermaltrace.climate.copernicus.eu/?agg=daily&anomaly=false&date=2025-08-20&heat=true&lat=51.46&lng=-0.97&variable=utci_daily_max_cat).\n\nIt is not possible to export the map view or download the corresponding data\n(global for one time stamp)\ndirectly from Thermal Trace.\nInstead, this can be done from the [ERA5-HEAT CDS catalogue entry](https://doi.org/10.24381/cds.553b7518).\n\n::::{tab-set}\n:::{tab-item} Map view screenshots\n:sync: example_map\n attachment:application_thermaltrace_example_map.png\n---\nname: application_thermaltrace_example_map\n---\nExample screenshots of the Thermal Trace map view, from top to bottom: \na) Daily peak heat stress on 20 August 2025, centered on Reading, UK (app view). \nb) Number of days with at least strong cold stress in December 2025–February 2026, centered on Reading, UK (app view).\n```\n:::\n:::{tab-item} Time series exports\n:sync: example_timeseries\n attachment:application_thermaltrace_example_timeseries.png\n---\nname: application_thermaltrace_example_timeseries\n---\nExample time series exports from the Thermal Trace, from top to bottom: \na) Hourly temperature and UTCI (feels-like temperature) with heat stress categories on 18–22 August 2025, in Reading, UK (app view). \nb) Yearly number of days with a given heat stress level, and yearly number of tropical nights, in Reading, UK (app view).\n```\n:::\n::::\n\nApplications are quality assessed separately from their underpinning datasets. For information on quality attributes of the underlying datasets, the reader is referred to the quality assessment of the datasets, which can be found in their respective CDS catalogue entries or [on the EQC hub](../intro).\n\nThe analysis and results are organised in the following use cases, which are detailed in the sections below:\n\n**[](section-events)**\n* Introduction\n* Assessment of Thermal Trace\n* Example: Southeastern Europe June 2024 Heatwave\n* Example: Northwestern Europe January 1963 Cold snap\n\n**[](section-trends)**\n* Introduction\n* Assessment of Thermal Trace\n* Examples", "text_with_prefix": "EQC Quality Assessment: \"Visualisation of heat and cold stress with Thermal Trace\"\nDataset: thermaltrace [CDS]\nAspect: extreme-events_q01 | Category: Applications\nSection: Visualisation of heat and cold stress with Thermal Trace > Methodology\n---\n=\"background-color:#993333; color:white;\">\n Extreme heat stress\n Higher than 46\n \n \n\n
\n\nOutputs from Thermal Trace can be exported in four ways:\n1. Screenshot of the map view (Figure {numref}`{number} `)\n2. Time series exports (Figure {numref}`{number} `)\n3. Time series data download\n4. Linking directly to a specific view in the application by copying the current URL, e.g. [daily peak heat stress in Reading, UK, on 20 August 2025](https://thermaltrace.climate.copernicus.eu/?agg=daily&anomaly=false&date=2025-08-20&heat=true&lat=51.46&lng=-0.97&variable=utci_daily_max_cat).\n\nIt is not possible to export the map view or download the corresponding data\n(global for one time stamp)\ndirectly from Thermal Trace.\nInstead, this can be done from the [ERA5-HEAT CDS catalogue entry](https://doi.org/10.24381/cds.553b7518).\n\n::::{tab-set}\n:::{tab-item} Map view screenshots\n:sync: example_map\n attachment:application_thermaltrace_example_map.png\n---\nname: application_thermaltrace_example_map\n---\nExample screenshots of the Thermal Trace map view, from top to bottom: \na) Daily peak heat stress on 20 August 2025, centered on Reading, UK (app view). \nb) Number of days with at least strong cold stress in December 2025–February 2026, centered on Reading, UK (app view).\n```\n:::\n:::{tab-item} Time series exports\n:sync: example_timeseries\n attachment:application_thermaltrace_example_timeseries.png\n---\nname: application_thermaltrace_example_timeseries\n---\nExample time series exports from the Thermal Trace, from top to bottom: \na) Hourly temperature and UTCI (feels-like temperature) with heat stress categories on 18–22 August 2025, in Reading, UK (app view). \nb) Yearly number of days with a given heat stress level, and yearly number of tropical nights, in Reading, UK (app view).\n```\n:::\n::::\n\nApplications are quality assessed separately from their underpinning datasets. For information on quality attributes of the underlying datasets, the reader is referred to the quality assessment of the datasets, which can be found in their respective CDS catalogue entries or [on the EQC hub](../intro).\n\nThe analysis and results are organised in the following use cases, which are detailed in the sections below:\n\n**[](section-events)**\n* Introduction\n* Assessment of Thermal Trace\n* Example: Southeastern Europe June 2024 Heatwave\n* Example: Northwestern Europe January 1963 Cold snap\n\n**[](section-trends)**\n* Introduction\n* Assessment of Thermal Trace\n* Examples"} {"chunk_id": "application_thermaltrace_extreme-events_q01__2843aba64752", "report_id": "application_thermaltrace_extreme-events_q01", "dataset_id": "thermaltrace", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extreme-events_q01", "aspect_base": "extreme-events", "category": "Applications", "match_confidence": "unmatched", "section": "Visualisation of heat and cold stress with Thermal Trace > Analysis and results", "title": "Visualisation of heat and cold stress with Thermal Trace", "chunk_index": 5, "token_count": 152, "text_raw": "This quality assessment evaluates the suitability of Thermal Trace for use by broad audiences to visualise heat and cold stress,\nboth long-term trends and specific extreme events (heatwaves and cold snaps).\nThermal Trace was assessed based on the availability of relevant data variables, granularity of the data, spatial and temporal coverage and resolution, and the timeliness of data updates.\nThe assessment is illustrated with practical examples of how these events and trends appear in Thermal Trace in real-world scenarios.\n\n(section-events)=", "text_with_prefix": "EQC Quality Assessment: \"Visualisation of heat and cold stress with Thermal Trace\"\nDataset: thermaltrace [CDS]\nAspect: extreme-events_q01 | Category: Applications\nSection: Visualisation of heat and cold stress with Thermal Trace > Analysis and results\n---\nThis quality assessment evaluates the suitability of Thermal Trace for use by broad audiences to visualise heat and cold stress,\nboth long-term trends and specific extreme events (heatwaves and cold snaps).\nThermal Trace was assessed based on the availability of relevant data variables, granularity of the data, spatial and temporal coverage and resolution, and the timeliness of data updates.\nThe assessment is illustrated with practical examples of how these events and trends appear in Thermal Trace in real-world scenarios.\n\n(section-events)="} {"chunk_id": "application_thermaltrace_extreme-events_q01__ff4b5b7eb04a", "report_id": "application_thermaltrace_extreme-events_q01", "dataset_id": "thermaltrace", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extreme-events_q01", "aspect_base": "extreme-events", "category": "Applications", "match_confidence": "unmatched", "section": "Visualisation of heat and cold stress with Thermal Trace > Analysis and results > 1. Extreme events > Introduction", "title": "Visualisation of heat and cold stress with Thermal Trace", "chunk_index": 6, "token_count": 535, "text_raw": "There is no universal definition of a [heatwave](https://climate.copernicus.eu/heatwaves-brief-introduction),\nwith different organisations and countries having established their own criteria.\nHowever, we can generally define them as a period of time, from a few days up to a few weeks,\nof temperatures significantly higher than the normal range for that location at that time of year\n[[Lhotka+22](https://doi.org/10.1029/2022EA002567), [Stefanon+12](https://doi.org/10.1088/1748-9326/7/1/014023)].\nFor example, the [European State of the Climate 2023](https://climate.copernicus.eu/heatwaves-brief-introduction) used the definition\n\"a period of at least three consecutive days when both the daily surface air temperature minima and maxima are higher than the highest 5% of values for the day in question during the 1991–2020 reference period\".\nAnalogously,\nwe can define cold snaps\n(also known as cold waves, cold spells, and cold air outbreaks)\nas a period of time with temperatures significantly lower than typical.\n\nDue to their significant economic and health impacts, as described above, these extreme events often generate significant news coverage.\nFor example,\nthe February–March 2018 cold snap in the UK and Ireland was widely reported on in the popular press under the nickname \"Beast from the East\",\nincluding real-time reporting of its death toll and economic damages [[Morris+18](https://www.theguardian.com/uk-news/2018/mar/02/death-toll-reaches-10-as-destructive-weather-batters-uk)].\nThe June 2024 Eastern Mediterranean heatwave led to similar real-time reporting of its increasing death toll,\nas well as journalistic investigations into the overall impact and the drivers of this extreme event\n[[Bali+24](https://www.dw.com/en/fatal-heat-wave-sweeps-greece-claims-more-lives/a-69485803), [World Weather Attribution+24](https://www.worldweatherattribution.org/deadly-mediterranean-heatwave-would-not-have-occurred-without-human-induced-climate-change/)].", "text_with_prefix": "EQC Quality Assessment: \"Visualisation of heat and cold stress with Thermal Trace\"\nDataset: thermaltrace [CDS]\nAspect: extreme-events_q01 | Category: Applications\nSection: Visualisation of heat and cold stress with Thermal Trace > Analysis and results > 1. Extreme events > Introduction\n---\nThere is no universal definition of a [heatwave](https://climate.copernicus.eu/heatwaves-brief-introduction),\nwith different organisations and countries having established their own criteria.\nHowever, we can generally define them as a period of time, from a few days up to a few weeks,\nof temperatures significantly higher than the normal range for that location at that time of year\n[[Lhotka+22](https://doi.org/10.1029/2022EA002567), [Stefanon+12](https://doi.org/10.1088/1748-9326/7/1/014023)].\nFor example, the [European State of the Climate 2023](https://climate.copernicus.eu/heatwaves-brief-introduction) used the definition\n\"a period of at least three consecutive days when both the daily surface air temperature minima and maxima are higher than the highest 5% of values for the day in question during the 1991–2020 reference period\".\nAnalogously,\nwe can define cold snaps\n(also known as cold waves, cold spells, and cold air outbreaks)\nas a period of time with temperatures significantly lower than typical.\n\nDue to their significant economic and health impacts, as described above, these extreme events often generate significant news coverage.\nFor example,\nthe February–March 2018 cold snap in the UK and Ireland was widely reported on in the popular press under the nickname \"Beast from the East\",\nincluding real-time reporting of its death toll and economic damages [[Morris+18](https://www.theguardian.com/uk-news/2018/mar/02/death-toll-reaches-10-as-destructive-weather-batters-uk)].\nThe June 2024 Eastern Mediterranean heatwave led to similar real-time reporting of its increasing death toll,\nas well as journalistic investigations into the overall impact and the drivers of this extreme event\n[[Bali+24](https://www.dw.com/en/fatal-heat-wave-sweeps-greece-claims-more-lives/a-69485803), [World Weather Attribution+24](https://www.worldweatherattribution.org/deadly-mediterranean-heatwave-would-not-have-occurred-without-human-induced-climate-change/)]."} {"chunk_id": "application_thermaltrace_extreme-events_q01__aa1492cee85c", "report_id": "application_thermaltrace_extreme-events_q01", "dataset_id": "thermaltrace", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extreme-events_q01", "aspect_base": "extreme-events", "category": "Applications", "match_confidence": "unmatched", "section": "Visualisation of heat and cold stress with Thermal Trace > Analysis and results > 1. Extreme events > Assessment of Thermal Trace", "title": "Visualisation of heat and cold stress with Thermal Trace", "chunk_index": 7, "token_count": 871, "text_raw": "Thermal Trace is determined to be suitable for visualising the impacts of heatwaves and cold snaps, depending on the date and location of the event, for the following reasons:\n* **Variables:** Thermal Trace displays multiple useful variables for visualising and characterising extreme heat and cold events, most importantly hourly (time series) or daily (map) UTCI and the derived heat/cold stress categories, as well as the number of days with _x_ stress category and tropical nights per month or season. Using the \"display extra variables\" option also provides stress indicators based on the opposite UTCI limit (i.e. minimum UTCI for heat, maximum for cold), which is useful for determining the amount of relief available, e.g. during a heatwave, are the nights cold enough for the body to rest? Note that Thermal Trace does not provide physical surface air temperatures, but users can view these in the [Climate Pulse application](https://apps.climate.copernicus.eu/overview?app=climate-pulse), for which there is [an equivalent quality assessment](./application_climate-pulse_extreme-events_q01).\n* **Spatial Coverage and Resolution:** Thermal Trace provides global coverage (coverage over the oceans can be toggled in the options) at a resolution (0.25° × 0.25°) suitable for showing regional, national, and multi-country events. Time series are available at the individual grid cell level. The spatial grid is very clearly indicated within the application, e.g. when a specific location is selected. The spatial resolution may be a limiting factor in locations with varied terrain, such as coastlines and cities (ERA5 does not account for the [urban heat island](https://stories.ecmwf.int/urban-heat-islands-and-heat-mortality/index.html) effect).\n* **Temporal Coverage and Resolution:** Data are available on an hourly to yearly basis, making it possible to characterise temperatures over the course of a heatwave or cold snap including daily variations (e.g. tropical nights), as well as comparisons to the rest of the year and the overall climatology of a site. Coverage extends back to 1940, providing a long record for comparison and enabling visualisation of historical events such as the 1963 European cold snap. Note that the accuracy and uncertainty of ERA5, the data underpinning this application, vary over time [[Soci+24](https://doi.org/10.1002/qj.4803), [Hersbach+20](https://doi.org/10.1002/qj.3803)].\n* **Timeliness:** Data become available in Thermal Trace on a delay of a few days (typically ~5), based on the update schedule of the ERA5-HEAT dataset. This enables monitoring of an ongoing event in near real-time, but not true real-time. For real-time and near-future UTCI, users may be interested in [ECMWF's experimental UTCI forecast](https://charts.ecmwf.int/products/medium-thermofeel).\n* **Uncertainty:** Thermal Trace does not show uncertainty estimates, because these are not available for the underpinning dataset (ERA5-HEAT). A general estimate of uncertainty in the data cannot be provided here, but users can refer to the relevant [documentation](https://confluence.ecmwf.int/x/ZSOgBg) and scientific literature for ERA5 [[Soci+24](https://doi.org/10.1002/qj.4803), [Simmons+21](https://doi.org/10.21957/ly5vbtbfd), [Hersbach+20](https://doi.org/10.1002/qj.3803)] to determine the confidence in specific times and locations. Overall, the data are most trustworthy after 1979 (the satellite era), and in locations with more observations (Europe and United States) before then.", "text_with_prefix": "EQC Quality Assessment: \"Visualisation of heat and cold stress with Thermal Trace\"\nDataset: thermaltrace [CDS]\nAspect: extreme-events_q01 | Category: Applications\nSection: Visualisation of heat and cold stress with Thermal Trace > Analysis and results > 1. Extreme events > Assessment of Thermal Trace\n---\nThermal Trace is determined to be suitable for visualising the impacts of heatwaves and cold snaps, depending on the date and location of the event, for the following reasons:\n* **Variables:** Thermal Trace displays multiple useful variables for visualising and characterising extreme heat and cold events, most importantly hourly (time series) or daily (map) UTCI and the derived heat/cold stress categories, as well as the number of days with _x_ stress category and tropical nights per month or season. Using the \"display extra variables\" option also provides stress indicators based on the opposite UTCI limit (i.e. minimum UTCI for heat, maximum for cold), which is useful for determining the amount of relief available, e.g. during a heatwave, are the nights cold enough for the body to rest? Note that Thermal Trace does not provide physical surface air temperatures, but users can view these in the [Climate Pulse application](https://apps.climate.copernicus.eu/overview?app=climate-pulse), for which there is [an equivalent quality assessment](./application_climate-pulse_extreme-events_q01).\n* **Spatial Coverage and Resolution:** Thermal Trace provides global coverage (coverage over the oceans can be toggled in the options) at a resolution (0.25° × 0.25°) suitable for showing regional, national, and multi-country events. Time series are available at the individual grid cell level. The spatial grid is very clearly indicated within the application, e.g. when a specific location is selected. The spatial resolution may be a limiting factor in locations with varied terrain, such as coastlines and cities (ERA5 does not account for the [urban heat island](https://stories.ecmwf.int/urban-heat-islands-and-heat-mortality/index.html) effect).\n* **Temporal Coverage and Resolution:** Data are available on an hourly to yearly basis, making it possible to characterise temperatures over the course of a heatwave or cold snap including daily variations (e.g. tropical nights), as well as comparisons to the rest of the year and the overall climatology of a site. Coverage extends back to 1940, providing a long record for comparison and enabling visualisation of historical events such as the 1963 European cold snap. Note that the accuracy and uncertainty of ERA5, the data underpinning this application, vary over time [[Soci+24](https://doi.org/10.1002/qj.4803), [Hersbach+20](https://doi.org/10.1002/qj.3803)].\n* **Timeliness:** Data become available in Thermal Trace on a delay of a few days (typically ~5), based on the update schedule of the ERA5-HEAT dataset. This enables monitoring of an ongoing event in near real-time, but not true real-time. For real-time and near-future UTCI, users may be interested in [ECMWF's experimental UTCI forecast](https://charts.ecmwf.int/products/medium-thermofeel).\n* **Uncertainty:** Thermal Trace does not show uncertainty estimates, because these are not available for the underpinning dataset (ERA5-HEAT). A general estimate of uncertainty in the data cannot be provided here, but users can refer to the relevant [documentation](https://confluence.ecmwf.int/x/ZSOgBg) and scientific literature for ERA5 [[Soci+24](https://doi.org/10.1002/qj.4803), [Simmons+21](https://doi.org/10.21957/ly5vbtbfd), [Hersbach+20](https://doi.org/10.1002/qj.3803)] to determine the confidence in specific times and locations. Overall, the data are most trustworthy after 1979 (the satellite era), and in locations with more observations (Europe and United States) before then."} {"chunk_id": "application_thermaltrace_extreme-events_q01__44d07c73c6e9", "report_id": "application_thermaltrace_extreme-events_q01", "dataset_id": "thermaltrace", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extreme-events_q01", "aspect_base": "extreme-events", "category": "Applications", "match_confidence": "unmatched", "section": "Visualisation of heat and cold stress with Thermal Trace > Analysis and results > 1. Extreme events > Example: Southeastern Europe June 2024 Heatwave", "title": "Visualisation of heat and cold stress with Thermal Trace", "chunk_index": 8, "token_count": 628, "text_raw": "Between 3 and 27 June 2024,\nseveral countries in Southeastern Europe and part of the Middle East were affected by a heatwave,\nwith the most extreme temperatures occurring between 11 and 14 June [[Dafis+24](https://www.climameter.org/20240611-14-june-eastern-mediterranean-heatwave)].\n\nIn Thermal Trace,\nthis heatwave is clearly visible in the daily maximum UTCI anomaly,\nwith UTCI values over 10 °C warmer than usual,\nand peak heat stress\n(Figure {numref}`{number} `).\nThese daily figures also show a clear progression from west to east over time,\nparticularly in North Africa.\nOver the whole month of June 2024,\nmost of the region experienced strong or very strong (Southeastern Europe and Türkiye) or extreme (North Africa and Middle East) heat stress at least once\n(Figure {numref}`{number} `).\nThe number of days with at least strong heat stress (max UTCI over 32 °C) and tropical nights (min UTCI over 20 °C) in June 2024 was\nmuch (>15 days) higher than normal.\nTropical nights are particularly impactful to human health as they prevent the body from recovering after a hot day [[Royé+21](https://doi.org/10.1097/EDE.0000000000001359)].\n\nIn addition to the map view,\nthe time series exports can be used to monitor the progression of the heatwave in a particular location or a custom area.\nFor example,\nthe exported time series in Figure {numref}`{number} `\nclearly display the daily cycle of temperature with 12 hours of strong to very strong heat stress in Athens, Greece during the peak of the heatwave.\nThe seasonal timeseries in Figure {numref}`{number} `\nshow that June 2024 was exceptionally warm in this region compared to the typical climatology,\nparticularly in Cyprus where the rest of the summer season followed the normal patterns more closely.\nWhile humans can adapt physically and behaviourally to a gradual seasonal rise in temperature,\nan intense heatwave like in June 2024 can increase heat stress levels much faster than usual\n(e.g. very strong heat stress in early June when one would expect only moderate),\noverwhelming this ability and posing serious health risks [[Di Napoli+21b](https://doi.org/10.1007/978-3-030-76716-7_10)].", "text_with_prefix": "EQC Quality Assessment: \"Visualisation of heat and cold stress with Thermal Trace\"\nDataset: thermaltrace [CDS]\nAspect: extreme-events_q01 | Category: Applications\nSection: Visualisation of heat and cold stress with Thermal Trace > Analysis and results > 1. Extreme events > Example: Southeastern Europe June 2024 Heatwave\n---\nBetween 3 and 27 June 2024,\nseveral countries in Southeastern Europe and part of the Middle East were affected by a heatwave,\nwith the most extreme temperatures occurring between 11 and 14 June [[Dafis+24](https://www.climameter.org/20240611-14-june-eastern-mediterranean-heatwave)].\n\nIn Thermal Trace,\nthis heatwave is clearly visible in the daily maximum UTCI anomaly,\nwith UTCI values over 10 °C warmer than usual,\nand peak heat stress\n(Figure {numref}`{number} `).\nThese daily figures also show a clear progression from west to east over time,\nparticularly in North Africa.\nOver the whole month of June 2024,\nmost of the region experienced strong or very strong (Southeastern Europe and Türkiye) or extreme (North Africa and Middle East) heat stress at least once\n(Figure {numref}`{number} `).\nThe number of days with at least strong heat stress (max UTCI over 32 °C) and tropical nights (min UTCI over 20 °C) in June 2024 was\nmuch (>15 days) higher than normal.\nTropical nights are particularly impactful to human health as they prevent the body from recovering after a hot day [[Royé+21](https://doi.org/10.1097/EDE.0000000000001359)].\n\nIn addition to the map view,\nthe time series exports can be used to monitor the progression of the heatwave in a particular location or a custom area.\nFor example,\nthe exported time series in Figure {numref}`{number} `\nclearly display the daily cycle of temperature with 12 hours of strong to very strong heat stress in Athens, Greece during the peak of the heatwave.\nThe seasonal timeseries in Figure {numref}`{number} `\nshow that June 2024 was exceptionally warm in this region compared to the typical climatology,\nparticularly in Cyprus where the rest of the summer season followed the normal patterns more closely.\nWhile humans can adapt physically and behaviourally to a gradual seasonal rise in temperature,\nan intense heatwave like in June 2024 can increase heat stress levels much faster than usual\n(e.g. very strong heat stress in early June when one would expect only moderate),\noverwhelming this ability and posing serious health risks [[Di Napoli+21b](https://doi.org/10.1007/978-3-030-76716-7_10)]."} {"chunk_id": "application_thermaltrace_extreme-events_q01__a7f378e745ba", "report_id": "application_thermaltrace_extreme-events_q01", "dataset_id": "thermaltrace", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extreme-events_q01", "aspect_base": "extreme-events", "category": "Applications", "match_confidence": "unmatched", "section": "Visualisation of heat and cold stress with Thermal Trace > Analysis and results > 1. Extreme events > Example: Southeastern Europe June 2024 Heatwave", "title": "Visualisation of heat and cold stress with Thermal Trace", "chunk_index": 9, "token_count": 1100, "text_raw": "+21b](https://doi.org/10.1007/978-3-030-76716-7_10)].\n\n::::{tab-set}\n:::{tab-item} Daily maps\n:sync: heat_dailymaps\n attachment:application_thermaltrace_heatwave_maps_over_time.png\n---\nname: application_thermaltrace_heatwave_maps_over_time\n---\nScreenshots of Thermal Trace showing: \na) Anomaly in daily maximum UTCI (top; app view). \nb) Daily peak heat stress category (bottom; app view). \nover the 11 to 14 (left to right) June 2024 peak heatwave period. \nThis figure is comparable to Figure {numref}`{number} ` in the quality assessment for Climate Pulse.\n```\n:::\n:::{tab-item} Monthly maps\n:sync: heat_monthlymaps\n attachment:application_thermaltrace_heatwave_maps_june_summary.png\n---\nname: application_thermaltrace_heatwave_maps_june_summary\n---\nScreenshots of Thermal Trace showing: \na) Monthly peak heat stress category (top left; app view). \nb) Anomaly in number of tropical nights (top right; app view). \nc) Number of days with at least strong heat stress (bottom left; app view). \nd) Anomaly in days with at least strong heat stress (bottom right; app view). \nover all of June 2024.\n```\n:::\n:::{tab-item} Time series\n:sync: heat_timeseries\n attachment:application_thermaltrace_heatwave_timeseries.png\n---\nname: application_thermaltrace_heatwave_timeseries\n---\nTime series, from top to bottom: \na) Hourly temperature (dashed) and UTCI (solid) in Athens, Greece during the peak of the 2024 heatwave (app view). \nb) Daily minimum/maximum (blue/red) UTCI over the 2024 summer season (June–August; solid) and average climatology (1991–2020; transparent)\nin Athens, Greece (app view). \nc) Same as b), in Nicosia, Cyprus (app view).\n```\n:::\n::::", "text_with_prefix": "EQC Quality Assessment: \"Visualisation of heat and cold stress with Thermal Trace\"\nDataset: thermaltrace [CDS]\nAspect: extreme-events_q01 | Category: Applications\nSection: Visualisation of heat and cold stress with Thermal Trace > Analysis and results > 1. Extreme events > Example: Southeastern Europe June 2024 Heatwave\n---\n+21b](https://doi.org/10.1007/978-3-030-76716-7_10)].\n\n::::{tab-set}\n:::{tab-item} Daily maps\n:sync: heat_dailymaps\n attachment:application_thermaltrace_heatwave_maps_over_time.png\n---\nname: application_thermaltrace_heatwave_maps_over_time\n---\nScreenshots of Thermal Trace showing: \na) Anomaly in daily maximum UTCI (top; app view). \nb) Daily peak heat stress category (bottom; app view). \nover the 11 to 14 (left to right) June 2024 peak heatwave period. \nThis figure is comparable to Figure {numref}`{number} ` in the quality assessment for Climate Pulse.\n```\n:::\n:::{tab-item} Monthly maps\n:sync: heat_monthlymaps\n attachment:application_thermaltrace_heatwave_maps_june_summary.png\n---\nname: application_thermaltrace_heatwave_maps_june_summary\n---\nScreenshots of Thermal Trace showing: \na) Monthly peak heat stress category (top left; app view). \nb) Anomaly in number of tropical nights (top right; app view). \nc) Number of days with at least strong heat stress (bottom left; app view). \nd) Anomaly in days with at least strong heat stress (bottom right; app view). \nover all of June 2024.\n```\n:::\n:::{tab-item} Time series\n:sync: heat_timeseries\n attachment:application_thermaltrace_heatwave_timeseries.png\n---\nname: application_thermaltrace_heatwave_timeseries\n---\nTime series, from top to bottom: \na) Hourly temperature (dashed) and UTCI (solid) in Athens, Greece during the peak of the 2024 heatwave (app view). \nb) Daily minimum/maximum (blue/red) UTCI over the 2024 summer season (June–August; solid) and average climatology (1991–2020; transparent)\nin Athens, Greece (app view). \nc) Same as b), in Nicosia, Cyprus (app view).\n```\n:::\n::::"} {"chunk_id": "application_thermaltrace_extreme-events_q01__5ce96eeb151d", "report_id": "application_thermaltrace_extreme-events_q01", "dataset_id": "thermaltrace", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extreme-events_q01", "aspect_base": "extreme-events", "category": "Applications", "match_confidence": "unmatched", "section": "Visualisation of heat and cold stress with Thermal Trace > Analysis and results > 1. Extreme events > Example: Northwestern Europe January 1963 Cold snap", "title": "Visualisation of heat and cold stress with Thermal Trace", "chunk_index": 10, "token_count": 816, "text_raw": "The winter of 1963 was one of the coldest in European history,\nsetting records in\nGermany (6.3 °C colder than the 1981–2010 average),\nthe UK (coldest winter since 1740 in central England),\nand much of the rest of Europe\n[[Sippel+24](https://doi.org/10.5194/wcd-5-943-2024), [Twardosz+16](https://doi.org/10.1515/acgeo-2016-0083), [Prior+11](https://doi.org/10.1002/wea.735)].\nIn the Netherlands, the national meteorological office KNMI defines a cold snap as a sequence of at least five days with a maximum temperature below 0.0 °C, of which at least three have a minimum temperature colder than –10.0 °C.\nFour such events were recorded in January 1963 alone [[KNMI](https://www.knmi.nl/nederland-nu/klimatologie/lijsten/koudegolven)].\nThe 1963 edition of the _Elfstedentocht_,\na nearly 200 km speed skating competition held irregularly in the Dutch province of Fryslân,\nhas been nicknamed \"The Hell of '63\" due to the extremely harsh conditions faced by competitors.\nOnly 127 out of 9 862 participants finished the race,\nwith many dropping out due to frozen eyes and frostbite before the race was cancelled for safety concerns [[De Friesche Elf Steden](https://elfstedentocht.frl/en/tour/the-tour-of-1963/), [Wielinga+24](https://www.groningentoen.nl/elfstedentocht/medische-gevolgen-elfstedentocht-1963/)].\n\nIn Thermal Trace,\nthe severe cold faced by Elfstedentocht participants on 18 January 1963 is clearly visible,\nwith strong to extreme cold stress across much of northwestern Europe\nand\nUTCI values around 20 °C colder than usual in the entire Netherlands\n(Figure {numref}`{number} `).\nOver the 1962–63 winter season,\nmost of this part of Europe experienced strong or extreme cold stress at least once,\nwith seasonal minimum UTCI values down to –45 °C,\nwhich is 10 to 15 °C colder than the typical seasonal minimum,\nand ≥70 days with strong cold stress\n(Figure {numref}`{number} `).\n\nThe hourly time series illustrates the difference between UTCI and physical temperature very well (Figure {numref}`{number} `).\nThe air temperature at 2 m increased from 17 to 18 January\nbut the UTCI fell sharply,\ncorresponding to known conditions (severe wind chill and lack of sunlight) that day [[De Friesche Elf Steden](https://elfstedentocht.frl/en/tour/the-tour-of-1963/)].\nSince UTCI is calculated for an average human performing standard tasks,\nnot someone speed skating 200 km,\nthere may be some difference between UTCI and the participants' actual experiences.\nFor example, exercise creates more body heat and thus reduces cold stress, but high speed increases wind chill which worsens cold stress.\nLastly, the seasonal time series in the same figure clearly shows the cold snap of 16–22 January as well as an earlier one around the turn of the year.", "text_with_prefix": "EQC Quality Assessment: \"Visualisation of heat and cold stress with Thermal Trace\"\nDataset: thermaltrace [CDS]\nAspect: extreme-events_q01 | Category: Applications\nSection: Visualisation of heat and cold stress with Thermal Trace > Analysis and results > 1. Extreme events > Example: Northwestern Europe January 1963 Cold snap\n---\nThe winter of 1963 was one of the coldest in European history,\nsetting records in\nGermany (6.3 °C colder than the 1981–2010 average),\nthe UK (coldest winter since 1740 in central England),\nand much of the rest of Europe\n[[Sippel+24](https://doi.org/10.5194/wcd-5-943-2024), [Twardosz+16](https://doi.org/10.1515/acgeo-2016-0083), [Prior+11](https://doi.org/10.1002/wea.735)].\nIn the Netherlands, the national meteorological office KNMI defines a cold snap as a sequence of at least five days with a maximum temperature below 0.0 °C, of which at least three have a minimum temperature colder than –10.0 °C.\nFour such events were recorded in January 1963 alone [[KNMI](https://www.knmi.nl/nederland-nu/klimatologie/lijsten/koudegolven)].\nThe 1963 edition of the _Elfstedentocht_,\na nearly 200 km speed skating competition held irregularly in the Dutch province of Fryslân,\nhas been nicknamed \"The Hell of '63\" due to the extremely harsh conditions faced by competitors.\nOnly 127 out of 9 862 participants finished the race,\nwith many dropping out due to frozen eyes and frostbite before the race was cancelled for safety concerns [[De Friesche Elf Steden](https://elfstedentocht.frl/en/tour/the-tour-of-1963/), [Wielinga+24](https://www.groningentoen.nl/elfstedentocht/medische-gevolgen-elfstedentocht-1963/)].\n\nIn Thermal Trace,\nthe severe cold faced by Elfstedentocht participants on 18 January 1963 is clearly visible,\nwith strong to extreme cold stress across much of northwestern Europe\nand\nUTCI values around 20 °C colder than usual in the entire Netherlands\n(Figure {numref}`{number} `).\nOver the 1962–63 winter season,\nmost of this part of Europe experienced strong or extreme cold stress at least once,\nwith seasonal minimum UTCI values down to –45 °C,\nwhich is 10 to 15 °C colder than the typical seasonal minimum,\nand ≥70 days with strong cold stress\n(Figure {numref}`{number} `).\n\nThe hourly time series illustrates the difference between UTCI and physical temperature very well (Figure {numref}`{number} `).\nThe air temperature at 2 m increased from 17 to 18 January\nbut the UTCI fell sharply,\ncorresponding to known conditions (severe wind chill and lack of sunlight) that day [[De Friesche Elf Steden](https://elfstedentocht.frl/en/tour/the-tour-of-1963/)].\nSince UTCI is calculated for an average human performing standard tasks,\nnot someone speed skating 200 km,\nthere may be some difference between UTCI and the participants' actual experiences.\nFor example, exercise creates more body heat and thus reduces cold stress, but high speed increases wind chill which worsens cold stress.\nLastly, the seasonal time series in the same figure clearly shows the cold snap of 16–22 January as well as an earlier one around the turn of the year."} {"chunk_id": "application_thermaltrace_extreme-events_q01__52faafc9e6f0", "report_id": "application_thermaltrace_extreme-events_q01", "dataset_id": "thermaltrace", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extreme-events_q01", "aspect_base": "extreme-events", "category": "Applications", "match_confidence": "unmatched", "section": "Visualisation of heat and cold stress with Thermal Trace > Analysis and results > 1. Extreme events > Example: Northwestern Europe January 1963 Cold snap", "title": "Visualisation of heat and cold stress with Thermal Trace", "chunk_index": 11, "token_count": 469, "text_raw": "cold stress.\nLastly, the seasonal time series in the same figure clearly shows the cold snap of 16–22 January as well as an earlier one around the turn of the year.\n\nUncertainty is particularly important to consider\nfor case studies early in the ERA5 data record.\nDue to the limited number of observational data available for assimilation in the early years,\nparticularly the absence of satellite data prior to 1979,\nthe uncertainty in individual ERA5 data can be markedly larger than in later years.\nThe accuracy may also be lower for the same reason.\nBoth accuracy and uncertainty are strongly tied to data availability and hence vary strongly with time and location.\nFor example, ERA5 is trustworthy over much of Europe and the United States as early as the 1940s,\nbut not until several decades later in the Southern Hemisphere [[Soci+24](https://doi.org/10.1002/qj.4803)].\nSince ERA5-HEAT, the dataset underpinning Thermal Trace, combines multiple ERA5 variables to determine UTCI,\ncovariance between these variables may increase or decrease the resulting uncertainties.\nFor a discussion on uncertainty in ERA5 and in UTCI, the reader is referred to\n[[Soci+24](https://doi.org/10.1002/qj.4803), [Di Napoli+21b](https://doi.org/10.1007/978-3-030-76716-7_10), [Hersbach+20](https://doi.org/10.1002/qj.3803), [Schreier+13](https://doi.org/10.1007/s00484-012-0525-y)].\nA general rule of thumb for using Thermal Trace would be to investigate the uncertainty in ERA5 for the preferred location and time,\nand to cross-reference with alternative data sources or descriptions of known events\n(such as heatwaves or cold snaps)\nfor increased confidence.", "text_with_prefix": "EQC Quality Assessment: \"Visualisation of heat and cold stress with Thermal Trace\"\nDataset: thermaltrace [CDS]\nAspect: extreme-events_q01 | Category: Applications\nSection: Visualisation of heat and cold stress with Thermal Trace > Analysis and results > 1. Extreme events > Example: Northwestern Europe January 1963 Cold snap\n---\ncold stress.\nLastly, the seasonal time series in the same figure clearly shows the cold snap of 16–22 January as well as an earlier one around the turn of the year.\n\nUncertainty is particularly important to consider\nfor case studies early in the ERA5 data record.\nDue to the limited number of observational data available for assimilation in the early years,\nparticularly the absence of satellite data prior to 1979,\nthe uncertainty in individual ERA5 data can be markedly larger than in later years.\nThe accuracy may also be lower for the same reason.\nBoth accuracy and uncertainty are strongly tied to data availability and hence vary strongly with time and location.\nFor example, ERA5 is trustworthy over much of Europe and the United States as early as the 1940s,\nbut not until several decades later in the Southern Hemisphere [[Soci+24](https://doi.org/10.1002/qj.4803)].\nSince ERA5-HEAT, the dataset underpinning Thermal Trace, combines multiple ERA5 variables to determine UTCI,\ncovariance between these variables may increase or decrease the resulting uncertainties.\nFor a discussion on uncertainty in ERA5 and in UTCI, the reader is referred to\n[[Soci+24](https://doi.org/10.1002/qj.4803), [Di Napoli+21b](https://doi.org/10.1007/978-3-030-76716-7_10), [Hersbach+20](https://doi.org/10.1002/qj.3803), [Schreier+13](https://doi.org/10.1007/s00484-012-0525-y)].\nA general rule of thumb for using Thermal Trace would be to investigate the uncertainty in ERA5 for the preferred location and time,\nand to cross-reference with alternative data sources or descriptions of known events\n(such as heatwaves or cold snaps)\nfor increased confidence."} {"chunk_id": "application_thermaltrace_extreme-events_q01__1f580c0f3bd5", "report_id": "application_thermaltrace_extreme-events_q01", "dataset_id": "thermaltrace", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extreme-events_q01", "aspect_base": "extreme-events", "category": "Applications", "match_confidence": "unmatched", "section": "Visualisation of heat and cold stress with Thermal Trace > Analysis and results > 1. Extreme events > Example: Northwestern Europe January 1963 Cold snap", "title": "Visualisation of heat and cold stress with Thermal Trace", "chunk_index": 12, "token_count": 1090, "text_raw": "of known events\n(such as heatwaves or cold snaps)\nfor increased confidence.\n\n::::{tab-set}\n:::{tab-item} Daily maps\n:sync: heat_dailymaps\n attachment:application_thermaltrace_cold_maps_daily.png\n---\nname: application_thermaltrace_cold_maps_daily\n---\nScreenshots of Thermal Trace showing: \na) Anomaly in daily minimum UTCI (top; app view). \nb) Daily peak cold stress category (bottom; app view). \nin Northwestern Europe on 18 January 1963, the day of the 1963 Elfstedentocht. The marker indicates Leeuwarden, where the race started and finished. \nNote that this figure uses the LAEA (EPSG:3035) projection and that the colour scale is different from Figure {numref}`{number} `.\n```\n:::\n:::{tab-item} Seasonal maps\n:sync: heat_monthlymaps\n attachment:application_thermaltrace_cold_maps_season.png\n---\nname: application_thermaltrace_cold_maps_season\n---\nScreenshots of Thermal Trace showing: \na) Seasonal peak cold stress (top; app view). \nb) Seasonal minimum UTCI (centre left; app view). \nc) Anomaly in seasonal minimum UTCI (centre right; app view). \nd) Number of strong cold stress days (bottom left; app view). \ne) Anomaly in strong cold stress days (bottom right; app view). \nover the December 1962–February 1963 winter season. \nNote that this figure uses the LAEA (EPSG:3035) projection.\n```\n:::\n:::{tab-item} Time series\n:sync: heat_timeseries\n attachment:application_thermaltrace_cold_maps_timeseries.png\n---\nname: application_thermaltrace_cold_maps_timeseries\n---\nTime series, from top to bottom: \na) Hourly temperature (dashed) and UTCI (solid) around the 18 January 1963 Elfstedentocht (app view). \nb) Daily minimum/maximum (blue/red) UTCI over the 1962–63 winter season (December–February; solid) and average climatology (1991–2020; transparent) ( Analysis and results > 1. Extreme events > Example: Northwestern Europe January 1963 Cold snap\n---\nof known events\n(such as heatwaves or cold snaps)\nfor increased confidence.\n\n::::{tab-set}\n:::{tab-item} Daily maps\n:sync: heat_dailymaps\n attachment:application_thermaltrace_cold_maps_daily.png\n---\nname: application_thermaltrace_cold_maps_daily\n---\nScreenshots of Thermal Trace showing: \na) Anomaly in daily minimum UTCI (top; app view). \nb) Daily peak cold stress category (bottom; app view). \nin Northwestern Europe on 18 January 1963, the day of the 1963 Elfstedentocht. The marker indicates Leeuwarden, where the race started and finished. \nNote that this figure uses the LAEA (EPSG:3035) projection and that the colour scale is different from Figure {numref}`{number} `.\n```\n:::\n:::{tab-item} Seasonal maps\n:sync: heat_monthlymaps\n attachment:application_thermaltrace_cold_maps_season.png\n---\nname: application_thermaltrace_cold_maps_season\n---\nScreenshots of Thermal Trace showing: \na) Seasonal peak cold stress (top; app view). \nb) Seasonal minimum UTCI (centre left; app view). \nc) Anomaly in seasonal minimum UTCI (centre right; app view). \nd) Number of strong cold stress days (bottom left; app view). \ne) Anomaly in strong cold stress days (bottom right; app view). \nover the December 1962–February 1963 winter season. \nNote that this figure uses the LAEA (EPSG:3035) projection.\n```\n:::\n:::{tab-item} Time series\n:sync: heat_timeseries\n attachment:application_thermaltrace_cold_maps_timeseries.png\n---\nname: application_thermaltrace_cold_maps_timeseries\n---\nTime series, from top to bottom: \na) Hourly temperature (dashed) and UTCI (solid) around the 18 January 1963 Elfstedentocht (app view). \nb) Daily minimum/maximum (blue/red) UTCI over the 1962–63 winter season (December–February; solid) and average climatology (1991–2020; transparent) ( Analysis and results > 1. Extreme events > Example: Northwestern Europe January 1963 Cold snap", "title": "Visualisation of heat and cold stress with Thermal Trace", "chunk_index": 13, "token_count": 172, "text_raw": "February; solid) and average climatology (1991–2020; transparent) (app view). \nin Leeuwarden, Netherlands.\n```\n:::\n::::", "text_with_prefix": "EQC Quality Assessment: \"Visualisation of heat and cold stress with Thermal Trace\"\nDataset: thermaltrace [CDS]\nAspect: extreme-events_q01 | Category: Applications\nSection: Visualisation of heat and cold stress with Thermal Trace > Analysis and results > 1. Extreme events > Example: Northwestern Europe January 1963 Cold snap\n---\nFebruary; solid) and average climatology (1991–2020; transparent) (app view). \nin Leeuwarden, Netherlands.\n```\n:::\n::::"} {"chunk_id": "application_thermaltrace_extreme-events_q01__cb6e9dd9159b", "report_id": "application_thermaltrace_extreme-events_q01", "dataset_id": "thermaltrace", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extreme-events_q01", "aspect_base": "extreme-events", "category": "Applications", "match_confidence": "unmatched", "section": "Visualisation of heat and cold stress with Thermal Trace > Analysis and results > 2. Long-term trends > Introduction", "title": "Visualisation of heat and cold stress with Thermal Trace", "chunk_index": 14, "token_count": 874, "text_raw": "Climate change is causing a general\nincrease in temperature across most (but not all) of the world,\nwhich naturally comes with increased heat stress and generally decreased cold stress [[Vargas Zeppetello+22](https://doi.org/10.1038/s43247-022-00524-4), [Van Oldenborgh+19](https://doi.org/10.1088/1748-9326/ab4867)].\nFor example,\nin the Caribbean Netherlands, heat stress is increasingly becoming \"unbearable\",\nand a recent report by health and climate experts urged the government to address this issue\n– for both Caribbean and European parts of the country –\nas a matter of priority for public health and the economy\n[[Gezondheidsraad+26](https://www.healthcouncil.nl/documents/2026/05/21/climate-change-and-health-directions-for-policy)].\nThis general trend towards increased heat stress is compounded by an increase in the frequency and severity extreme events,\nboth heatwaves and (in some locations) cold snaps\n[[Thompson+22](https://doi.org/10.1126/sciadv.abm6860), [Ma+19](https://doi.org/10.1175/JCLI-D-18-0234.1), [Kundzewicz+16](https://doi.org/10.1515/igbp-2016-0005)].\n\nIn popular media,\nthe increase in heat stress has been a particularly common discussion topic relating to climate change.\nExplainer pieces on climate change have become a mainstay in journalism,\nwith many published every summer and during heatwaves\n[[Hassan+26](https://www.theguardian.com/world/2026/may/26/high-temperatures-millions-workers-impacted-by-heat-india-asia), [Overton+25](https://www.mainepublic.org/climate/2025-08-07/heat-danger-maines-most-vulnerable-at-risk-from-rising-summer-temps), [Rowlatt+23](https://www.bbc.co.uk/news/science-environment-66143682)].\nThe same is true for heat health alerts,\nissued and reported on when temperatures (physical or feels-like) hit certain thresholds [[Young+26](https://www.bbc.co.uk/news/articles/cp8ppdk5zp2o)].\nThe press also frequently report on projected outcomes of climate change,\nsuch as migration patterns likely to emerge from increased heat stress [[Hall+26](https://www.smh.com.au/environment/climate-change/how-lethal-humidity-threatens-to-displace-millions-in-our-region-20260525-p600j0.html)].\n\nThermal Trace does not provide future projections of UTCI,\nbut can be used to explore historical trends up to the present.\nMany studies have looked into historical trends for different variables and locations and based on different datasets,\ne.g.\nUTCI across the Arabian Peninsula [[Ullah+24](https://doi.org/10.1038/s41598-024-54766-7)]\nand\ntropical nights in Ukraine [[Klok+23](https://doi.org/10.1088/1755-1315/1126/1/012023)] and Spain [[Correa+24](https://doi.org/10.1002/joc.8510)].\nIn one study, Thermal Trace was used to assess the long-term (1940–2024) trend of UTCI and days with very strong/extreme heat stress in Oman \n[[Hereher+26](https://doi.org/10.3390/su18041800)].\nFor future projections of temperature including daily thresholds, but not UTCI, the user is referred to the [_Copernicus Interactive Climate Atlas_](https://apps.climate.copernicus.eu/overview?app=climate-atlas).", "text_with_prefix": "EQC Quality Assessment: \"Visualisation of heat and cold stress with Thermal Trace\"\nDataset: thermaltrace [CDS]\nAspect: extreme-events_q01 | Category: Applications\nSection: Visualisation of heat and cold stress with Thermal Trace > Analysis and results > 2. Long-term trends > Introduction\n---\nClimate change is causing a general\nincrease in temperature across most (but not all) of the world,\nwhich naturally comes with increased heat stress and generally decreased cold stress [[Vargas Zeppetello+22](https://doi.org/10.1038/s43247-022-00524-4), [Van Oldenborgh+19](https://doi.org/10.1088/1748-9326/ab4867)].\nFor example,\nin the Caribbean Netherlands, heat stress is increasingly becoming \"unbearable\",\nand a recent report by health and climate experts urged the government to address this issue\n– for both Caribbean and European parts of the country –\nas a matter of priority for public health and the economy\n[[Gezondheidsraad+26](https://www.healthcouncil.nl/documents/2026/05/21/climate-change-and-health-directions-for-policy)].\nThis general trend towards increased heat stress is compounded by an increase in the frequency and severity extreme events,\nboth heatwaves and (in some locations) cold snaps\n[[Thompson+22](https://doi.org/10.1126/sciadv.abm6860), [Ma+19](https://doi.org/10.1175/JCLI-D-18-0234.1), [Kundzewicz+16](https://doi.org/10.1515/igbp-2016-0005)].\n\nIn popular media,\nthe increase in heat stress has been a particularly common discussion topic relating to climate change.\nExplainer pieces on climate change have become a mainstay in journalism,\nwith many published every summer and during heatwaves\n[[Hassan+26](https://www.theguardian.com/world/2026/may/26/high-temperatures-millions-workers-impacted-by-heat-india-asia), [Overton+25](https://www.mainepublic.org/climate/2025-08-07/heat-danger-maines-most-vulnerable-at-risk-from-rising-summer-temps), [Rowlatt+23](https://www.bbc.co.uk/news/science-environment-66143682)].\nThe same is true for heat health alerts,\nissued and reported on when temperatures (physical or feels-like) hit certain thresholds [[Young+26](https://www.bbc.co.uk/news/articles/cp8ppdk5zp2o)].\nThe press also frequently report on projected outcomes of climate change,\nsuch as migration patterns likely to emerge from increased heat stress [[Hall+26](https://www.smh.com.au/environment/climate-change/how-lethal-humidity-threatens-to-displace-millions-in-our-region-20260525-p600j0.html)].\n\nThermal Trace does not provide future projections of UTCI,\nbut can be used to explore historical trends up to the present.\nMany studies have looked into historical trends for different variables and locations and based on different datasets,\ne.g.\nUTCI across the Arabian Peninsula [[Ullah+24](https://doi.org/10.1038/s41598-024-54766-7)]\nand\ntropical nights in Ukraine [[Klok+23](https://doi.org/10.1088/1755-1315/1126/1/012023)] and Spain [[Correa+24](https://doi.org/10.1002/joc.8510)].\nIn one study, Thermal Trace was used to assess the long-term (1940–2024) trend of UTCI and days with very strong/extreme heat stress in Oman \n[[Hereher+26](https://doi.org/10.3390/su18041800)].\nFor future projections of temperature including daily thresholds, but not UTCI, the user is referred to the [_Copernicus Interactive Climate Atlas_](https://apps.climate.copernicus.eu/overview?app=climate-atlas)."} {"chunk_id": "application_thermaltrace_extreme-events_q01__fc70d2901ab8", "report_id": "application_thermaltrace_extreme-events_q01", "dataset_id": "thermaltrace", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extreme-events_q01", "aspect_base": "extreme-events", "category": "Applications", "match_confidence": "unmatched", "section": "Visualisation of heat and cold stress with Thermal Trace > Analysis and results > 2. Long-term trends > Assessment of Thermal Trace", "title": "Visualisation of heat and cold stress with Thermal Trace", "chunk_index": 15, "token_count": 836, "text_raw": "Thermal Trace is determined to be suitable for visualising long-term trends in heat and cold stress for the following reasons:\n* **Variables:** Thermal Trace displays multiple variables that address common questions in the scientific and popular literature, as well as in policymaking, particularly the number of heat/cold stress days per year and the number of tropical nights per year. For similar information based on physical (rather than feels-like) temperature, users may be interested in the [ERA Explorer](https://apps.climate.copernicus.eu/overview?app=era-explorer) application.\n* **Consistency:** Temporal consistency is key for (long-term) trend analysis, as one needs to be confident that apparent trends are real, not artefacts of the data production. As such, this was a key goal in the production of ERA5 and its extension back to 1940, but the ability to ensure consistency is inherently limited by changes in availability of observations over time [[Soci+24](https://doi.org/10.1002/qj.4803)]. There are some known biases and patterns in ERA5, such as a cold bias over land before 1967 [[Simmons+21](https://doi.org/10.21957/ly5vbtbfd)], differences between the pre- and post-satellite (1979) eras [[Liu+24](https://doi.org/10.1016/j.ecolind.2024.112481), [Lussana+24](https://doi.org/10.1002/asl.1239)], and spatial differences in temporal consistency [[Wang+25](https://doi.org/10.3390/rs17071317)]. Keeping these caveats in mind, ERA5 and its derivative ERA5-HEAT can be reasonably used for long-term trend analysis, and hence Thermal Trace is suitable for visualisation of these trends.\n* **Spatial Coverage and Resolution:** Thermal Trace provides global coverage at a resolution (0.25° × 0.25°) suitable for investigating specific locations on the scale of counties. The ability to draw custom areas also enables larger-scale analysis, e.g. for a small country or wider region. The spatial resolution may be a limiting factor in locations with varied terrain, such as coastlines and cities (ERA5 does not account for the [urban heat island](https://stories.ecmwf.int/urban-heat-islands-and-heat-mortality/index.html) effect). For example, the grid cell containing Bonaire, Netherlands (see example below) is mixed land-sea and its data therefore may not fully represent conditions on the island itself.\n* **Temporal Coverage and Resolution:** Time series are available at hourly, daily, and yearly (summary) resolution, allowing for trend analysis at different temporal scales. For long-term trend analysis, the yearly summary statistics are particularly useful.\n* **Uncertainty:** Thermal Trace does not show uncertainty estimates, because these are not available for the underpinning dataset (ERA5-HEAT). A general estimate of uncertainty in the data cannot be provided here, but users can refer to the relevant [documentation](https://confluence.ecmwf.int/x/ZSOgBg) and scientific literature for ERA5 [[Soci+24](https://doi.org/10.1002/qj.4803), [Simmons+21](https://doi.org/10.21957/ly5vbtbfd), [Hersbach+20](https://doi.org/10.1002/qj.3803)] to determine the confidence in specific times and locations. Overall, the data are most trustworthy after 1979 (the satellite era), and in locations with more observations (Europe and United States) before then.", "text_with_prefix": "EQC Quality Assessment: \"Visualisation of heat and cold stress with Thermal Trace\"\nDataset: thermaltrace [CDS]\nAspect: extreme-events_q01 | Category: Applications\nSection: Visualisation of heat and cold stress with Thermal Trace > Analysis and results > 2. Long-term trends > Assessment of Thermal Trace\n---\nThermal Trace is determined to be suitable for visualising long-term trends in heat and cold stress for the following reasons:\n* **Variables:** Thermal Trace displays multiple variables that address common questions in the scientific and popular literature, as well as in policymaking, particularly the number of heat/cold stress days per year and the number of tropical nights per year. For similar information based on physical (rather than feels-like) temperature, users may be interested in the [ERA Explorer](https://apps.climate.copernicus.eu/overview?app=era-explorer) application.\n* **Consistency:** Temporal consistency is key for (long-term) trend analysis, as one needs to be confident that apparent trends are real, not artefacts of the data production. As such, this was a key goal in the production of ERA5 and its extension back to 1940, but the ability to ensure consistency is inherently limited by changes in availability of observations over time [[Soci+24](https://doi.org/10.1002/qj.4803)]. There are some known biases and patterns in ERA5, such as a cold bias over land before 1967 [[Simmons+21](https://doi.org/10.21957/ly5vbtbfd)], differences between the pre- and post-satellite (1979) eras [[Liu+24](https://doi.org/10.1016/j.ecolind.2024.112481), [Lussana+24](https://doi.org/10.1002/asl.1239)], and spatial differences in temporal consistency [[Wang+25](https://doi.org/10.3390/rs17071317)]. Keeping these caveats in mind, ERA5 and its derivative ERA5-HEAT can be reasonably used for long-term trend analysis, and hence Thermal Trace is suitable for visualisation of these trends.\n* **Spatial Coverage and Resolution:** Thermal Trace provides global coverage at a resolution (0.25° × 0.25°) suitable for investigating specific locations on the scale of counties. The ability to draw custom areas also enables larger-scale analysis, e.g. for a small country or wider region. The spatial resolution may be a limiting factor in locations with varied terrain, such as coastlines and cities (ERA5 does not account for the [urban heat island](https://stories.ecmwf.int/urban-heat-islands-and-heat-mortality/index.html) effect). For example, the grid cell containing Bonaire, Netherlands (see example below) is mixed land-sea and its data therefore may not fully represent conditions on the island itself.\n* **Temporal Coverage and Resolution:** Time series are available at hourly, daily, and yearly (summary) resolution, allowing for trend analysis at different temporal scales. For long-term trend analysis, the yearly summary statistics are particularly useful.\n* **Uncertainty:** Thermal Trace does not show uncertainty estimates, because these are not available for the underpinning dataset (ERA5-HEAT). A general estimate of uncertainty in the data cannot be provided here, but users can refer to the relevant [documentation](https://confluence.ecmwf.int/x/ZSOgBg) and scientific literature for ERA5 [[Soci+24](https://doi.org/10.1002/qj.4803), [Simmons+21](https://doi.org/10.21957/ly5vbtbfd), [Hersbach+20](https://doi.org/10.1002/qj.3803)] to determine the confidence in specific times and locations. Overall, the data are most trustworthy after 1979 (the satellite era), and in locations with more observations (Europe and United States) before then."} {"chunk_id": "application_thermaltrace_extreme-events_q01__05bc02417993", "report_id": "application_thermaltrace_extreme-events_q01", "dataset_id": "thermaltrace", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extreme-events_q01", "aspect_base": "extreme-events", "category": "Applications", "match_confidence": "unmatched", "section": "Visualisation of heat and cold stress with Thermal Trace > Analysis and results > 2. Long-term trends > Examples", "title": "Visualisation of heat and cold stress with Thermal Trace", "chunk_index": 16, "token_count": 599, "text_raw": "Thermal Trace allows the user to generate and export time series for a given grid cell,\ne.g. by searching for a particular town or clicking a location on the map,\nor a custom area drawn directly on the map.\nFor long-term trends,\nthis creates a stacked bar chart of yearly statistics,\ni.e. the number of days with _x_ level of heat or cold stress per year.\nThis chart can be used to visually identify trends.\nThe displayed data can be exported in CSV format for quantitative analysis or custom visualisation.\nHere, we assess only Thermal Trace's suitability for visualising trends within the application;\nthe quality of the underpinning ERA5-HEAT dataset is assessed separately\n(see [](../Indicators/indicator)).\n\nVisually comparing time series exported from Thermal Trace\n(Figure {numref}`{number} `)\nto the literature shows general agreement.\nFor example,\nthe time series show\nan increase in very strong heat stress days in Bonaire, Caribbean Netherlands [[Gezondheidsraad+26](https://www.healthcouncil.nl/documents/2026/05/21/climate-change-and-health-directions-for-policy)]\nand\nan increase in the number of tropical nights,\nas well as a potential increase in the number of heat stress days,\nin Madrid, Spain [[Correa+24](https://doi.org/10.1002/joc.8510)].\nKherson, Ukraine, also appears to show an increase in tropical nights and heat stress days [[Klok+23](https://doi.org/10.1088/1755-1315/1126/1/012023)],\nbut this is more ambiguous in the time series.\nIn Hyderabad, India, the number of tropical nights seems to have increased, but the number of heat stress days appears stable [[Hassan+26](https://www.theguardian.com/world/2026/may/26/high-temperatures-millions-workers-impacted-by-heat-india-asia)].\nSimilarly,\nThermal Trace shows a slight decline in cold stress days in recent years\n(Figure {numref}`{number} `)\nin the Midwest region of the United States [[Van Oldenborgh+19](https://doi.org/10.1088/1748-9326/ab4867)].\nIn each of these cases,\nThermal Trace is a very useful tool for a first glance at the data,\ne.g. in a new location or timeframe,\nprior to a quantitative analysis.", "text_with_prefix": "EQC Quality Assessment: \"Visualisation of heat and cold stress with Thermal Trace\"\nDataset: thermaltrace [CDS]\nAspect: extreme-events_q01 | Category: Applications\nSection: Visualisation of heat and cold stress with Thermal Trace > Analysis and results > 2. Long-term trends > Examples\n---\nThermal Trace allows the user to generate and export time series for a given grid cell,\ne.g. by searching for a particular town or clicking a location on the map,\nor a custom area drawn directly on the map.\nFor long-term trends,\nthis creates a stacked bar chart of yearly statistics,\ni.e. the number of days with _x_ level of heat or cold stress per year.\nThis chart can be used to visually identify trends.\nThe displayed data can be exported in CSV format for quantitative analysis or custom visualisation.\nHere, we assess only Thermal Trace's suitability for visualising trends within the application;\nthe quality of the underpinning ERA5-HEAT dataset is assessed separately\n(see [](../Indicators/indicator)).\n\nVisually comparing time series exported from Thermal Trace\n(Figure {numref}`{number} `)\nto the literature shows general agreement.\nFor example,\nthe time series show\nan increase in very strong heat stress days in Bonaire, Caribbean Netherlands [[Gezondheidsraad+26](https://www.healthcouncil.nl/documents/2026/05/21/climate-change-and-health-directions-for-policy)]\nand\nan increase in the number of tropical nights,\nas well as a potential increase in the number of heat stress days,\nin Madrid, Spain [[Correa+24](https://doi.org/10.1002/joc.8510)].\nKherson, Ukraine, also appears to show an increase in tropical nights and heat stress days [[Klok+23](https://doi.org/10.1088/1755-1315/1126/1/012023)],\nbut this is more ambiguous in the time series.\nIn Hyderabad, India, the number of tropical nights seems to have increased, but the number of heat stress days appears stable [[Hassan+26](https://www.theguardian.com/world/2026/may/26/high-temperatures-millions-workers-impacted-by-heat-india-asia)].\nSimilarly,\nThermal Trace shows a slight decline in cold stress days in recent years\n(Figure {numref}`{number} `)\nin the Midwest region of the United States [[Van Oldenborgh+19](https://doi.org/10.1088/1748-9326/ab4867)].\nIn each of these cases,\nThermal Trace is a very useful tool for a first glance at the data,\ne.g. in a new location or timeframe,\nprior to a quantitative analysis."} {"chunk_id": "application_thermaltrace_extreme-events_q01__ad2518c9aa23", "report_id": "application_thermaltrace_extreme-events_q01", "dataset_id": "thermaltrace", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extreme-events_q01", "aspect_base": "extreme-events", "category": "Applications", "match_confidence": "unmatched", "section": "Visualisation of heat and cold stress with Thermal Trace > Analysis and results > 2. Long-term trends > Examples", "title": "Visualisation of heat and cold stress with Thermal Trace", "chunk_index": 17, "token_count": 694, "text_raw": "useful tool for a first glance at the data,\ne.g. in a new location or timeframe,\nprior to a quantitative analysis.\n\n::::{tab-set}\n:::{tab-item} Heat stress\n:sync: trends_heat\n attachment:application_thermaltrace_trends_heat.png\n---\nname: application_thermaltrace_trends_heat\n---\nTime series exports showing yearly number of days with a given heat stress level, and yearly number of tropical nights, in (from top to bottom): \na) Bonaire, Caribbean Netherlands (app view). Note that this requires the ocean mask to be toggled off. \nb) Hyderabad, India (app view). \nc) Kherson, Ukraine (app view). \nd) Madrid, Spain (app view).\n```\n:::\n:::{tab-item} Cold stress\n:sync: trends_cold\n attachment:application_thermaltrace_trends_cold.png\n---\nname: application_thermaltrace_trends_cold\n---\nTime series exports showing yearly number of days with a given cold stress level in a custom drawn region stretching from Charleston, West Virginia to Cleveland, Ohio (north) and Omaha, Nebraska (east), all in the United States. From top to bottom: \na) Time series. \nb) Region shown on the map, zoomed in. \nc) Region shown on the map, in the wider American context. \nThese figures can be reproduced by navigating to the same location\n(app view)\nand manually re-drawing the region.\n```\n:::\n::::", "text_with_prefix": "EQC Quality Assessment: \"Visualisation of heat and cold stress with Thermal Trace\"\nDataset: thermaltrace [CDS]\nAspect: extreme-events_q01 | Category: Applications\nSection: Visualisation of heat and cold stress with Thermal Trace > Analysis and results > 2. Long-term trends > Examples\n---\nuseful tool for a first glance at the data,\ne.g. in a new location or timeframe,\nprior to a quantitative analysis.\n\n::::{tab-set}\n:::{tab-item} Heat stress\n:sync: trends_heat\n attachment:application_thermaltrace_trends_heat.png\n---\nname: application_thermaltrace_trends_heat\n---\nTime series exports showing yearly number of days with a given heat stress level, and yearly number of tropical nights, in (from top to bottom): \na) Bonaire, Caribbean Netherlands (app view). Note that this requires the ocean mask to be toggled off. \nb) Hyderabad, India (app view). \nc) Kherson, Ukraine (app view). \nd) Madrid, Spain (app view).\n```\n:::\n:::{tab-item} Cold stress\n:sync: trends_cold\n attachment:application_thermaltrace_trends_cold.png\n---\nname: application_thermaltrace_trends_cold\n---\nTime series exports showing yearly number of days with a given cold stress level in a custom drawn region stretching from Charleston, West Virginia to Cleveland, Ohio (north) and Omaha, Nebraska (east), all in the United States. From top to bottom: \na) Time series. \nb) Region shown on the map, zoomed in. \nc) Region shown on the map, in the wider American context. \nThese figures can be reproduced by navigating to the same location\n(app view)\nand manually re-drawing the region.\n```\n:::\n::::"} {"chunk_id": "application_thermaltrace_extreme-events_q01__368337bd4c58", "report_id": "application_thermaltrace_extreme-events_q01", "dataset_id": "thermaltrace", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extreme-events_q01", "aspect_base": "extreme-events", "category": "Applications", "match_confidence": "unmatched", "section": "Visualisation of heat and cold stress with Thermal Trace > ℹ️ If you want to know more > Key resources", "title": "Visualisation of heat and cold stress with Thermal Trace", "chunk_index": 18, "token_count": 981, "text_raw": "More about the ERA5 reanalysis and ERA5-HEAT dataset:\n* [The ERA5 global reanalysis](https://doi.org/10.1002/qj.3803)\n* [The ERA5 global reanalysis from 1940 to 2022](https://doi.org/10.1002/qj.4803)\n* [ERA5-HEAT: A global gridded historical dataset of human thermal comfort indices from climate reanalysis](https://doi.org/10.1002/gdj3.102)\n* [Thermal comfort indices derived from ERA5 reanalysis (ERA5-HEAT)](https://doi.org/10.24381/cds.553b7518) on the CDS\n * [Time series optimised version of the dataset](https://cds.climate.copernicus.eu/datasets/derived-utci-historical-timeseries)\n\nRelated datasets on the CDS:\n* [Climate extreme indices and heat stress indicators derived from CMIP6 global climate projections](https://doi.org/10.24381/cds.776e08bd)\n* [Heat waves and cold spells in Europe derived from climate projections](https://doi.org/10.24381/cds.9e7ca677)\n\nMore about feels-like temperature and UTCI:\n* [Literature Review on UTCI Applications](https://doi.org/10.1007/978-3-030-76716-7_3)\n* [Principles of the New Universal Thermal Climate Index (UTCI) and its Application to Bioclimatic Research in European Scale](https://doi.org/10.2478/mgrsd-2010-0009)\n* [Thermofeel: A python thermal comfort indices library](https://doi.org/10.1016/j.softx.2022.101005)\n* [UTCI—Why another thermal index?](https://doi.org/10.1007/s00484-011-0513-7)\n\nMore about cold snaps:\n* [Could an extremely cold central European winter such as 1963 happen again despite climate change?](https://doi.org/10.5194/wcd-5-943-2024)\n* [The UK winter of 2009/2010 compared with severe winters of the last 100 years](https://doi.org/10.1002/wea.735)\n* [Where Do Cold Air Outbreaks Occur, and How Have They Changed Over Time?](https://doi.org/10.1029/2020GL086983)\n\nMore about heatwaves:\n* [Heatwaves – a brief introduction](https://climate.copernicus.eu/heatwaves-brief-introduction)\n* [Heat Waves: Physical Understanding and Scientific Challenges](https://doi.org/10.1029/2022RG000780)\n* [World Health Organization: \"Heat and health\"](https://www.who.int/news-room/fact-sheets/detail/climate-change-heat-and-health)\n\nMore about long-term trends in heat and cold stress:\n* [Climate change and health: Directions for policy](https://www.healthcouncil.nl/documents/2026/05/21/climate-change-and-health-directions-for-policy)\n* [Cold waves are getting milder in the northern midlatitudes](https://doi.org/10.1088/1748-9326/ab4867)\n* [Extreme Cold Wave over East Asia in January 2016: A Possible Response to the Larger Internal Atmospheric Variability Induced by Arctic Warming](https://doi.org/10.1175/JCLI-D-18-0234.1)\n* [Probabilistic projections of increased heat stress driven by climate change](https://doi.org/10.1038/s43247-022-00524-4)\n\nMore applications for visualising temperature and thermal comfort:\n* [Climate Pulse](https://apps.climate.copernicus.eu/overview?app=climate-pulse)\n * [](./application_climate-pulse_extreme-events_q01)\n* [ECMWF's experimental UTCI forecast](https://charts.ecmwf.int/products/medium-thermofeel)\n* [ERA Explorer](https://apps.climate.copernicus.eu/overview?app=era-explorer)\n* [Weather Replay](https://apps.climate.copernicus.eu/overview?app=weather-replay)", "text_with_prefix": "EQC Quality Assessment: \"Visualisation of heat and cold stress with Thermal Trace\"\nDataset: thermaltrace [CDS]\nAspect: extreme-events_q01 | Category: Applications\nSection: Visualisation of heat and cold stress with Thermal Trace > ℹ️ If you want to know more > Key resources\n---\nMore about the ERA5 reanalysis and ERA5-HEAT dataset:\n* [The ERA5 global reanalysis](https://doi.org/10.1002/qj.3803)\n* [The ERA5 global reanalysis from 1940 to 2022](https://doi.org/10.1002/qj.4803)\n* [ERA5-HEAT: A global gridded historical dataset of human thermal comfort indices from climate reanalysis](https://doi.org/10.1002/gdj3.102)\n* [Thermal comfort indices derived from ERA5 reanalysis (ERA5-HEAT)](https://doi.org/10.24381/cds.553b7518) on the CDS\n * [Time series optimised version of the dataset](https://cds.climate.copernicus.eu/datasets/derived-utci-historical-timeseries)\n\nRelated datasets on the CDS:\n* [Climate extreme indices and heat stress indicators derived from CMIP6 global climate projections](https://doi.org/10.24381/cds.776e08bd)\n* [Heat waves and cold spells in Europe derived from climate projections](https://doi.org/10.24381/cds.9e7ca677)\n\nMore about feels-like temperature and UTCI:\n* [Literature Review on UTCI Applications](https://doi.org/10.1007/978-3-030-76716-7_3)\n* [Principles of the New Universal Thermal Climate Index (UTCI) and its Application to Bioclimatic Research in European Scale](https://doi.org/10.2478/mgrsd-2010-0009)\n* [Thermofeel: A python thermal comfort indices library](https://doi.org/10.1016/j.softx.2022.101005)\n* [UTCI—Why another thermal index?](https://doi.org/10.1007/s00484-011-0513-7)\n\nMore about cold snaps:\n* [Could an extremely cold central European winter such as 1963 happen again despite climate change?](https://doi.org/10.5194/wcd-5-943-2024)\n* [The UK winter of 2009/2010 compared with severe winters of the last 100 years](https://doi.org/10.1002/wea.735)\n* [Where Do Cold Air Outbreaks Occur, and How Have They Changed Over Time?](https://doi.org/10.1029/2020GL086983)\n\nMore about heatwaves:\n* [Heatwaves – a brief introduction](https://climate.copernicus.eu/heatwaves-brief-introduction)\n* [Heat Waves: Physical Understanding and Scientific Challenges](https://doi.org/10.1029/2022RG000780)\n* [World Health Organization: \"Heat and health\"](https://www.who.int/news-room/fact-sheets/detail/climate-change-heat-and-health)\n\nMore about long-term trends in heat and cold stress:\n* [Climate change and health: Directions for policy](https://www.healthcouncil.nl/documents/2026/05/21/climate-change-and-health-directions-for-policy)\n* [Cold waves are getting milder in the northern midlatitudes](https://doi.org/10.1088/1748-9326/ab4867)\n* [Extreme Cold Wave over East Asia in January 2016: A Possible Response to the Larger Internal Atmospheric Variability Induced by Arctic Warming](https://doi.org/10.1175/JCLI-D-18-0234.1)\n* [Probabilistic projections of increased heat stress driven by climate change](https://doi.org/10.1038/s43247-022-00524-4)\n\nMore applications for visualising temperature and thermal comfort:\n* [Climate Pulse](https://apps.climate.copernicus.eu/overview?app=climate-pulse)\n * [](./application_climate-pulse_extreme-events_q01)\n* [ECMWF's experimental UTCI forecast](https://charts.ecmwf.int/products/medium-thermofeel)\n* [ERA Explorer](https://apps.climate.copernicus.eu/overview?app=era-explorer)\n* [Weather Replay](https://apps.climate.copernicus.eu/overview?app=weather-replay)"} {"chunk_id": "application_thermaltrace_extreme-events_q01__8c71a25c5359", "report_id": "application_thermaltrace_extreme-events_q01", "dataset_id": "thermaltrace", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extreme-events_q01", "aspect_base": "extreme-events", "category": "Applications", "match_confidence": "unmatched", "section": "Visualisation of heat and cold stress with Thermal Trace > ℹ️ If you want to know more > References", "title": "Visualisation of heat and cold stress with Thermal Trace", "chunk_index": 19, "token_count": 955, "text_raw": "[[Bell+18](https://doi.org/10.1080/10962247.2017.1401017)] J. E. Bell et al., ‘Changes in extreme events and the potential impacts on human health’, Journal of the Air & Waste Management Association, vol. 68, no. 4, pp. 265–287, Apr. 2018, doi: 10.1080/10962247.2017.1401017.\n\n[[Błażejczyk+10](https://doi.org/10.2478/mgrsd-2010-0009)] K. Błażejczyk et al., ‘Principles of the New Universal Thermal Climate Index (UTCI) and its Application to Bioclimatic Research in European Scale’, Miscellanea Geographica, vol. 14, no. 1, pp. 91–102, Dec. 2010, doi: 10.2478/mgrsd-2010-0009.\n\n[[Correa+24](https://doi.org/10.1002/joc.8510)] J. Correa, P. Dorta, A. López-Díez, and J. Díaz-Pacheco, ‘Analysis of tropical nights in Spain (1970–2023): Minimum temperatures as an indicator of climate change’, International Journal of Climatology, vol. 44, no. 9, pp. 3006–3027, Jun. 2024, doi: 10.1002/joc.8510.\n\n[[De Friesche Elf Steden](https://elfstedentocht.frl/en/tour/the-tour-of-1963/)] De Friesche Elf Steden, ‘Tour of 1963: The Toughest Tour in the History of the Frisian Elfstedentocht’, Koninklijke Vereniging De Friesche Elf Steden. Accessed: May 18, 2026. [Online]. Available: https://elfstedentocht.frl/en/tour/the-tour-of-1963/\n\n[[Di Napoli+19](https://doi.org/10.1175/JAMC-D-18-0246.1)] C. Di Napoli, F. Pappenberger, and H. L. Cloke, ‘Verification of Heat Stress Thresholds for a Health-Based Heat-Wave Definition’, Journal of Applied Meteorology and Climatology, vol. 58, no. 6, pp. 1177–1194, Jun. 2019, doi: 10.1175/JAMC-D-18-0246.1.\n\n[[Di Napoli+21a](https://doi.org/10.1002/gdj3.102)] C. Di Napoli, C. Barnard, C. Prudhomme, H. L. Cloke, and F. Pappenberger, ‘ERA5-HEAT: A global gridded historical dataset of human thermal comfort indices from climate reanalysis’, Geoscience Data Journal, vol. 8, no. 1, pp. 2–10, 2021, doi: 10.1002/gdj3.102.\n\n[[Di Napoli+21b](https://doi.org/10.1007/978-3-030-76716-7_10)] C. Di Napoli et al., ‘The Universal Thermal Climate Index as an Operational Forecasting Tool of Human Biometeorological Conditions in Europe’, in Applications of the Universal Thermal Climate Index UTCI in Biometeorology: Latest Developments and Case Studies, E. L. Krüger, Ed., Cham: Springer International Publishing, 2021, pp. 193–208. doi: 10.1007/978-3-030-76716-7_10.\n\n[[Fiala+12](https://doi.org/10.1007/s00484-011-0424-7)] G. Havenith et al., ‘The UTCI-clothing model’, International Journal of Biometeorology, vol. 56, no. 3, pp. 461–470, 2012, doi: 10.1007/s00484-011-0451-4.", "text_with_prefix": "EQC Quality Assessment: \"Visualisation of heat and cold stress with Thermal Trace\"\nDataset: thermaltrace [CDS]\nAspect: extreme-events_q01 | Category: Applications\nSection: Visualisation of heat and cold stress with Thermal Trace > ℹ️ If you want to know more > References\n---\n[[Bell+18](https://doi.org/10.1080/10962247.2017.1401017)] J. E. Bell et al., ‘Changes in extreme events and the potential impacts on human health’, Journal of the Air & Waste Management Association, vol. 68, no. 4, pp. 265–287, Apr. 2018, doi: 10.1080/10962247.2017.1401017.\n\n[[Błażejczyk+10](https://doi.org/10.2478/mgrsd-2010-0009)] K. Błażejczyk et al., ‘Principles of the New Universal Thermal Climate Index (UTCI) and its Application to Bioclimatic Research in European Scale’, Miscellanea Geographica, vol. 14, no. 1, pp. 91–102, Dec. 2010, doi: 10.2478/mgrsd-2010-0009.\n\n[[Correa+24](https://doi.org/10.1002/joc.8510)] J. Correa, P. Dorta, A. López-Díez, and J. Díaz-Pacheco, ‘Analysis of tropical nights in Spain (1970–2023): Minimum temperatures as an indicator of climate change’, International Journal of Climatology, vol. 44, no. 9, pp. 3006–3027, Jun. 2024, doi: 10.1002/joc.8510.\n\n[[De Friesche Elf Steden](https://elfstedentocht.frl/en/tour/the-tour-of-1963/)] De Friesche Elf Steden, ‘Tour of 1963: The Toughest Tour in the History of the Frisian Elfstedentocht’, Koninklijke Vereniging De Friesche Elf Steden. Accessed: May 18, 2026. [Online]. Available: https://elfstedentocht.frl/en/tour/the-tour-of-1963/\n\n[[Di Napoli+19](https://doi.org/10.1175/JAMC-D-18-0246.1)] C. Di Napoli, F. Pappenberger, and H. L. Cloke, ‘Verification of Heat Stress Thresholds for a Health-Based Heat-Wave Definition’, Journal of Applied Meteorology and Climatology, vol. 58, no. 6, pp. 1177–1194, Jun. 2019, doi: 10.1175/JAMC-D-18-0246.1.\n\n[[Di Napoli+21a](https://doi.org/10.1002/gdj3.102)] C. Di Napoli, C. Barnard, C. Prudhomme, H. L. Cloke, and F. Pappenberger, ‘ERA5-HEAT: A global gridded historical dataset of human thermal comfort indices from climate reanalysis’, Geoscience Data Journal, vol. 8, no. 1, pp. 2–10, 2021, doi: 10.1002/gdj3.102.\n\n[[Di Napoli+21b](https://doi.org/10.1007/978-3-030-76716-7_10)] C. Di Napoli et al., ‘The Universal Thermal Climate Index as an Operational Forecasting Tool of Human Biometeorological Conditions in Europe’, in Applications of the Universal Thermal Climate Index UTCI in Biometeorology: Latest Developments and Case Studies, E. L. Krüger, Ed., Cham: Springer International Publishing, 2021, pp. 193–208. doi: 10.1007/978-3-030-76716-7_10.\n\n[[Fiala+12](https://doi.org/10.1007/s00484-011-0424-7)] G. Havenith et al., ‘The UTCI-clothing model’, International Journal of Biometeorology, vol. 56, no. 3, pp. 461–470, 2012, doi: 10.1007/s00484-011-0451-4."} {"chunk_id": "application_thermaltrace_extreme-events_q01__8d9f1e9fc48c", "report_id": "application_thermaltrace_extreme-events_q01", "dataset_id": "thermaltrace", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extreme-events_q01", "aspect_base": "extreme-events", "category": "Applications", "match_confidence": "unmatched", "section": "Visualisation of heat and cold stress with Thermal Trace > ℹ️ If you want to know more > References", "title": "Visualisation of heat and cold stress with Thermal Trace", "chunk_index": 20, "token_count": 1095, "text_raw": "Biometeorology, vol. 56, no. 3, pp. 461–470, 2012, doi: 10.1007/s00484-011-0451-4.\n\n[[Gezondheidsraad+26](https://www.healthcouncil.nl/documents/2026/05/21/climate-change-and-health-directions-for-policy)] Gezondheidsraad and Wetenschappelijke Klimaatraad, ‘Climate change and health: Directions for policy’, Den Haag, Netherlands, 2026/05, WKR-advies 009, May 2026. [Online]. Available: https://www.healthcouncil.nl/documents/2026/05/21/climate-change-and-health-directions-for-policy\n\n[[Hall+26](https://www.smh.com.au/environment/climate-change/how-lethal-humidity-threatens-to-displace-millions-in-our-region-20260525-p600j0.html)] B. Hall, ‘How lethal humidity threatens to displace millions in our region’, The Sydney Morning Herald, May 27, 2026. Accessed: May 27, 2026. [Online]. Available: https://www.smh.com.au/environment/climate-change/how-lethal-humidity-threatens-to-displace-millions-in-our-region-20260525-p600j0.html\n\n[[Hassan+26](https://www.theguardian.com/world/2026/may/26/high-temperatures-millions-workers-impacted-by-heat-india-asia)] A. Hassan, ‘“My head spins with the heat”: India’s gig workers battle exhaustion amid soaring temperatures’, The Guardian, May 26, 2026. Accessed: May 27, 2026. [Online]. Available: https://www.theguardian.com/world/2026/may/26/high-temperatures-millions-workers-impacted-by-heat-india-asia\n\n[[Havenith+12](https://doi.org/10.1007/s00484-011-0451-4)] D. Fiala, G. Havenith, P. Bröde, B. Kampmann, and G. Jendritzky, ‘UTCI-Fiala multi-node model of human heat transfer and temperature regulation’, International Journal of Biometeorology, vol. 56, no. 3, pp. 429–441, 2012, doi: 10.1007/s00484-011-0424-7.\n\n[[Hereher+26](https://doi.org/10.3390/su18041800)] M. E. Hereher, ‘Assessment of Seasonal Patterns of Apparent Heat Stress in Oman Using ERA5 Climatic Data’, Sustainability, vol. 18, no. 4, p. 1800, Feb. 2026, doi: 10.3390/su18041800.\n\n[[Hersbach+20](https://doi.org/10.1002/qj.3803)] H. Hersbach et al., ‘The ERA5 global reanalysis’, Quarterly Journal of the Royal Meteorological Society, vol. 146, no. 730, pp. 1999–2049, Jul. 2020, doi: 10.1002/qj.3803.\n\n[[Jendritzky+12](https://doi.org/10.1007/s00484-011-0513-7)] G. Jendritzky, R. de Dear, and G. Havenith, ‘UTCI—Why another thermal index?’, International Journal of Biometeorology, vol. 56, no. 3, pp. 421–428, 2012, doi: 10.1007/s00484-011-0513-7.\n\n[[Klok+23](https://doi.org/10.1088/1755-1315/1126/1/012023)] S. Klok, A. Kornus, O. Kornus, O. Danylchenko, and O. Skyba, ‘Tropical nights (1976–2019) as an indicator of climate change in Ukraine’, in IOP Conference Series: Earth and Environmental Science, Riga, Latvia: IOP Publishing, Jan. 2023, p. 012023. doi: 10.1088/1755-1315/1126/1/012023.\n\n[[KNMI](https://www.knmi.nl/nederland-nu/klimatologie/lijsten/koudegolven)] KNMI, ‘Koudegolven’, KNMI – Koninklijk Nederlands Meteorologisch Instituut. Accessed: May 22, 2026. [Online]. Available: https://www.knmi.nl/nederland-nu/klimatologie/lijsten/koudegolven", "text_with_prefix": "EQC Quality Assessment: \"Visualisation of heat and cold stress with Thermal Trace\"\nDataset: thermaltrace [CDS]\nAspect: extreme-events_q01 | Category: Applications\nSection: Visualisation of heat and cold stress with Thermal Trace > ℹ️ If you want to know more > References\n---\nBiometeorology, vol. 56, no. 3, pp. 461–470, 2012, doi: 10.1007/s00484-011-0451-4.\n\n[[Gezondheidsraad+26](https://www.healthcouncil.nl/documents/2026/05/21/climate-change-and-health-directions-for-policy)] Gezondheidsraad and Wetenschappelijke Klimaatraad, ‘Climate change and health: Directions for policy’, Den Haag, Netherlands, 2026/05, WKR-advies 009, May 2026. [Online]. Available: https://www.healthcouncil.nl/documents/2026/05/21/climate-change-and-health-directions-for-policy\n\n[[Hall+26](https://www.smh.com.au/environment/climate-change/how-lethal-humidity-threatens-to-displace-millions-in-our-region-20260525-p600j0.html)] B. Hall, ‘How lethal humidity threatens to displace millions in our region’, The Sydney Morning Herald, May 27, 2026. Accessed: May 27, 2026. [Online]. Available: https://www.smh.com.au/environment/climate-change/how-lethal-humidity-threatens-to-displace-millions-in-our-region-20260525-p600j0.html\n\n[[Hassan+26](https://www.theguardian.com/world/2026/may/26/high-temperatures-millions-workers-impacted-by-heat-india-asia)] A. Hassan, ‘“My head spins with the heat”: India’s gig workers battle exhaustion amid soaring temperatures’, The Guardian, May 26, 2026. Accessed: May 27, 2026. [Online]. Available: https://www.theguardian.com/world/2026/may/26/high-temperatures-millions-workers-impacted-by-heat-india-asia\n\n[[Havenith+12](https://doi.org/10.1007/s00484-011-0451-4)] D. Fiala, G. Havenith, P. Bröde, B. Kampmann, and G. Jendritzky, ‘UTCI-Fiala multi-node model of human heat transfer and temperature regulation’, International Journal of Biometeorology, vol. 56, no. 3, pp. 429–441, 2012, doi: 10.1007/s00484-011-0424-7.\n\n[[Hereher+26](https://doi.org/10.3390/su18041800)] M. E. Hereher, ‘Assessment of Seasonal Patterns of Apparent Heat Stress in Oman Using ERA5 Climatic Data’, Sustainability, vol. 18, no. 4, p. 1800, Feb. 2026, doi: 10.3390/su18041800.\n\n[[Hersbach+20](https://doi.org/10.1002/qj.3803)] H. Hersbach et al., ‘The ERA5 global reanalysis’, Quarterly Journal of the Royal Meteorological Society, vol. 146, no. 730, pp. 1999–2049, Jul. 2020, doi: 10.1002/qj.3803.\n\n[[Jendritzky+12](https://doi.org/10.1007/s00484-011-0513-7)] G. Jendritzky, R. de Dear, and G. Havenith, ‘UTCI—Why another thermal index?’, International Journal of Biometeorology, vol. 56, no. 3, pp. 421–428, 2012, doi: 10.1007/s00484-011-0513-7.\n\n[[Klok+23](https://doi.org/10.1088/1755-1315/1126/1/012023)] S. Klok, A. Kornus, O. Kornus, O. Danylchenko, and O. Skyba, ‘Tropical nights (1976–2019) as an indicator of climate change in Ukraine’, in IOP Conference Series: Earth and Environmental Science, Riga, Latvia: IOP Publishing, Jan. 2023, p. 012023. doi: 10.1088/1755-1315/1126/1/012023.\n\n[[KNMI](https://www.knmi.nl/nederland-nu/klimatologie/lijsten/koudegolven)] KNMI, ‘Koudegolven’, KNMI – Koninklijk Nederlands Meteorologisch Instituut. Accessed: May 22, 2026. [Online]. Available: https://www.knmi.nl/nederland-nu/klimatologie/lijsten/koudegolven"} {"chunk_id": "application_thermaltrace_extreme-events_q01__e7bb9dd3bfb5", "report_id": "application_thermaltrace_extreme-events_q01", "dataset_id": "thermaltrace", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extreme-events_q01", "aspect_base": "extreme-events", "category": "Applications", "match_confidence": "unmatched", "section": "Visualisation of heat and cold stress with Thermal Trace > ℹ️ If you want to know more > References", "title": "Visualisation of heat and cold stress with Thermal Trace", "chunk_index": 21, "token_count": 1037, "text_raw": "lijk Nederlands Meteorologisch Instituut. Accessed: May 22, 2026. [Online]. Available: https://www.knmi.nl/nederland-nu/klimatologie/lijsten/koudegolven\n\n[[Kundzewicz+16](https://doi.org/10.1515/igbp-2016-0005)] Z. W. Kundzewicz, ‘Extreme Weather Events and their Consequences’, Papers on Global Change IGBP, vol. 23, no. 1, pp. 59–69, Jan. 2016, doi: 10.1515/igbp-2016-0005.\n\n[[Liu+24](https://doi.org/10.1016/j.ecolind.2024.112481)] R. Liu et al., ‘Global-scale ERA5 product precipitation and temperature evaluation’, Ecological Indicators, vol. 166, p. 112481, Sep. 2024, doi: 10.1016/j.ecolind.2024.112481.\n\n[[Lussana+24](https://doi.org/10.1002/asl.1239)] C. Lussana, F. Cavalleri, M. Brunetti, V. Manara, and M. Maugeri, ‘Evaluating long-term trends in annual precipitation: A temporal consistency analysis of ERA5 data in the Alps and Italy’, Atmospheric Science Letters, vol. 25, no. 9, p. e1239, Apr. 2024, doi: 10.1002/asl.1239.\n\n[[Ma+19](https://doi.org/10.1175/JCLI-D-18-0234.1)] S. Ma and C. Zhu, ‘Extreme Cold Wave over East Asia in January 2016: A Possible Response to the Larger Internal Atmospheric Variability Induced by Arctic Warming’, Journal of Climate, vol. 32, no. 4, pp. 1203–1216, Feb. 2019, doi: 10.1175/JCLI-D-18-0234.1.\n\n[[Menary+25](https://www.ecmwf.int/en/newsletter/185/news/thermal-trace-health-related-weather-and-climate-monitoring)] M. Menary, R. Emerton, C. Barnard, A. Lombardi, J. Varndell, and C. Cagnazzo, ‘Thermal Trace: health-related weather and climate monitoring’, ECMWF Newsletter, vol. 185, pp. 5–6, Oct. 2025.\n\n[[Morris+18](https://www.theguardian.com/uk-news/2018/mar/02/death-toll-reaches-10-as-destructive-weather-batters-uk)] S. Morris, M. Weaver, and J. Halliday, ‘UK storm death toll reaches 10 as ice warnings follow snow chaos’, The Guardian, Mar. 02, 2018. Accessed: May 22, 2026. [Online]. Available: https://www.theguardian.com/uk-news/2018/mar/02/death-toll-reaches-10-as-destructive-weather-batters-uk\n\n[[Overton+25](https://www.mainepublic.org/climate/2025-08-07/heat-danger-maines-most-vulnerable-at-risk-from-rising-summer-temps)] P. Overton, ‘Heat danger: Maine’s most vulnerable at risk from rising summer temps’, Maine Public, Aug. 07, 2025. Accessed: May 27, 2026. [Online]. Available: https://www.mainepublic.org/climate/2025-08-07/heat-danger-maines-most-vulnerable-at-risk-from-rising-summer-temps\n\n[[Prior+11](https://doi.org/10.1002/wea.735)] J. Prior and M. Kendon, ‘The UK winter of 2009/2010 compared with severe winters of the last 100 years’, Weather, vol. 66, no. 1, pp. 4–10, 2011, doi: 10.1002/wea.735.\n\n[[Rowlatt+23](https://www.bbc.co.uk/news/science-environment-66143682)] J. Rowlatt, ‘Excessive heat: Why this summer has been so hot’, BBC, Jul. 13, 2023. Accessed: Jun. 16, 2025. [Online]. Available: https://www.bbc.co.uk/news/science-environment-66143682", "text_with_prefix": "EQC Quality Assessment: \"Visualisation of heat and cold stress with Thermal Trace\"\nDataset: thermaltrace [CDS]\nAspect: extreme-events_q01 | Category: Applications\nSection: Visualisation of heat and cold stress with Thermal Trace > ℹ️ If you want to know more > References\n---\nlijk Nederlands Meteorologisch Instituut. Accessed: May 22, 2026. [Online]. Available: https://www.knmi.nl/nederland-nu/klimatologie/lijsten/koudegolven\n\n[[Kundzewicz+16](https://doi.org/10.1515/igbp-2016-0005)] Z. W. Kundzewicz, ‘Extreme Weather Events and their Consequences’, Papers on Global Change IGBP, vol. 23, no. 1, pp. 59–69, Jan. 2016, doi: 10.1515/igbp-2016-0005.\n\n[[Liu+24](https://doi.org/10.1016/j.ecolind.2024.112481)] R. Liu et al., ‘Global-scale ERA5 product precipitation and temperature evaluation’, Ecological Indicators, vol. 166, p. 112481, Sep. 2024, doi: 10.1016/j.ecolind.2024.112481.\n\n[[Lussana+24](https://doi.org/10.1002/asl.1239)] C. Lussana, F. Cavalleri, M. Brunetti, V. Manara, and M. Maugeri, ‘Evaluating long-term trends in annual precipitation: A temporal consistency analysis of ERA5 data in the Alps and Italy’, Atmospheric Science Letters, vol. 25, no. 9, p. e1239, Apr. 2024, doi: 10.1002/asl.1239.\n\n[[Ma+19](https://doi.org/10.1175/JCLI-D-18-0234.1)] S. Ma and C. Zhu, ‘Extreme Cold Wave over East Asia in January 2016: A Possible Response to the Larger Internal Atmospheric Variability Induced by Arctic Warming’, Journal of Climate, vol. 32, no. 4, pp. 1203–1216, Feb. 2019, doi: 10.1175/JCLI-D-18-0234.1.\n\n[[Menary+25](https://www.ecmwf.int/en/newsletter/185/news/thermal-trace-health-related-weather-and-climate-monitoring)] M. Menary, R. Emerton, C. Barnard, A. Lombardi, J. Varndell, and C. Cagnazzo, ‘Thermal Trace: health-related weather and climate monitoring’, ECMWF Newsletter, vol. 185, pp. 5–6, Oct. 2025.\n\n[[Morris+18](https://www.theguardian.com/uk-news/2018/mar/02/death-toll-reaches-10-as-destructive-weather-batters-uk)] S. Morris, M. Weaver, and J. Halliday, ‘UK storm death toll reaches 10 as ice warnings follow snow chaos’, The Guardian, Mar. 02, 2018. Accessed: May 22, 2026. [Online]. Available: https://www.theguardian.com/uk-news/2018/mar/02/death-toll-reaches-10-as-destructive-weather-batters-uk\n\n[[Overton+25](https://www.mainepublic.org/climate/2025-08-07/heat-danger-maines-most-vulnerable-at-risk-from-rising-summer-temps)] P. Overton, ‘Heat danger: Maine’s most vulnerable at risk from rising summer temps’, Maine Public, Aug. 07, 2025. Accessed: May 27, 2026. [Online]. Available: https://www.mainepublic.org/climate/2025-08-07/heat-danger-maines-most-vulnerable-at-risk-from-rising-summer-temps\n\n[[Prior+11](https://doi.org/10.1002/wea.735)] J. Prior and M. Kendon, ‘The UK winter of 2009/2010 compared with severe winters of the last 100 years’, Weather, vol. 66, no. 1, pp. 4–10, 2011, doi: 10.1002/wea.735.\n\n[[Rowlatt+23](https://www.bbc.co.uk/news/science-environment-66143682)] J. Rowlatt, ‘Excessive heat: Why this summer has been so hot’, BBC, Jul. 13, 2023. Accessed: Jun. 16, 2025. [Online]. Available: https://www.bbc.co.uk/news/science-environment-66143682"} {"chunk_id": "application_thermaltrace_extreme-events_q01__fa15c11925f1", "report_id": "application_thermaltrace_extreme-events_q01", "dataset_id": "thermaltrace", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extreme-events_q01", "aspect_base": "extreme-events", "category": "Applications", "match_confidence": "unmatched", "section": "Visualisation of heat and cold stress with Thermal Trace > ℹ️ If you want to know more > References", "title": "Visualisation of heat and cold stress with Thermal Trace", "chunk_index": 22, "token_count": 1025, "text_raw": "so hot’, BBC, Jul. 13, 2023. Accessed: Jun. 16, 2025. [Online]. Available: https://www.bbc.co.uk/news/science-environment-66143682\n\n[[Royé+21](https://doi.org/10.1097/EDE.0000000000001359)] D. Royé et al., ‘Effects of Hot Nights on Mortality in Southern Europe’, Epidemiology, vol. 32, no. 4, pp. 487–498, Jul. 2021, doi: 10.1097/EDE.0000000000001359.\n\n[[Schreier+13](https://doi.org/10.1007/s00484-012-0525-y)] S. F. Schreier et al., ‘The uncertainty of UTCI due to uncertainties in the determination of radiation fluxes derived from numerical weather prediction and regional climate model simulations’, International Journal of Biometeorology, vol. 57, no. 2, pp. 207–223, Mar. 2013, doi: 10.1007/s00484-012-0525-y.\n\n[[Simmons+21](https://doi.org/10.21957/ly5vbtbfd)] A. Simmons et al., ‘Low frequency variability and trends in surface air temperature and humidity from ERA5 and other datasets’, European Centre for Medium-Range Weather Forecasts, Reading, UK, Technical Memo 881, Feb. 2021. doi: 10.21957/ly5vbtbfd.\n\n[[Sippel+24](https://doi.org/10.5194/wcd-5-943-2024)] S. Sippel et al., ‘Could an extremely cold central European winter such as 1963 happen again despite climate change?’, Weather and Climate Dynamics, vol. 5, no. 3, pp. 943–957, Jul. 2024, doi: 10.5194/wcd-5-943-2024.\n\n[[Soci+24](https://doi.org/10.1002/qj.4803)] C. Soci et al., ‘The ERA5 global reanalysis from 1940 to 2022’, Quarterly Journal of the Royal Meteorological Society, vol. 150, no. 764, pp. 4014–4048, Jul. 2024, doi: 10.1002/qj.4803.\n\n[[Steadman+84](https://doi.org/10.1175/1520-0450%281984%29023%3C1674:AUSOAT%3E2.0.CO;2)] R. G. Steadman, ‘A Universal Scale of Apparent Temperature’, Journal of Applied Meteorology and Climatology, vol. 23, no. 12, pp. 1674–1687, Dec. 1984, doi: 10.1175/1520-0450(1984)023<1674:AUSOAT>2.0.​CO;2.\n\n[[Thompson+22](https://doi.org/10.1126/sciadv.abm6860)] V. Thompson et al., ‘The 2021 western North America heat wave among the most extreme events ever recorded globally’, Science Advances, vol. 8, no. 18, p. eabm6860, May 2022, doi: 10.1126/sciadv.abm6860.\n\n[[Tousi+24](https://doi.org/10.3390/urbansci8040193)] E. Tousi, A. Mela, and A. Tseliou, ‘Thermal Stress in Outdoor Spaces During Mediterranean Heatwaves: A PET and UTCI Analysis of Different Demographics’, Urban Science, vol. 8, no. 4, p. 193, Oct. 2024, doi: 10.3390/urbansci8040193.\n\n[[Twardosz+16](https://doi.org/10.1515/acgeo-2016-0083)] R. Twardosz, U. Kossowska-Cezak, and S. Pełech, ‘Extremely Cold Winter Months in Europe (1951–2010)’, Acta Geophys., vol. 64, no. 6, pp. 2609–2629, Dec. 2016, doi: 10.1515/acgeo-2016-0083.", "text_with_prefix": "EQC Quality Assessment: \"Visualisation of heat and cold stress with Thermal Trace\"\nDataset: thermaltrace [CDS]\nAspect: extreme-events_q01 | Category: Applications\nSection: Visualisation of heat and cold stress with Thermal Trace > ℹ️ If you want to know more > References\n---\nso hot’, BBC, Jul. 13, 2023. Accessed: Jun. 16, 2025. [Online]. Available: https://www.bbc.co.uk/news/science-environment-66143682\n\n[[Royé+21](https://doi.org/10.1097/EDE.0000000000001359)] D. Royé et al., ‘Effects of Hot Nights on Mortality in Southern Europe’, Epidemiology, vol. 32, no. 4, pp. 487–498, Jul. 2021, doi: 10.1097/EDE.0000000000001359.\n\n[[Schreier+13](https://doi.org/10.1007/s00484-012-0525-y)] S. F. Schreier et al., ‘The uncertainty of UTCI due to uncertainties in the determination of radiation fluxes derived from numerical weather prediction and regional climate model simulations’, International Journal of Biometeorology, vol. 57, no. 2, pp. 207–223, Mar. 2013, doi: 10.1007/s00484-012-0525-y.\n\n[[Simmons+21](https://doi.org/10.21957/ly5vbtbfd)] A. Simmons et al., ‘Low frequency variability and trends in surface air temperature and humidity from ERA5 and other datasets’, European Centre for Medium-Range Weather Forecasts, Reading, UK, Technical Memo 881, Feb. 2021. doi: 10.21957/ly5vbtbfd.\n\n[[Sippel+24](https://doi.org/10.5194/wcd-5-943-2024)] S. Sippel et al., ‘Could an extremely cold central European winter such as 1963 happen again despite climate change?’, Weather and Climate Dynamics, vol. 5, no. 3, pp. 943–957, Jul. 2024, doi: 10.5194/wcd-5-943-2024.\n\n[[Soci+24](https://doi.org/10.1002/qj.4803)] C. Soci et al., ‘The ERA5 global reanalysis from 1940 to 2022’, Quarterly Journal of the Royal Meteorological Society, vol. 150, no. 764, pp. 4014–4048, Jul. 2024, doi: 10.1002/qj.4803.\n\n[[Steadman+84](https://doi.org/10.1175/1520-0450%281984%29023%3C1674:AUSOAT%3E2.0.CO;2)] R. G. Steadman, ‘A Universal Scale of Apparent Temperature’, Journal of Applied Meteorology and Climatology, vol. 23, no. 12, pp. 1674–1687, Dec. 1984, doi: 10.1175/1520-0450(1984)023<1674:AUSOAT>2.0.​CO;2.\n\n[[Thompson+22](https://doi.org/10.1126/sciadv.abm6860)] V. Thompson et al., ‘The 2021 western North America heat wave among the most extreme events ever recorded globally’, Science Advances, vol. 8, no. 18, p. eabm6860, May 2022, doi: 10.1126/sciadv.abm6860.\n\n[[Tousi+24](https://doi.org/10.3390/urbansci8040193)] E. Tousi, A. Mela, and A. Tseliou, ‘Thermal Stress in Outdoor Spaces During Mediterranean Heatwaves: A PET and UTCI Analysis of Different Demographics’, Urban Science, vol. 8, no. 4, p. 193, Oct. 2024, doi: 10.3390/urbansci8040193.\n\n[[Twardosz+16](https://doi.org/10.1515/acgeo-2016-0083)] R. Twardosz, U. Kossowska-Cezak, and S. Pełech, ‘Extremely Cold Winter Months in Europe (1951–2010)’, Acta Geophys., vol. 64, no. 6, pp. 2609–2629, Dec. 2016, doi: 10.1515/acgeo-2016-0083."} {"chunk_id": "application_thermaltrace_extreme-events_q01__50f231f31aa8", "report_id": "application_thermaltrace_extreme-events_q01", "dataset_id": "thermaltrace", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extreme-events_q01", "aspect_base": "extreme-events", "category": "Applications", "match_confidence": "unmatched", "section": "Visualisation of heat and cold stress with Thermal Trace > ℹ️ If you want to know more > References", "title": "Visualisation of heat and cold stress with Thermal Trace", "chunk_index": 23, "token_count": 882, "text_raw": "a Geophys., vol. 64, no. 6, pp. 2609–2629, Dec. 2016, doi: 10.1515/acgeo-2016-0083.\n\n[[Ullah+24](https://doi.org/10.1038/s41598-024-54766-7)] S. Ullah, A. Aldossary, W. Ullah, and S. G. Al-Ghamdi, ‘Augmented human thermal discomfort in urban centers of the Arabian Peninsula’, Scientific Reports, vol. 14, no. 1, p. 3974, Feb. 2024, doi: 10.1038/s41598-024-54766-7.\n\n[[Van Oldenborgh+19](https://doi.org/10.1088/1748-9326/ab4867)] G. J. van Oldenborgh, E. Mitchell-Larson, G. A. Vecchi, H. de Vries, R. Vautard, and F. Otto, ‘Cold waves are getting milder in the northern midlatitudes’, Environmental Research Letters, vol. 14, no. 11, p. 114004, Oct. 2019, doi: 10.1088/1748-9326/ab4867.\n\n[[Vargas Zeppetello+22](https://doi.org/10.1038/s43247-022-00524-4)] L. R. Vargas Zeppetello, A. E. Raftery, and D. S. Battisti, ‘Probabilistic projections of increased heat stress driven by climate change’, Communications Earth & Environment, vol. 3, p. 183, Aug. 2022, doi: 10.1038/s43247-022-00524-4.\n\n[[Wang+25](https://doi.org/10.3390/rs17071317)] H. Wang and Y. Wang, ‘Evaluation of the Accuracy and Trend Consistency of Hourly Surface Solar Radiation Datasets of ERA5, MERRA-2, SARAH-E, CERES, and Solcast over China’, Remote Sensing, vol. 17, no. 7, p. 1317, Apr. 2025, doi: 10.3390/rs17071317.\n\n[[Wielinga+24](https://www.groningentoen.nl/elfstedentocht/medische-gevolgen-elfstedentocht-1963/)] M. Wielinga, ‘Medische aspecten Elfstedentocht 1963’, Groningen Toen. Accessed: May 22, 2026. [Online]. Available: https://www.groningentoen.nl/elfstedentocht/medische-gevolgen-elfstedentocht-1963/\n\n[[Young+26](https://www.bbc.co.uk/news/articles/cp8ppdk5zp2o)] L. Young, ‘Amber heat health alert issued for the South West by UKHSA’, BBC, May 26, 2026. Accessed: May 27, 2026. [Online]. Available: https://www.bbc.co.uk/news/articles/cp8ppdk5zp2o\n\n[[Zhao+21](https://doi.org/10.1016/S2542-5196%2821%2900081-4)] Q. Zhao et al., ‘Global, regional, and national burden of mortality associated with non-optimal ambient temperatures from 2000 to 2019: a three-stage modelling study’, The Lancet Planetary Health, vol. 5, no. 7, pp. e415–e425, Jul. 2021, doi: 10.1016/S2542-5196(21)00081-4.", "text_with_prefix": "EQC Quality Assessment: \"Visualisation of heat and cold stress with Thermal Trace\"\nDataset: thermaltrace [CDS]\nAspect: extreme-events_q01 | Category: Applications\nSection: Visualisation of heat and cold stress with Thermal Trace > ℹ️ If you want to know more > References\n---\na Geophys., vol. 64, no. 6, pp. 2609–2629, Dec. 2016, doi: 10.1515/acgeo-2016-0083.\n\n[[Ullah+24](https://doi.org/10.1038/s41598-024-54766-7)] S. Ullah, A. Aldossary, W. Ullah, and S. G. Al-Ghamdi, ‘Augmented human thermal discomfort in urban centers of the Arabian Peninsula’, Scientific Reports, vol. 14, no. 1, p. 3974, Feb. 2024, doi: 10.1038/s41598-024-54766-7.\n\n[[Van Oldenborgh+19](https://doi.org/10.1088/1748-9326/ab4867)] G. J. van Oldenborgh, E. Mitchell-Larson, G. A. Vecchi, H. de Vries, R. Vautard, and F. Otto, ‘Cold waves are getting milder in the northern midlatitudes’, Environmental Research Letters, vol. 14, no. 11, p. 114004, Oct. 2019, doi: 10.1088/1748-9326/ab4867.\n\n[[Vargas Zeppetello+22](https://doi.org/10.1038/s43247-022-00524-4)] L. R. Vargas Zeppetello, A. E. Raftery, and D. S. Battisti, ‘Probabilistic projections of increased heat stress driven by climate change’, Communications Earth & Environment, vol. 3, p. 183, Aug. 2022, doi: 10.1038/s43247-022-00524-4.\n\n[[Wang+25](https://doi.org/10.3390/rs17071317)] H. Wang and Y. Wang, ‘Evaluation of the Accuracy and Trend Consistency of Hourly Surface Solar Radiation Datasets of ERA5, MERRA-2, SARAH-E, CERES, and Solcast over China’, Remote Sensing, vol. 17, no. 7, p. 1317, Apr. 2025, doi: 10.3390/rs17071317.\n\n[[Wielinga+24](https://www.groningentoen.nl/elfstedentocht/medische-gevolgen-elfstedentocht-1963/)] M. Wielinga, ‘Medische aspecten Elfstedentocht 1963’, Groningen Toen. Accessed: May 22, 2026. [Online]. Available: https://www.groningentoen.nl/elfstedentocht/medische-gevolgen-elfstedentocht-1963/\n\n[[Young+26](https://www.bbc.co.uk/news/articles/cp8ppdk5zp2o)] L. Young, ‘Amber heat health alert issued for the South West by UKHSA’, BBC, May 26, 2026. Accessed: May 27, 2026. [Online]. Available: https://www.bbc.co.uk/news/articles/cp8ppdk5zp2o\n\n[[Zhao+21](https://doi.org/10.1016/S2542-5196%2821%2900081-4)] Q. Zhao et al., ‘Global, regional, and national burden of mortality associated with non-optimal ambient temperatures from 2000 to 2019: a three-stage modelling study’, The Lancet Planetary Health, vol. 5, no. 7, pp. e415–e425, Jul. 2021, doi: 10.1016/S2542-5196(21)00081-4."} {"chunk_id": "climate_projections-cmip6_climate-and-weather-extremes_q01__d3a5fe8864e9", "report_id": "climate_projections-cmip6_climate-and-weather-extremes_q01", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q01", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Bias in extreme temperature indices for the reinsurance sector", "title": "Bias in extreme temperature indices for the reinsurance sector", "chunk_index": 0, "token_count": 96, "text_raw": "Production date: 23-05-2024\n\nProduced by: CMCC foundation - Euro-Mediterranean Center on Climate Change. Albert Martinez Boti.", "text_with_prefix": "EQC Quality Assessment: \"Bias in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-and-weather-extremes_q01 | Category: Climate_Projections\nSection: Bias in extreme temperature indices for the reinsurance sector\n---\nProduction date: 23-05-2024\n\nProduced by: CMCC foundation - Euro-Mediterranean Center on Climate Change. Albert Martinez Boti."} {"chunk_id": "climate_projections-cmip6_climate-and-weather-extremes_q01__926f2e951e87", "report_id": "climate_projections-cmip6_climate-and-weather-extremes_q01", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q01", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Bias in extreme temperature indices for the reinsurance sector > Quality assessment question", "title": "Bias in extreme temperature indices for the reinsurance sector", "chunk_index": 1, "token_count": 640, "text_raw": "* **How well do CMIP6 projections represent climatology and trends of air temperature extremes in Europe?**\n\nClimate change has a major impact on the reinsurance market [[1]](https://doi.org/10.3390/atmos11020146)[[2]](https://doi.org/10.5194/nhess-22-659-2022). In the third assessment report of the IPCC, hot temperature extremes were already presented as relevant to insurance and related services [[3]](https://www.ipcc.ch/site/assets/uploads/2018/03/wg2TARchap8.pdf). Consequently, the need for reliable regional and global climate projections has become paramount, offering valuable insights for optimising reinsurance strategies in the face of a changing climate landscape. Nonetheless, despite their pivotal role, uncertainties inherent in these projections can potentially lead to misuse [[4]](https://doi.org/10.1002/wcc.71)[[5]](https://doi.org/10.1002/wcc.579). This underscores the importance of accurately calculating and accounting for uncertainties to ensure their appropriate consideration. This notebook utilises data from a subset of models from [CMIP6](https://cds.climate.copernicus.eu/datasets/projections-cmip6?tab=overview) Global Climate Models (GCMs) and compares them with [ERA5](https://cds.climate.copernicus.eu/datasets/reanalysis-era5-complete?tab=doc) reanalysis, serving as the reference product. Two maximum-temperature-based indices from [ECA&D](https://www.ecad.eu/indicesextremes/) indices (one of physical nature and the other of statistical nature) are computed using the [icclim](https://icclim.readthedocs.io/en/stable/) Python package. The first index, identified by the ETCCDI short name 'SU', quantifies the occurrence of summer days (i.e., with daily maximum temperatures exceeding 25°C) within a year or a season (JJA in this notebook). The second index, labeled 'TX90p', describes the number of days with daily maximum temperatures exceeding the daily 90th percentile of maximum temperature for a 5-day moving window. Within this notebook, these calculations are performed over the historical period spanning from 1971 to 2000. It is important to mention that the results presented here pertain to a specific subset of the CMIP6 ensemble and may not be generalisable to the entire dataset. Also note that a separate assessment examines the representation of trends of these indices for the same models during a fixed future period (2015-2099), while another assessment looks at the projected climate signal of these indices for the same models at a 2°C Global Warming Level.", "text_with_prefix": "EQC Quality Assessment: \"Bias in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-and-weather-extremes_q01 | Category: Climate_Projections\nSection: Bias in extreme temperature indices for the reinsurance sector > Quality assessment question\n---\n* **How well do CMIP6 projections represent climatology and trends of air temperature extremes in Europe?**\n\nClimate change has a major impact on the reinsurance market [[1]](https://doi.org/10.3390/atmos11020146)[[2]](https://doi.org/10.5194/nhess-22-659-2022). In the third assessment report of the IPCC, hot temperature extremes were already presented as relevant to insurance and related services [[3]](https://www.ipcc.ch/site/assets/uploads/2018/03/wg2TARchap8.pdf). Consequently, the need for reliable regional and global climate projections has become paramount, offering valuable insights for optimising reinsurance strategies in the face of a changing climate landscape. Nonetheless, despite their pivotal role, uncertainties inherent in these projections can potentially lead to misuse [[4]](https://doi.org/10.1002/wcc.71)[[5]](https://doi.org/10.1002/wcc.579). This underscores the importance of accurately calculating and accounting for uncertainties to ensure their appropriate consideration. This notebook utilises data from a subset of models from [CMIP6](https://cds.climate.copernicus.eu/datasets/projections-cmip6?tab=overview) Global Climate Models (GCMs) and compares them with [ERA5](https://cds.climate.copernicus.eu/datasets/reanalysis-era5-complete?tab=doc) reanalysis, serving as the reference product. Two maximum-temperature-based indices from [ECA&D](https://www.ecad.eu/indicesextremes/) indices (one of physical nature and the other of statistical nature) are computed using the [icclim](https://icclim.readthedocs.io/en/stable/) Python package. The first index, identified by the ETCCDI short name 'SU', quantifies the occurrence of summer days (i.e., with daily maximum temperatures exceeding 25°C) within a year or a season (JJA in this notebook). The second index, labeled 'TX90p', describes the number of days with daily maximum temperatures exceeding the daily 90th percentile of maximum temperature for a 5-day moving window. Within this notebook, these calculations are performed over the historical period spanning from 1971 to 2000. It is important to mention that the results presented here pertain to a specific subset of the CMIP6 ensemble and may not be generalisable to the entire dataset. Also note that a separate assessment examines the representation of trends of these indices for the same models during a fixed future period (2015-2099), while another assessment looks at the projected climate signal of these indices for the same models at a 2°C Global Warming Level."} {"chunk_id": "climate_projections-cmip6_climate-and-weather-extremes_q01__0d3ad264729a", "report_id": "climate_projections-cmip6_climate-and-weather-extremes_q01", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q01", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Bias in extreme temperature indices for the reinsurance sector > Quality assessment statement", "title": "Bias in extreme temperature indices for the reinsurance sector", "chunk_index": 2, "token_count": 488, "text_raw": "These are the key outcomes of this assessment\n\n* CMIP6 projections based on maximum temperature indices can be valuable for designing strategies to optimise reinsurance protections. However, it is crucial for users to acknowledge that all climate projections inherently contain certain biases. This is especially significant when using projections in regions with complex orography or extensive continental areas, where biases are more pronounced and impact the accuracy of the projections. Therefore, users should exercise caution and consider these limitations when applying the projections to their specific use cases, understanding that biases in the mean climate and trends affect indices differently and that the choice of bias correction method should be tailored accordingly [[6]](https://doi.org/10.1002/asl.1072)[[7]](https://ibicus.readthedocs.io/en/latest/).\n\n* GCMs exhibit biases in capturing mean values of extreme temperature indices like the number of summer days ('SU') and days exceeding the 90th percentile threshold ('TX90p'), with underestimation of index magnitudes, especially in continental areas and regions with complex terrain.\n\n* ERA5 shows that the magnitude of the historical trend varies spatially. The trend bias assessment suggests that, overall, the trend derived from the CMIP6 projections is underestimated with slight overestimations observed in certain regions. Hence, users need to consider this regional variation in both the trend and its bias.\n\n* Despite limitations and biases, the considered subset of CMIP6 models generally reproduces historical trends reasonably well for the summer season (JJA), providing a foundation for understanding past climate behavior. These findings also increase confidence (though they do not ensure accuracy) when analysing future trends using these models.\n```\n\nattachment:1c0e8cf0-01f8-4c79-a63c-395662058d5e.png\n---\nwidth: 900px\nalt: Mean Bias SU\n---\nNumber of summer days ('SU') for the temporal aggregation of 'JJA'. Mean bias for the historical period (1971 - 2000) of each individual CMIP6 model.\n```", "text_with_prefix": "EQC Quality Assessment: \"Bias in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-and-weather-extremes_q01 | Category: Climate_Projections\nSection: Bias in extreme temperature indices for the reinsurance sector > Quality assessment statement\n---\nThese are the key outcomes of this assessment\n\n* CMIP6 projections based on maximum temperature indices can be valuable for designing strategies to optimise reinsurance protections. However, it is crucial for users to acknowledge that all climate projections inherently contain certain biases. This is especially significant when using projections in regions with complex orography or extensive continental areas, where biases are more pronounced and impact the accuracy of the projections. Therefore, users should exercise caution and consider these limitations when applying the projections to their specific use cases, understanding that biases in the mean climate and trends affect indices differently and that the choice of bias correction method should be tailored accordingly [[6]](https://doi.org/10.1002/asl.1072)[[7]](https://ibicus.readthedocs.io/en/latest/).\n\n* GCMs exhibit biases in capturing mean values of extreme temperature indices like the number of summer days ('SU') and days exceeding the 90th percentile threshold ('TX90p'), with underestimation of index magnitudes, especially in continental areas and regions with complex terrain.\n\n* ERA5 shows that the magnitude of the historical trend varies spatially. The trend bias assessment suggests that, overall, the trend derived from the CMIP6 projections is underestimated with slight overestimations observed in certain regions. Hence, users need to consider this regional variation in both the trend and its bias.\n\n* Despite limitations and biases, the considered subset of CMIP6 models generally reproduces historical trends reasonably well for the summer season (JJA), providing a foundation for understanding past climate behavior. These findings also increase confidence (though they do not ensure accuracy) when analysing future trends using these models.\n```\n\nattachment:1c0e8cf0-01f8-4c79-a63c-395662058d5e.png\n---\nwidth: 900px\nalt: Mean Bias SU\n---\nNumber of summer days ('SU') for the temporal aggregation of 'JJA'. Mean bias for the historical period (1971 - 2000) of each individual CMIP6 model.\n```"} {"chunk_id": "climate_projections-cmip6_climate-and-weather-extremes_q01__a71ad31eb563", "report_id": "climate_projections-cmip6_climate-and-weather-extremes_q01", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q01", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Bias in extreme temperature indices for the reinsurance sector > Methodology", "title": "Bias in extreme temperature indices for the reinsurance sector", "chunk_index": 3, "token_count": 435, "text_raw": "This notebook provides an assessment of the systematic errors (trend and climate mean) in a subset of 16 models from [CMIP6](https://cds.climate.copernicus.eu/datasets/projections-cmip6?tab=overview). It achieves this by comparing the model predictions with the ERA5 reanalysis for the maximum-temperature-based indices of 'SU' and 'TX90p', calculated over the temporal aggregation of JJA and for the historical period spanning from 1971 to 2000 (chosen to allow comparison to [CORDEX](https://cds.climate.copernicus.eu/datasets/projections-cordex-domains-single-levels?tab=overview) models in another assessment). In particular, spatial patterns of climate mean and trend, along with biases, are examined and displayed for each model and the ensemble median (calculated for each grid cell). Additionally, spatially-averaged trend values are analysed and presented using box plots to provide an overview of trend behavior across the distribution of the chosen subset of models when averaged across Europe.\n\nThe analysis and results follow the next outline:\n\n**[](section-1)**\n * [](section-1.1)\n * [](section-1.2)\n * [](section-1.3)\n * [](section-1.4)\n * [](section-1.5)\n * [](section-1.6)\n\n**[](section-2)**\n * [](section-2.1)\n * [](section-2.2)\n * [](section-2.3)\n\n**[](section-3)**\n * [](section-3.1)\n * [](section-3.2)\n * [](section-3.3)\n * [](section-3.4)\n * [](section-3.5)\n * [](section-3.6)", "text_with_prefix": "EQC Quality Assessment: \"Bias in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-and-weather-extremes_q01 | Category: Climate_Projections\nSection: Bias in extreme temperature indices for the reinsurance sector > Methodology\n---\nThis notebook provides an assessment of the systematic errors (trend and climate mean) in a subset of 16 models from [CMIP6](https://cds.climate.copernicus.eu/datasets/projections-cmip6?tab=overview). It achieves this by comparing the model predictions with the ERA5 reanalysis for the maximum-temperature-based indices of 'SU' and 'TX90p', calculated over the temporal aggregation of JJA and for the historical period spanning from 1971 to 2000 (chosen to allow comparison to [CORDEX](https://cds.climate.copernicus.eu/datasets/projections-cordex-domains-single-levels?tab=overview) models in another assessment). In particular, spatial patterns of climate mean and trend, along with biases, are examined and displayed for each model and the ensemble median (calculated for each grid cell). Additionally, spatially-averaged trend values are analysed and presented using box plots to provide an overview of trend behavior across the distribution of the chosen subset of models when averaged across Europe.\n\nThe analysis and results follow the next outline:\n\n**[](section-1)**\n * [](section-1.1)\n * [](section-1.2)\n * [](section-1.3)\n * [](section-1.4)\n * [](section-1.5)\n * [](section-1.6)\n\n**[](section-2)**\n * [](section-2.1)\n * [](section-2.2)\n * [](section-2.3)\n\n**[](section-3)**\n * [](section-3.1)\n * [](section-3.2)\n * [](section-3.3)\n * [](section-3.4)\n * [](section-3.5)\n * [](section-3.6)"} {"chunk_id": "climate_projections-cmip6_climate-and-weather-extremes_q01__0725b7ec5603", "report_id": "climate_projections-cmip6_climate-and-weather-extremes_q01", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q01", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Bias in extreme temperature indices for the reinsurance sector > Analysis and results > 1. Parameters, requests and functions definition > 1.2. Define Parameters", "title": "Bias in extreme temperature indices for the reinsurance sector", "chunk_index": 4, "token_count": 334, "text_raw": "In the \"Define Parameters\" section, various customisable options for the notebook are specified:\n\n- The initial and ending year used for the historical period can be specified by changing the parameters `year_start` and `year_stop` (1971-2000 is chosen for consistency between CMIP6 and CORDEX).\n- The `timeseries` set the temporal aggregation. For instance, selecting \"JJA\" implies considering only the JJA season.\n- `collection_id` provides the choice between Global Climate Models CMIP6 or Regional Climate Models CORDEX. Although the code allows choosing between CMIP6 or CORDEX, the example provided in this notebook deals with CMIP6.\n- `area` allows specifying the geographical domain of interest.\n- The `interpolation_method` parameter allows selecting the interpolation method when regridding is performed over the indices.\n- The `chunk` selection allows the user to define if dividing into chunks when downloading the data on their local machine. Although it does not significantly affect the analysis, it is recommended to keep the default value for optimal performance.\n\nChoose annual or seasonal timeseries\nChoose CORDEX or CMIP6\nInterpolation method\nArea to show\nChunks for download\n\n(section-1.3)=", "text_with_prefix": "EQC Quality Assessment: \"Bias in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-and-weather-extremes_q01 | Category: Climate_Projections\nSection: Bias in extreme temperature indices for the reinsurance sector > Analysis and results > 1. Parameters, requests and functions definition > 1.2. Define Parameters\n---\nIn the \"Define Parameters\" section, various customisable options for the notebook are specified:\n\n- The initial and ending year used for the historical period can be specified by changing the parameters `year_start` and `year_stop` (1971-2000 is chosen for consistency between CMIP6 and CORDEX).\n- The `timeseries` set the temporal aggregation. For instance, selecting \"JJA\" implies considering only the JJA season.\n- `collection_id` provides the choice between Global Climate Models CMIP6 or Regional Climate Models CORDEX. Although the code allows choosing between CMIP6 or CORDEX, the example provided in this notebook deals with CMIP6.\n- `area` allows specifying the geographical domain of interest.\n- The `interpolation_method` parameter allows selecting the interpolation method when regridding is performed over the indices.\n- The `chunk` selection allows the user to define if dividing into chunks when downloading the data on their local machine. Although it does not significantly affect the analysis, it is recommended to keep the default value for optimal performance.\n\nChoose annual or seasonal timeseries\nChoose CORDEX or CMIP6\nInterpolation method\nArea to show\nChunks for download\n\n(section-1.3)="} {"chunk_id": "climate_projections-cmip6_climate-and-weather-extremes_q01__7d33e83db972", "report_id": "climate_projections-cmip6_climate-and-weather-extremes_q01", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q01", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Bias in extreme temperature indices for the reinsurance sector > Analysis and results > 1. Parameters, requests and functions definition > 1.3. Define models", "title": "Bias in extreme temperature indices for the reinsurance sector", "chunk_index": 5, "token_count": 213, "text_raw": "The following climate analyses are performed considering a subset of GCMs from CMIP6. Models names are listed in the parameters below. Some variable-dependent parameters are also selected, as the `index_names` parameter, which specifies the temperature-based indices ('SU' and 'TX90p' in our case) from the [icclim](https://icclim.readthedocs.io/en/stable/) Python package.\n\nThe selected CMIP6 models have available both the historical and SSP5-8.5 experiments.\n\nDefine dictionaries to use in titles and caption\nDefine dictionaries to use in titles and caption\n\n(section-1.4)=", "text_with_prefix": "EQC Quality Assessment: \"Bias in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-and-weather-extremes_q01 | Category: Climate_Projections\nSection: Bias in extreme temperature indices for the reinsurance sector > Analysis and results > 1. Parameters, requests and functions definition > 1.3. Define models\n---\nThe following climate analyses are performed considering a subset of GCMs from CMIP6. Models names are listed in the parameters below. Some variable-dependent parameters are also selected, as the `index_names` parameter, which specifies the temperature-based indices ('SU' and 'TX90p' in our case) from the [icclim](https://icclim.readthedocs.io/en/stable/) Python package.\n\nThe selected CMIP6 models have available both the historical and SSP5-8.5 experiments.\n\nDefine dictionaries to use in titles and caption\nDefine dictionaries to use in titles and caption\n\n(section-1.4)="} {"chunk_id": "climate_projections-cmip6_climate-and-weather-extremes_q01__4382fcb98ca8", "report_id": "climate_projections-cmip6_climate-and-weather-extremes_q01", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q01", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Bias in extreme temperature indices for the reinsurance sector > Analysis and results > 1. Parameters, requests and functions definition > 1.4. Define ERA5 request", "title": "Bias in extreme temperature indices for the reinsurance sector", "chunk_index": 6, "token_count": 123, "text_raw": "Within this notebook, ERA5 serves as the reference product. In this section, we set the required parameters for the cds-api data-request of ERA5.\n\n(section-1.5)=", "text_with_prefix": "EQC Quality Assessment: \"Bias in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-and-weather-extremes_q01 | Category: Climate_Projections\nSection: Bias in extreme temperature indices for the reinsurance sector > Analysis and results > 1. Parameters, requests and functions definition > 1.4. Define ERA5 request\n---\nWithin this notebook, ERA5 serves as the reference product. In this section, we set the required parameters for the cds-api data-request of ERA5.\n\n(section-1.5)="} {"chunk_id": "climate_projections-cmip6_climate-and-weather-extremes_q01__aa09b0c08e9a", "report_id": "climate_projections-cmip6_climate-and-weather-extremes_q01", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q01", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Bias in extreme temperature indices for the reinsurance sector > Analysis and results > 1. Parameters, requests and functions definition > 1.5. Define model requests", "title": "Bias in extreme temperature indices for the reinsurance sector", "chunk_index": 7, "token_count": 190, "text_raw": "In this section we set the required parameters for the cds-api data-request.\n\nWhen `weights = True`, spatial weighting is applied for calculations requiring spatial data aggregation. This is particularly relevant for CMIP6 GCMs with regular lon-lat grids that do not consider varying surface extensions at different latitudes. In contrast, CORDEX RCMs, using rotated grids, inherently account for different cell surfaces based on latitude, eliminating the need for a latitude cosine multiplicative factor (`weights = False`).\n\n(section-1.6)=", "text_with_prefix": "EQC Quality Assessment: \"Bias in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-and-weather-extremes_q01 | Category: Climate_Projections\nSection: Bias in extreme temperature indices for the reinsurance sector > Analysis and results > 1. Parameters, requests and functions definition > 1.5. Define model requests\n---\nIn this section we set the required parameters for the cds-api data-request.\n\nWhen `weights = True`, spatial weighting is applied for calculations requiring spatial data aggregation. This is particularly relevant for CMIP6 GCMs with regular lon-lat grids that do not consider varying surface extensions at different latitudes. In contrast, CORDEX RCMs, using rotated grids, inherently account for different cell surfaces based on latitude, eliminating the need for a latitude cosine multiplicative factor (`weights = False`).\n\n(section-1.6)="} {"chunk_id": "climate_projections-cmip6_climate-and-weather-extremes_q01__052d6703af6f", "report_id": "climate_projections-cmip6_climate-and-weather-extremes_q01", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q01", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Bias in extreme temperature indices for the reinsurance sector > Analysis and results > 1. Parameters, requests and functions definition > 1.6. Functions to cache", "title": "Bias in extreme temperature indices for the reinsurance sector", "chunk_index": 8, "token_count": 316, "text_raw": "In this section, functions that will be executed in the caching phase are defined. Caching is the process of storing copies of files in a temporary storage location, so that they can be accessed more quickly. This process also checks if the user has already downloaded a file, avoiding redundant downloads.\n\nFunctions description:\n\n- The `select_timeseries` function subsets the dataset based on the chosen `timeseries` parameter.\n\n- The `compute_indices` function utilises the icclim package to calculate the selected maximum-temperature-based indices.\n\n- The `compute_trends` function employs the Mann-Kendall test for trend calculation.\n\n- Finally, the `compute_indices_and_trends` function selects the temporal aggregation using the `select_timeseries` function. It then computes daily maximum temperature (only if we are dealing with ERA5), calculates maximum-temperature-based indices via the `compute_indices` function, determines the indices mean over the historical period (1971-2000), obtain the trends using the `compute_trends` function, and offers an option for regridding to ERA5 if required.\n\nOriginal bounds for conservative interpolation\n\n(section-2)=", "text_with_prefix": "EQC Quality Assessment: \"Bias in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-and-weather-extremes_q01 | Category: Climate_Projections\nSection: Bias in extreme temperature indices for the reinsurance sector > Analysis and results > 1. Parameters, requests and functions definition > 1.6. Functions to cache\n---\nIn this section, functions that will be executed in the caching phase are defined. Caching is the process of storing copies of files in a temporary storage location, so that they can be accessed more quickly. This process also checks if the user has already downloaded a file, avoiding redundant downloads.\n\nFunctions description:\n\n- The `select_timeseries` function subsets the dataset based on the chosen `timeseries` parameter.\n\n- The `compute_indices` function utilises the icclim package to calculate the selected maximum-temperature-based indices.\n\n- The `compute_trends` function employs the Mann-Kendall test for trend calculation.\n\n- Finally, the `compute_indices_and_trends` function selects the temporal aggregation using the `select_timeseries` function. It then computes daily maximum temperature (only if we are dealing with ERA5), calculates maximum-temperature-based indices via the `compute_indices` function, determines the indices mean over the historical period (1971-2000), obtain the trends using the `compute_trends` function, and offers an option for regridding to ERA5 if required.\n\nOriginal bounds for conservative interpolation\n\n(section-2)="} {"chunk_id": "climate_projections-cmip6_climate-and-weather-extremes_q01__4c22d1ad6549", "report_id": "climate_projections-cmip6_climate-and-weather-extremes_q01", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q01", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Bias in extreme temperature indices for the reinsurance sector > Analysis and results > 2. Downloading and processing > 2.1. Download and transform ERA5", "title": "Bias in extreme temperature indices for the reinsurance sector", "chunk_index": 9, "token_count": 182, "text_raw": "In this section, the `download.download_and_transform` function from the 'c3s_eqc_automatic_quality_control' package is employed to download ERA5 reference data, select the temporal aggregation (\"JJA\" in our example), compute daily maximum temperature from hourly data, compute the maximum-temperature-based indices, calculate the mean and trend for the historical period (1971-2000) and cache the result (to avoid redundant downloads and processing).\n\n(section-2.2)=", "text_with_prefix": "EQC Quality Assessment: \"Bias in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-and-weather-extremes_q01 | Category: Climate_Projections\nSection: Bias in extreme temperature indices for the reinsurance sector > Analysis and results > 2. Downloading and processing > 2.1. Download and transform ERA5\n---\nIn this section, the `download.download_and_transform` function from the 'c3s_eqc_automatic_quality_control' package is employed to download ERA5 reference data, select the temporal aggregation (\"JJA\" in our example), compute daily maximum temperature from hourly data, compute the maximum-temperature-based indices, calculate the mean and trend for the historical period (1971-2000) and cache the result (to avoid redundant downloads and processing).\n\n(section-2.2)="} {"chunk_id": "climate_projections-cmip6_climate-and-weather-extremes_q01__619513fc3751", "report_id": "climate_projections-cmip6_climate-and-weather-extremes_q01", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q01", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Bias in extreme temperature indices for the reinsurance sector > Analysis and results > 2. Downloading and processing > 2.2. Download and transform models", "title": "Bias in extreme temperature indices for the reinsurance sector", "chunk_index": 10, "token_count": 348, "text_raw": "In this section, the `download.download_and_transform` function from the 'c3s_eqc_automatic_quality_control' package is employed to download daily data from the CMIP6 models, compute the maximum-temperature-based indices for the selected temporal aggregation, calculate the mean and trend over the historical period (1971-2000), interpolate to ERA5's grid (only for the cases in which it is specified, in the other cases, the original model's grid is mantained), and cache the result (to avoid redundant downloads and processing).\n\nOriginal model\nInterpolated model\n\n```text\nmodel='access_cm2'\nmodel='awi_cm_1_1_mr'\nmodel='cmcc_esm2'\nmodel='cnrm_cm6_1_hr'\nmodel='cnrm_esm2_1'\nmodel='ec_earth3_cc'\nmodel='gfdl_esm4'\nmodel='inm_cm4_8'\nmodel='inm_cm5_0'\nmodel='kiost_esm'\nmodel='mpi_esm1_2_lr'\nmodel='miroc6'\nmodel='miroc_es2l'\nmodel='mri_esm2_0'\nmodel='noresm2_mm'\nmodel='nesm3'\n```\n\n(section-2.3)=", "text_with_prefix": "EQC Quality Assessment: \"Bias in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-and-weather-extremes_q01 | Category: Climate_Projections\nSection: Bias in extreme temperature indices for the reinsurance sector > Analysis and results > 2. Downloading and processing > 2.2. Download and transform models\n---\nIn this section, the `download.download_and_transform` function from the 'c3s_eqc_automatic_quality_control' package is employed to download daily data from the CMIP6 models, compute the maximum-temperature-based indices for the selected temporal aggregation, calculate the mean and trend over the historical period (1971-2000), interpolate to ERA5's grid (only for the cases in which it is specified, in the other cases, the original model's grid is mantained), and cache the result (to avoid redundant downloads and processing).\n\nOriginal model\nInterpolated model\n\n```text\nmodel='access_cm2'\nmodel='awi_cm_1_1_mr'\nmodel='cmcc_esm2'\nmodel='cnrm_cm6_1_hr'\nmodel='cnrm_esm2_1'\nmodel='ec_earth3_cc'\nmodel='gfdl_esm4'\nmodel='inm_cm4_8'\nmodel='inm_cm5_0'\nmodel='kiost_esm'\nmodel='mpi_esm1_2_lr'\nmodel='miroc6'\nmodel='miroc_es2l'\nmodel='mri_esm2_0'\nmodel='noresm2_mm'\nmodel='nesm3'\n```\n\n(section-2.3)="} {"chunk_id": "climate_projections-cmip6_climate-and-weather-extremes_q01__7f450810a8f3", "report_id": "climate_projections-cmip6_climate-and-weather-extremes_q01", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q01", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Bias in extreme temperature indices for the reinsurance sector > Analysis and results > 2. Downloading and processing > 2.3. Apply land-sea mask, change attributes and cut the region to show", "title": "Bias in extreme temperature indices for the reinsurance sector", "chunk_index": 11, "token_count": 260, "text_raw": "This section performs the following tasks:\n\n1. Cut the region of interest.\n2. Downloads the sea mask for ERA5.\n3. Applies the sea mask to both ERA5 data and the model data, which were previously regridded to ERA5's grid (i.e., it applies the ERA5 sea mask to `ds_interpolated`).\n4. Regrids the ERA5 land-sea mask to the model's grid and applies it to them.\n5. Change some variable attributes for plotting purposes.\n\n**Note:** `ds_interpolated` contains data from the models (mean and trend over the historical period, p-value of the trends...) regridded to ERA5. `model_datasets` contain the same data but in the original grid of each model.\n\nEdit attributes\n\n(section-3)=", "text_with_prefix": "EQC Quality Assessment: \"Bias in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-and-weather-extremes_q01 | Category: Climate_Projections\nSection: Bias in extreme temperature indices for the reinsurance sector > Analysis and results > 2. Downloading and processing > 2.3. Apply land-sea mask, change attributes and cut the region to show\n---\nThis section performs the following tasks:\n\n1. Cut the region of interest.\n2. Downloads the sea mask for ERA5.\n3. Applies the sea mask to both ERA5 data and the model data, which were previously regridded to ERA5's grid (i.e., it applies the ERA5 sea mask to `ds_interpolated`).\n4. Regrids the ERA5 land-sea mask to the model's grid and applies it to them.\n5. Change some variable attributes for plotting purposes.\n\n**Note:** `ds_interpolated` contains data from the models (mean and trend over the historical period, p-value of the trends...) regridded to ERA5. `model_datasets` contain the same data but in the original grid of each model.\n\nEdit attributes\n\n(section-3)="} {"chunk_id": "climate_projections-cmip6_climate-and-weather-extremes_q01__4d067af1225e", "report_id": "climate_projections-cmip6_climate-and-weather-extremes_q01", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q01", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Bias in extreme temperature indices for the reinsurance sector > Analysis and results > 3. Plot and describe results", "title": "Bias in extreme temperature indices for the reinsurance sector", "chunk_index": 12, "token_count": 312, "text_raw": "This section will display the following results:\n\n- Maps representing the spatial distribution of the **historical mean values** (1971-2000) of the 'SU' index for ERA5, each model individually, the ensemble median (understood as the median of the mean values of the chosen subset of models calculated for each grid cell), and the ensemble spread (derived as the standard deviation of the distribution of the chosen subset of models).\n- Maps representing the spatial distribution of the **historical trends** (1971-2000) of the indices 'SU' and 'TX90p'. Similar to the first analysis, this includes ERA5, each model individually, the ensemble median (understood as the median of the trend values of the chosen subset of models calculated for each grid cell), and the ensemble spread (derived as the standard deviation of the distribution of the chosen subset of models).\n- **Bias maps of the historical mean values**.\n- **Trend bias maps**. \n- **Boxplots** representing statistical distributions (PDF) built on the spatially-averaged historical trends from each considered model, displayed together with ERA5.\n\n(section-3.1)=", "text_with_prefix": "EQC Quality Assessment: \"Bias in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-and-weather-extremes_q01 | Category: Climate_Projections\nSection: Bias in extreme temperature indices for the reinsurance sector > Analysis and results > 3. Plot and describe results\n---\nThis section will display the following results:\n\n- Maps representing the spatial distribution of the **historical mean values** (1971-2000) of the 'SU' index for ERA5, each model individually, the ensemble median (understood as the median of the mean values of the chosen subset of models calculated for each grid cell), and the ensemble spread (derived as the standard deviation of the distribution of the chosen subset of models).\n- Maps representing the spatial distribution of the **historical trends** (1971-2000) of the indices 'SU' and 'TX90p'. Similar to the first analysis, this includes ERA5, each model individually, the ensemble median (understood as the median of the trend values of the chosen subset of models calculated for each grid cell), and the ensemble spread (derived as the standard deviation of the distribution of the chosen subset of models).\n- **Bias maps of the historical mean values**.\n- **Trend bias maps**. \n- **Boxplots** representing statistical distributions (PDF) built on the spatially-averaged historical trends from each considered model, displayed together with ERA5.\n\n(section-3.1)="} {"chunk_id": "climate_projections-cmip6_climate-and-weather-extremes_q01__c36147a2c919", "report_id": "climate_projections-cmip6_climate-and-weather-extremes_q01", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q01", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Bias in extreme temperature indices for the reinsurance sector > Analysis and results > 3. Plot and describe results > 3.1. Define plotting functions", "title": "Bias in extreme temperature indices for the reinsurance sector", "chunk_index": 13, "token_count": 651, "text_raw": "The functions presented here are used to plot the mean values and trends calculated over the historical period (1971-2000) for each of the indices ('SU' and 'TX90p').\n\nFor a selected index, three layout types can be displayed, depending on the chosen function:\n\n1. Layout including the reference ERA5 product, the ensemble median, the bias of the ensemble median, and the ensemble spread: `plot_ensemble()` is used.\n2. Layout including every model (for the trend and mean values): `plot_models()` is employed.\n3. Layout including the bias of every model (for the trend and mean values): `plot_models()` is used again.\n\n`trend==True` argument allows displaying trend values over the historical period, while `trend==False` will show mean values. When the `trend` argument is set to `True`, regions with no significance are hatched. For individual models and ERA5, a grid point is considered to have a statistically significant trend when the p-value is lower than 0.05 (in such cases, no hatching is shown). However, for determining trend significance for the ensemble median (understood as the median of the trend values of the chosen subset of models calculated for each grid cell), reliance is placed on agreement categories, following the advanced approach proposed in AR6 [IPCC](https://www.ipcc.ch/report/ar6/wg1/downloads/report/IPCC_AR6_WGI_Atlas.pdf) on pages 1945-1950. The `hatch_p_value_ensemble()` function is used to distinguish, for each grid point, between three possible cases:\n\n1. If more than 66% of the models are statistically significant (p-value < 0.05) and more than 80% of the models share the same sign, we consider the ensemble median trend to be statistically significant, and there is agreement on the sign. To represent this, no hatching is used.\n2. If less than 66% of the models are statistically significant, regardless of agreement on the sign of the trend, hatching is applied (indicating that the ensemble median trend is not statistically significant).\n3. If more than 66% of the models are statistically significant but less than 80% of the models share the same sign, we consider the ensemble median trend to be statistically significant, but there is no agreement on the sign of the trend. This is represented using crosses.\n\nDefine function to plot the caption of the figures (for the ensemble case)\nAdd caption to the figure\nAdd each line to the figure\nend captioning\nDefine function to plot the caption of the figures (for the individual models case)\nAdd caption to the figure\nAdd each line to the figure\n\n(section-3.2)=", "text_with_prefix": "EQC Quality Assessment: \"Bias in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-and-weather-extremes_q01 | Category: Climate_Projections\nSection: Bias in extreme temperature indices for the reinsurance sector > Analysis and results > 3. Plot and describe results > 3.1. Define plotting functions\n---\nThe functions presented here are used to plot the mean values and trends calculated over the historical period (1971-2000) for each of the indices ('SU' and 'TX90p').\n\nFor a selected index, three layout types can be displayed, depending on the chosen function:\n\n1. Layout including the reference ERA5 product, the ensemble median, the bias of the ensemble median, and the ensemble spread: `plot_ensemble()` is used.\n2. Layout including every model (for the trend and mean values): `plot_models()` is employed.\n3. Layout including the bias of every model (for the trend and mean values): `plot_models()` is used again.\n\n`trend==True` argument allows displaying trend values over the historical period, while `trend==False` will show mean values. When the `trend` argument is set to `True`, regions with no significance are hatched. For individual models and ERA5, a grid point is considered to have a statistically significant trend when the p-value is lower than 0.05 (in such cases, no hatching is shown). However, for determining trend significance for the ensemble median (understood as the median of the trend values of the chosen subset of models calculated for each grid cell), reliance is placed on agreement categories, following the advanced approach proposed in AR6 [IPCC](https://www.ipcc.ch/report/ar6/wg1/downloads/report/IPCC_AR6_WGI_Atlas.pdf) on pages 1945-1950. The `hatch_p_value_ensemble()` function is used to distinguish, for each grid point, between three possible cases:\n\n1. If more than 66% of the models are statistically significant (p-value < 0.05) and more than 80% of the models share the same sign, we consider the ensemble median trend to be statistically significant, and there is agreement on the sign. To represent this, no hatching is used.\n2. If less than 66% of the models are statistically significant, regardless of agreement on the sign of the trend, hatching is applied (indicating that the ensemble median trend is not statistically significant).\n3. If more than 66% of the models are statistically significant but less than 80% of the models share the same sign, we consider the ensemble median trend to be statistically significant, but there is no agreement on the sign of the trend. This is represented using crosses.\n\nDefine function to plot the caption of the figures (for the ensemble case)\nAdd caption to the figure\nAdd each line to the figure\nend captioning\nDefine function to plot the caption of the figures (for the individual models case)\nAdd caption to the figure\nAdd each line to the figure\n\n(section-3.2)="} {"chunk_id": "climate_projections-cmip6_climate-and-weather-extremes_q01__445414a70dfe", "report_id": "climate_projections-cmip6_climate-and-weather-extremes_q01", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q01", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Bias in extreme temperature indices for the reinsurance sector > Analysis and results > 3. Plot and describe results > 3.2. Plot ensemble maps", "title": "Bias in extreme temperature indices for the reinsurance sector", "chunk_index": 14, "token_count": 370, "text_raw": "In this section, we invoke the `plot_ensemble()` function to visualise the mean values and trends calculated over the historical period (1971-2000) for the model ensemble and ERA5 reference product across Europe. Note that the model data used in this section has previously been interpolated to the ERA5 grid.\n\nSpecifically, for each of the indices ('SU' and 'TX90p'), this section presents two layouts:\n\n1. Only for the 'SU' index: mean values of the historical period (1971-2000) for: (a) the reference ERA5 product, (b) the ensemble median (understood as the median of the mean values of the chosen subset of models calculated for each grid cell), (c) the bias of the ensemble median, and (d) the ensemble spread (derived as the standard deviation of the distribution of the chosen subset of models).\n\n2. Trend values of the historical period (1971-2000) for: (a) the reference ERA5 product, (b) the ensemble median (understood as the median of the trend values of the chosen subset of models calculated for each grid cell), (c) the bias of the ensemble median, and (d) the ensemble spread (derived as the standard deviation of the distribution of the chosen subset of models).\n\nChange default colorbars\nFig number counter\nCommon title\n\n(section-3.3)=", "text_with_prefix": "EQC Quality Assessment: \"Bias in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-and-weather-extremes_q01 | Category: Climate_Projections\nSection: Bias in extreme temperature indices for the reinsurance sector > Analysis and results > 3. Plot and describe results > 3.2. Plot ensemble maps\n---\nIn this section, we invoke the `plot_ensemble()` function to visualise the mean values and trends calculated over the historical period (1971-2000) for the model ensemble and ERA5 reference product across Europe. Note that the model data used in this section has previously been interpolated to the ERA5 grid.\n\nSpecifically, for each of the indices ('SU' and 'TX90p'), this section presents two layouts:\n\n1. Only for the 'SU' index: mean values of the historical period (1971-2000) for: (a) the reference ERA5 product, (b) the ensemble median (understood as the median of the mean values of the chosen subset of models calculated for each grid cell), (c) the bias of the ensemble median, and (d) the ensemble spread (derived as the standard deviation of the distribution of the chosen subset of models).\n\n2. Trend values of the historical period (1971-2000) for: (a) the reference ERA5 product, (b) the ensemble median (understood as the median of the trend values of the chosen subset of models calculated for each grid cell), (c) the bias of the ensemble median, and (d) the ensemble spread (derived as the standard deviation of the distribution of the chosen subset of models).\n\nChange default colorbars\nFig number counter\nCommon title\n\n(section-3.3)="} {"chunk_id": "climate_projections-cmip6_climate-and-weather-extremes_q01__600ae7a1d68e", "report_id": "climate_projections-cmip6_climate-and-weather-extremes_q01", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q01", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Bias in extreme temperature indices for the reinsurance sector > Analysis and results > 3. Plot and describe results > 3.3. Plot model maps", "title": "Bias in extreme temperature indices for the reinsurance sector", "chunk_index": 15, "token_count": 211, "text_raw": "In this section, we invoke the `plot_models()` function to visualise the mean values and trends calculated over the historical period (1971-2000) for every model individually across Europe. Note that the model data used in this section maintains its original grid.\n\nSpecifically, for each of the indices ('SU' and 'TX90p'), this section presents two layouts:\n\n1. A layout including the historical mean (1971-2000) of every model (only for the 'SU' index).\n\n2. A layout including the historical trend (1971-2000) of every model.\n\n(section-3.4)=", "text_with_prefix": "EQC Quality Assessment: \"Bias in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-and-weather-extremes_q01 | Category: Climate_Projections\nSection: Bias in extreme temperature indices for the reinsurance sector > Analysis and results > 3. Plot and describe results > 3.3. Plot model maps\n---\nIn this section, we invoke the `plot_models()` function to visualise the mean values and trends calculated over the historical period (1971-2000) for every model individually across Europe. Note that the model data used in this section maintains its original grid.\n\nSpecifically, for each of the indices ('SU' and 'TX90p'), this section presents two layouts:\n\n1. A layout including the historical mean (1971-2000) of every model (only for the 'SU' index).\n\n2. A layout including the historical trend (1971-2000) of every model.\n\n(section-3.4)="} {"chunk_id": "climate_projections-cmip6_climate-and-weather-extremes_q01__15f2db66d3bb", "report_id": "climate_projections-cmip6_climate-and-weather-extremes_q01", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q01", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Bias in extreme temperature indices for the reinsurance sector > Analysis and results > 3. Plot and describe results > 3.4. Plot bias maps", "title": "Bias in extreme temperature indices for the reinsurance sector", "chunk_index": 16, "token_count": 225, "text_raw": "In this section, we invoke the `plot_models()` function to visualise the bias for the mean values and trends calculated over the historical period (1971-2000) for every model individually across Europe. Note that the model data used in this section has previously been interpolated to the ERA5 grid.\n\nSpecifically, for each of the indices ('SU' and 'TX90p'), this section presents two layouts:\n\n1. A layout including the bias for the historical mean (1971-2000) of every model (only for the 'SU' index).\n\n2. A layout including the bias for the historical trend (1971-2000) of every model.\n\n(section-3.5)=", "text_with_prefix": "EQC Quality Assessment: \"Bias in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-and-weather-extremes_q01 | Category: Climate_Projections\nSection: Bias in extreme temperature indices for the reinsurance sector > Analysis and results > 3. Plot and describe results > 3.4. Plot bias maps\n---\nIn this section, we invoke the `plot_models()` function to visualise the bias for the mean values and trends calculated over the historical period (1971-2000) for every model individually across Europe. Note that the model data used in this section has previously been interpolated to the ERA5 grid.\n\nSpecifically, for each of the indices ('SU' and 'TX90p'), this section presents two layouts:\n\n1. A layout including the bias for the historical mean (1971-2000) of every model (only for the 'SU' index).\n\n2. A layout including the bias for the historical trend (1971-2000) of every model.\n\n(section-3.5)="} {"chunk_id": "climate_projections-cmip6_climate-and-weather-extremes_q01__fb2ae07ab93e", "report_id": "climate_projections-cmip6_climate-and-weather-extremes_q01", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q01", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Bias in extreme temperature indices for the reinsurance sector > Analysis and results > 3. Plot and describe results > 3.5. Boxplots of the historical trend", "title": "Bias in extreme temperature indices for the reinsurance sector", "chunk_index": 17, "token_count": 391, "text_raw": "In this last section, we compare the trends of the climate models with the reference trend from ERA5.\n\nDots represent the spatially-averaged historical trend over the selected region (change of the number of days per decade) for each model (grey), the ensemble mean (blue), and the reference product (orange). The ensemble median is shown as a green line. Note that the spatially averaged values are calculated for each model from its original grid (i.e., no interpolated data has been used here).\n\nThe boxplot visually illustrates the distribution of trends among the climate models, with the box covering the first quartile (Q1 = 25th percentile) to the third quartile (Q3 = 75th percentile), and a green line indicating the ensemble median (Q2 = 50th percentile). Whiskers extend from the edges of the box to show the full data range.\n\nEnsemble mean\n\n
\n
\n

Fig 10. Boxplots illustrating the historical trends of the distribution of the chosen subset of models and ERA5 for: (a) the 'SU' index and (b) the 'TX90p' index. The distribution is created by considering spatially averaged trends across Europe. The ensemble mean and the ensemble median trends are both included. Outliers in the distribution are denoted by a grey circle with a black contour.

\n
\n\n(section-3.6)=", "text_with_prefix": "EQC Quality Assessment: \"Bias in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-and-weather-extremes_q01 | Category: Climate_Projections\nSection: Bias in extreme temperature indices for the reinsurance sector > Analysis and results > 3. Plot and describe results > 3.5. Boxplots of the historical trend\n---\nIn this last section, we compare the trends of the climate models with the reference trend from ERA5.\n\nDots represent the spatially-averaged historical trend over the selected region (change of the number of days per decade) for each model (grey), the ensemble mean (blue), and the reference product (orange). The ensemble median is shown as a green line. Note that the spatially averaged values are calculated for each model from its original grid (i.e., no interpolated data has been used here).\n\nThe boxplot visually illustrates the distribution of trends among the climate models, with the box covering the first quartile (Q1 = 25th percentile) to the third quartile (Q3 = 75th percentile), and a green line indicating the ensemble median (Q2 = 50th percentile). Whiskers extend from the edges of the box to show the full data range.\n\nEnsemble mean\n\n
\n
\n

Fig 10. Boxplots illustrating the historical trends of the distribution of the chosen subset of models and ERA5 for: (a) the 'SU' index and (b) the 'TX90p' index. The distribution is created by considering spatially averaged trends across Europe. The ensemble mean and the ensemble median trends are both included. Outliers in the distribution are denoted by a grey circle with a black contour.

\n
\n\n(section-3.6)="} {"chunk_id": "climate_projections-cmip6_climate-and-weather-extremes_q01__13b154ade7c5", "report_id": "climate_projections-cmip6_climate-and-weather-extremes_q01", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q01", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Bias in extreme temperature indices for the reinsurance sector > Analysis and results > 3. Plot and describe results > 3.6. Results summary and discussion", "title": "Bias in extreme temperature indices for the reinsurance sector", "chunk_index": 18, "token_count": 737, "text_raw": "- The selected subset of CMIP6 Global Climate Models (GCMs) exhibit biases in capturing both mean values and trends for the 'number of summer days' (SU) and 'the number of days with daily maximum temperature exceeding the daily 90th percentile' (TX90p) indices. These indices are calculated over the temporal aggregation of JJA and for the historical period spanning from 1971 to 2000.\n\n- GCMs struggle to accurately represent the climatology of the 'SU' index in regions with complex orography. Furthermore, underestimating the magnitude of this index is common, particularly in continental areas.\n\n- While the mean climate bias is mainly affected by orography, the regional variation in the trend bias is broader. ERA5 shows that the magnitude of the historical trend varies spatially. The trend bias assessment suggests that, overall, the trend is underestimated, with slight overestimations observed in certain regions. Hence, users need to consider this regional variation in both the trend and its bias.\n\n- Boxplots reveal a consistent positive spatial-averaged historical trend characterising the distribution of the chosen subset of models. Magnitudes, however, are slightly underestimated when compared to ERA5. For the 'SU' index, the CMIP6 ensemble median displays approximately 1.0 days/10 years, contrasting with ERA5's ~2 days/10 years. The 'TX90p' index reveals around 1.4 days/10 years for the CMIP6 ensemble median in contrast to the ~2 days/10 years observed in ERA5. The interquantile range of the ensemble ranges from 0.9 to 1.2 days per decade for the 'SU' index (with outliers closer to 1.9 days per decade) and from 1.1 to more than 1.7 days per decade for the 'TX90p' index.\n\n- What do the results mean for users? Are the biases relevant?\n\n- These results indicate that CMIP6 projections of maximum temperature indices can be useful in applications like designing strategies to optimise reinsurance protections. The findings increase confidence (though they do not ensure accuracy) when analysing future trends using these models.\n\n- However, users must recognise that all climate projections inherently include certain biases. This is particularly important when using projections of maximum temperature indices in areas with complex orography or large continental regions. In such cases, the biases may be more pronounced and could impact the accuracy of the projections. Therefore, users should exercise caution and consider these limitations when applying the projections to their specific use cases, recognising that biases in the mean climate and trends affect indices differently and that the choice of bias correction method should be tailored accordingly [[6]](https://doi.org/10.1002/asl.1072)[[7]](https://ibicus.readthedocs.io/en/latest/).\n\n\n
\nRESULTS NOTE:
\nIt is important to note that the results presented are specific to the 16 models chosen, and users should aim to assess as wide a range of models as possible before making a sub-selection.", "text_with_prefix": "EQC Quality Assessment: \"Bias in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-and-weather-extremes_q01 | Category: Climate_Projections\nSection: Bias in extreme temperature indices for the reinsurance sector > Analysis and results > 3. Plot and describe results > 3.6. Results summary and discussion\n---\n- The selected subset of CMIP6 Global Climate Models (GCMs) exhibit biases in capturing both mean values and trends for the 'number of summer days' (SU) and 'the number of days with daily maximum temperature exceeding the daily 90th percentile' (TX90p) indices. These indices are calculated over the temporal aggregation of JJA and for the historical period spanning from 1971 to 2000.\n\n- GCMs struggle to accurately represent the climatology of the 'SU' index in regions with complex orography. Furthermore, underestimating the magnitude of this index is common, particularly in continental areas.\n\n- While the mean climate bias is mainly affected by orography, the regional variation in the trend bias is broader. ERA5 shows that the magnitude of the historical trend varies spatially. The trend bias assessment suggests that, overall, the trend is underestimated, with slight overestimations observed in certain regions. Hence, users need to consider this regional variation in both the trend and its bias.\n\n- Boxplots reveal a consistent positive spatial-averaged historical trend characterising the distribution of the chosen subset of models. Magnitudes, however, are slightly underestimated when compared to ERA5. For the 'SU' index, the CMIP6 ensemble median displays approximately 1.0 days/10 years, contrasting with ERA5's ~2 days/10 years. The 'TX90p' index reveals around 1.4 days/10 years for the CMIP6 ensemble median in contrast to the ~2 days/10 years observed in ERA5. The interquantile range of the ensemble ranges from 0.9 to 1.2 days per decade for the 'SU' index (with outliers closer to 1.9 days per decade) and from 1.1 to more than 1.7 days per decade for the 'TX90p' index.\n\n- What do the results mean for users? Are the biases relevant?\n\n- These results indicate that CMIP6 projections of maximum temperature indices can be useful in applications like designing strategies to optimise reinsurance protections. The findings increase confidence (though they do not ensure accuracy) when analysing future trends using these models.\n\n- However, users must recognise that all climate projections inherently include certain biases. This is particularly important when using projections of maximum temperature indices in areas with complex orography or large continental regions. In such cases, the biases may be more pronounced and could impact the accuracy of the projections. Therefore, users should exercise caution and consider these limitations when applying the projections to their specific use cases, recognising that biases in the mean climate and trends affect indices differently and that the choice of bias correction method should be tailored accordingly [[6]](https://doi.org/10.1002/asl.1072)[[7]](https://ibicus.readthedocs.io/en/latest/).\n\n\n
\nRESULTS NOTE:
\nIt is important to note that the results presented are specific to the 16 models chosen, and users should aim to assess as wide a range of models as possible before making a sub-selection."} {"chunk_id": "climate_projections-cmip6_climate-and-weather-extremes_q01__4079d0943790", "report_id": "climate_projections-cmip6_climate-and-weather-extremes_q01", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q01", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Bias in extreme temperature indices for the reinsurance sector > ℹ️ If you want to know more > Key resources", "title": "Bias in extreme temperature indices for the reinsurance sector", "chunk_index": 19, "token_count": 270, "text_raw": "Some key resources and further reading were linked throughout this assessment.\n\nThe CDS catalogue entries for the data used were:\n* CMIP6 climate projections (Daily - Daily maximum near-surface air temperature): https://cds.climate.copernicus.eu/datasets/projections-cmip6?tab=overview\n* ERA5 hourly data on single levels from 1940 to present (2m temperature): https://cds.climate.copernicus.eu/datasets/reanalysis-era5-single-levels?tab=overview\n\nCode libraries used:\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)\n* [icclim](https://icclim.readthedocs.io/en/stable/) Python package", "text_with_prefix": "EQC Quality Assessment: \"Bias in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-and-weather-extremes_q01 | Category: Climate_Projections\nSection: Bias in extreme temperature indices for the reinsurance sector > ℹ️ If you want to know more > Key resources\n---\nSome key resources and further reading were linked throughout this assessment.\n\nThe CDS catalogue entries for the data used were:\n* CMIP6 climate projections (Daily - Daily maximum near-surface air temperature): https://cds.climate.copernicus.eu/datasets/projections-cmip6?tab=overview\n* ERA5 hourly data on single levels from 1940 to present (2m temperature): https://cds.climate.copernicus.eu/datasets/reanalysis-era5-single-levels?tab=overview\n\nCode libraries used:\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)\n* [icclim](https://icclim.readthedocs.io/en/stable/) Python package"} {"chunk_id": "climate_projections-cmip6_climate-and-weather-extremes_q01__b67579670c14", "report_id": "climate_projections-cmip6_climate-and-weather-extremes_q01", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q01", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Bias in extreme temperature indices for the reinsurance sector > ℹ️ If you want to know more > References", "title": "Bias in extreme temperature indices for the reinsurance sector", "chunk_index": 20, "token_count": 791, "text_raw": "[[1]](https://doi.org/10.3390/atmos11020146) Tesselaar, M., Botzen, W.J.W., Aerts, J.C.J.H. (2020). Impacts of Climate Change and Remote Natural Catastrophes on EU Flood Insurance Markets: An Analysis of Soft and Hard Reinsurance Markets for Flood Coverage. Atmosphere 2020, 11, 146. https://doi.org/10.3390/atmos11020146\n\n[[2]](https://doi.org/10.5194/nhess-22-659-2022) Rädler, A. T. (2022). Invited perspectives: how does climate change affect the risk of natural hazards? Challenges and step changes from the reinsurance perspective. Nat. Hazards Earth Syst. Sci., 22, 659–664. https://doi.org/10.5194/nhess-22-659-2022\n\n[[3]](https://www.ipcc.ch/site/assets/uploads/2018/03/wg2TARchap8.pdf) Vellinga, P., Mills, E., Bowers, L., Berz, G.A., Huq, S., Kozak, L.M., Paultikof, J., Schanzenbacker, B., Shida, S., Soler, G., Benson, C., Bidan, P., Bruce, J.W., Huyck, P.M., Lemcke, G., Peara, A., Radevsky, R., Schoubroeck, C.V., Dlugolecki, A.F. (2001). Insurance and other financial services. In J. J. McCarthy, O. F. Canziani, N. A. Leary, D. J. Dokken, & K. S. White (Eds.), Climate change 2001: impacts, adaptation, and vulnerability. Contribution of working group 2 to the third assessment report of the intergovernmental panel on climate change. (pp. 417-450). Cambridge University Press.\n\n[[4]](https://doi.org/10.1002/wcc.71) Lemos, M.C. and Rood, R.B. (2010). Climate projections and their impact on policy and practice. WIREs Clim Chg, 1: 670-682. https://doi.org/10.1002/wcc.71\n\n[[5]](https://doi.org/10.1002/wcc.579) Nissan, H., Goddard, L., de Perez, E.C., et al. (2019). On the use and misuse of climate change projections in international development. WIREs Clim Change, 10:e579. https://doi.org/10.1002/wcc.579\n\n[[6]](https://doi.org/10.1002/asl.1072) Iturbide, M., Casanueva, A., Bedia, J., Herrera, S., Milovac, J., Gutiérrez, J.M. (2021). On the need of bias adjustment for more plausible climate change projections of extreme heat. Atmos. Sci. Lett., 23, e1072. https://doi.org/10.1002/asl.1072\n\n[[7]](https://ibicus.readthedocs.io/en/latest/) https://ibicus.readthedocs.io/en/latest/", "text_with_prefix": "EQC Quality Assessment: \"Bias in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-and-weather-extremes_q01 | Category: Climate_Projections\nSection: Bias in extreme temperature indices for the reinsurance sector > ℹ️ If you want to know more > References\n---\n[[1]](https://doi.org/10.3390/atmos11020146) Tesselaar, M., Botzen, W.J.W., Aerts, J.C.J.H. (2020). Impacts of Climate Change and Remote Natural Catastrophes on EU Flood Insurance Markets: An Analysis of Soft and Hard Reinsurance Markets for Flood Coverage. Atmosphere 2020, 11, 146. https://doi.org/10.3390/atmos11020146\n\n[[2]](https://doi.org/10.5194/nhess-22-659-2022) Rädler, A. T. (2022). Invited perspectives: how does climate change affect the risk of natural hazards? Challenges and step changes from the reinsurance perspective. Nat. Hazards Earth Syst. Sci., 22, 659–664. https://doi.org/10.5194/nhess-22-659-2022\n\n[[3]](https://www.ipcc.ch/site/assets/uploads/2018/03/wg2TARchap8.pdf) Vellinga, P., Mills, E., Bowers, L., Berz, G.A., Huq, S., Kozak, L.M., Paultikof, J., Schanzenbacker, B., Shida, S., Soler, G., Benson, C., Bidan, P., Bruce, J.W., Huyck, P.M., Lemcke, G., Peara, A., Radevsky, R., Schoubroeck, C.V., Dlugolecki, A.F. (2001). Insurance and other financial services. In J. J. McCarthy, O. F. Canziani, N. A. Leary, D. J. Dokken, & K. S. White (Eds.), Climate change 2001: impacts, adaptation, and vulnerability. Contribution of working group 2 to the third assessment report of the intergovernmental panel on climate change. (pp. 417-450). Cambridge University Press.\n\n[[4]](https://doi.org/10.1002/wcc.71) Lemos, M.C. and Rood, R.B. (2010). Climate projections and their impact on policy and practice. WIREs Clim Chg, 1: 670-682. https://doi.org/10.1002/wcc.71\n\n[[5]](https://doi.org/10.1002/wcc.579) Nissan, H., Goddard, L., de Perez, E.C., et al. (2019). On the use and misuse of climate change projections in international development. WIREs Clim Change, 10:e579. https://doi.org/10.1002/wcc.579\n\n[[6]](https://doi.org/10.1002/asl.1072) Iturbide, M., Casanueva, A., Bedia, J., Herrera, S., Milovac, J., Gutiérrez, J.M. (2021). On the need of bias adjustment for more plausible climate change projections of extreme heat. Atmos. Sci. Lett., 23, e1072. https://doi.org/10.1002/asl.1072\n\n[[7]](https://ibicus.readthedocs.io/en/latest/) https://ibicus.readthedocs.io/en/latest/"} {"chunk_id": "climate_projections-cmip6_climate-and-weather-extremes_q02__db94c4617171", "report_id": "climate_projections-cmip6_climate-and-weather-extremes_q02", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q02", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector", "title": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector", "chunk_index": 0, "token_count": 104, "text_raw": "Production date: 24-05-2024\n\nProduced by: CMCC foundation - Euro-Mediterranean Center on Climate Change. Albert Martinez Boti.", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-and-weather-extremes_q02 | Category: Climate_Projections\nSection: Uncertainty in projected changes in extreme temperature indices for the reinsurance sector\n---\nProduction date: 24-05-2024\n\nProduced by: CMCC foundation - Euro-Mediterranean Center on Climate Change. Albert Martinez Boti."} {"chunk_id": "climate_projections-cmip6_climate-and-weather-extremes_q02__2b55dcdf8649", "report_id": "climate_projections-cmip6_climate-and-weather-extremes_q02", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q02", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector > Quality assessment question", "title": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector", "chunk_index": 1, "token_count": 671, "text_raw": "* **What are the projected future changes and associated uncertainties in air temperature extremes in Europe?**\n\nClimate change has a major impact on the reinsurance market [[1]](https://doi.org/10.3390/atmos11020146)[[2]](https://doi.org/10.5194/nhess-22-659-2022). In the third assessment report of the IPCC, hot temperature extremes were already presented as relevant to insurance and related services [[3]](https://www.ipcc.ch/site/assets/uploads/2018/03/wg2TARchap8.pdf). Consequently, the need for reliable regional and global climate projections has become paramount, offering valuable insights for optimising reinsurance strategies in the face of a changing climate landscape. Nonetheless, despite their pivotal role, uncertainties inherent in these projections can potentially lead to misuse [[4]](https://doi.org/10.1002/wcc.71)[[5]](https://doi.org/10.1002/wcc.579). This underscores the importance of accurately calculating and accounting for uncertainties to ensure their appropriate consideration. This notebook utilises data from a subset of models from [CMIP6](https://cds.climate.copernicus.eu/datasets/projections-cmip6?tab=overview) Global Climate Models (GCMs) and explores the uncertainty in future projections of maximum temperature-based extreme indices by considering the ensemble inter-model spread of projected changes. Two maximum temperature-based indices from [ECA&D](https://www.ecad.eu/indicesextremes/) indices (one of physical nature and the other of statistical nature) are computed using the [icclim](https://icclim.readthedocs.io/en/stable/) Python package. The first index, identified by the ETCCDI short name 'SU', quantifies the occurrence of summer days (i.e., with daily maximum temperatures exceeding 25°C) within a year or a season (JJA in this notebook). The second index, labeled 'TX90p', describes the number of days with daily maximum temperatures exceeding the daily 90th percentile of maximum temperature for a 5-day moving window. For this notebook, the daily 90th percentile threshold is calculated for the historical period spanning from 1971 to 2000. The index calculations, though, are performed over the future period from 2015 to 2099, following the Shared Socioeconomic Pathways SSP5-8.5. It is important to mention that the results presented here pertain to a specific subset of the CMIP6 ensemble and may not be generalisable to the entire dataset. Also note that a separate assessment examines the representation of climatology and trends of these indices for the same models during the historical period (1971-2000), while another assessment looks at the projected climate signal of these indices for the same models at a 2°C Global Warming Level.", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-and-weather-extremes_q02 | Category: Climate_Projections\nSection: Uncertainty in projected changes in extreme temperature indices for the reinsurance sector > Quality assessment question\n---\n* **What are the projected future changes and associated uncertainties in air temperature extremes in Europe?**\n\nClimate change has a major impact on the reinsurance market [[1]](https://doi.org/10.3390/atmos11020146)[[2]](https://doi.org/10.5194/nhess-22-659-2022). In the third assessment report of the IPCC, hot temperature extremes were already presented as relevant to insurance and related services [[3]](https://www.ipcc.ch/site/assets/uploads/2018/03/wg2TARchap8.pdf). Consequently, the need for reliable regional and global climate projections has become paramount, offering valuable insights for optimising reinsurance strategies in the face of a changing climate landscape. Nonetheless, despite their pivotal role, uncertainties inherent in these projections can potentially lead to misuse [[4]](https://doi.org/10.1002/wcc.71)[[5]](https://doi.org/10.1002/wcc.579). This underscores the importance of accurately calculating and accounting for uncertainties to ensure their appropriate consideration. This notebook utilises data from a subset of models from [CMIP6](https://cds.climate.copernicus.eu/datasets/projections-cmip6?tab=overview) Global Climate Models (GCMs) and explores the uncertainty in future projections of maximum temperature-based extreme indices by considering the ensemble inter-model spread of projected changes. Two maximum temperature-based indices from [ECA&D](https://www.ecad.eu/indicesextremes/) indices (one of physical nature and the other of statistical nature) are computed using the [icclim](https://icclim.readthedocs.io/en/stable/) Python package. The first index, identified by the ETCCDI short name 'SU', quantifies the occurrence of summer days (i.e., with daily maximum temperatures exceeding 25°C) within a year or a season (JJA in this notebook). The second index, labeled 'TX90p', describes the number of days with daily maximum temperatures exceeding the daily 90th percentile of maximum temperature for a 5-day moving window. For this notebook, the daily 90th percentile threshold is calculated for the historical period spanning from 1971 to 2000. The index calculations, though, are performed over the future period from 2015 to 2099, following the Shared Socioeconomic Pathways SSP5-8.5. It is important to mention that the results presented here pertain to a specific subset of the CMIP6 ensemble and may not be generalisable to the entire dataset. Also note that a separate assessment examines the representation of climatology and trends of these indices for the same models during the historical period (1971-2000), while another assessment looks at the projected climate signal of these indices for the same models at a 2°C Global Warming Level."} {"chunk_id": "climate_projections-cmip6_climate-and-weather-extremes_q02__fadbf163929c", "report_id": "climate_projections-cmip6_climate-and-weather-extremes_q02", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q02", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector > Quality assessment statement", "title": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector", "chunk_index": 2, "token_count": 588, "text_raw": "These are the key outcomes of this assessment\n\n* Looking ahead to future projections (2015-2099), all models within the subset agree on projecting general positive trends for both indices across Europe during the temporal aggregation of JJA, with a particularly notable positive projected trend in the Mediterranean Basin for 'TX90p'. This finding is consistent with the results of Josep Cos et al. (2022) [[6]](https://doi.org/10.5194/esd-13-321-2022), who evaluated the Mediterranean climate change hotspot using CMIP6 projections.\n\n* While certain regions exhibit near-zero trends for the 'SU' index, possibly due to threshold temperature constraints, others show higher values, highlighting the importance of considering both statistically and physically based extreme indices for comprehensive assessments.\n\n* Utilising CMIP6 projections presents a valuable opportunity to anticipate future trends in air temperature extremes across Europe, enabling the insurance industry to refine risk management strategies. While all considered models show a positive trend for these indices, the magnitude of these trends and their uncertainty (quantified by the inter-model spread) vary spatially and need to be considered.\n\n* A separate assessment evaluates the biases in climatology and trends of these indices for the historical period from 1971 to 2000 (\"CMIP6 Climate Projections: evaluating bias in extreme temperature indices for the reinsurance sector\"). The results of that assessment show an overall underestimation of the trends for both indices and the climatology of 'SU', as well as difficulty in correctly representing the spatial distribution, particularly over regions with complex orography. These biases may affect future projections and should be taken into account before using them.\n```\n\nattachment:00e7ccee-ce24-4448-a32d-d37b6cce9162.png\n---\nalt: trend_future_TX90p\nwidth: 850px\n---\nNumber of days with daily maximum temperatures exceeding the daily 90th percentile of maximum temperature for a 5-day moving window ('TX90p') for the temporal aggregation of 'JJA'. Trend for the future period (2015-2099). For this index, the reference daily 90th percentile threshold is calculated based on the historical period (1971-2000). The layout includes data corresponding to: (a) the ensemble median (understood as the median of the trend values of the chosen subset of models calculated for each grid cell) and (b) the ensemble spread (derived as the standard deviation of the distribution of the chosen subset of models).\n```", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-and-weather-extremes_q02 | Category: Climate_Projections\nSection: Uncertainty in projected changes in extreme temperature indices for the reinsurance sector > Quality assessment statement\n---\nThese are the key outcomes of this assessment\n\n* Looking ahead to future projections (2015-2099), all models within the subset agree on projecting general positive trends for both indices across Europe during the temporal aggregation of JJA, with a particularly notable positive projected trend in the Mediterranean Basin for 'TX90p'. This finding is consistent with the results of Josep Cos et al. (2022) [[6]](https://doi.org/10.5194/esd-13-321-2022), who evaluated the Mediterranean climate change hotspot using CMIP6 projections.\n\n* While certain regions exhibit near-zero trends for the 'SU' index, possibly due to threshold temperature constraints, others show higher values, highlighting the importance of considering both statistically and physically based extreme indices for comprehensive assessments.\n\n* Utilising CMIP6 projections presents a valuable opportunity to anticipate future trends in air temperature extremes across Europe, enabling the insurance industry to refine risk management strategies. While all considered models show a positive trend for these indices, the magnitude of these trends and their uncertainty (quantified by the inter-model spread) vary spatially and need to be considered.\n\n* A separate assessment evaluates the biases in climatology and trends of these indices for the historical period from 1971 to 2000 (\"CMIP6 Climate Projections: evaluating bias in extreme temperature indices for the reinsurance sector\"). The results of that assessment show an overall underestimation of the trends for both indices and the climatology of 'SU', as well as difficulty in correctly representing the spatial distribution, particularly over regions with complex orography. These biases may affect future projections and should be taken into account before using them.\n```\n\nattachment:00e7ccee-ce24-4448-a32d-d37b6cce9162.png\n---\nalt: trend_future_TX90p\nwidth: 850px\n---\nNumber of days with daily maximum temperatures exceeding the daily 90th percentile of maximum temperature for a 5-day moving window ('TX90p') for the temporal aggregation of 'JJA'. Trend for the future period (2015-2099). For this index, the reference daily 90th percentile threshold is calculated based on the historical period (1971-2000). The layout includes data corresponding to: (a) the ensemble median (understood as the median of the trend values of the chosen subset of models calculated for each grid cell) and (b) the ensemble spread (derived as the standard deviation of the distribution of the chosen subset of models).\n```"} {"chunk_id": "climate_projections-cmip6_climate-and-weather-extremes_q02__5111d54a08f5", "report_id": "climate_projections-cmip6_climate-and-weather-extremes_q02", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q02", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector > Methodology", "title": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector", "chunk_index": 3, "token_count": 397, "text_raw": "This notebook offers an assessment of the projected changes and their associated uncertainties using a subset of 16 models from [CMIP6](https://cds.climate.copernicus.eu/datasets/projections-cmip6?tab=overview). The uncertainty is examined by analysing the ensemble inter-model spread of projected changes for the maximum-temperature-based indices 'SU' and 'TX90p,' calculated over the temporal aggregation of JJA for the future period spanning from 2015 to 2099. In particular, spatial patterns of climate projected trends are examined and displayed for each model individually and for the ensemble median (calculated for each grid cell), alongside the ensemble inter-model spread to account for projected uncertainty. Additionally, spatially-averaged trend values are analysed and presented using box plots to provide an overview of trend behavior across the distribution of the chosen subset of models when averaged across Europe.\n\nThe analysis and results follow the next outline:\n\n**[](section-1)**\n * [](section-1.1)\n * [](section-1.2)\n * [](section-1.3)\n * [](section-1.4)\n * [](section-1.5)\n * [](section-1.6)\n\n**[](section-2)**\n * [](section-2.1)\n * [](section-2.2)\n * [](section-2.3)\n\n**[](section-3)**\n * [](section-3.1)\n * [](section-3.2)\n * [](section-3.3)\n * [](section-3.4)\n * [](section-3.5)", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-and-weather-extremes_q02 | Category: Climate_Projections\nSection: Uncertainty in projected changes in extreme temperature indices for the reinsurance sector > Methodology\n---\nThis notebook offers an assessment of the projected changes and their associated uncertainties using a subset of 16 models from [CMIP6](https://cds.climate.copernicus.eu/datasets/projections-cmip6?tab=overview). The uncertainty is examined by analysing the ensemble inter-model spread of projected changes for the maximum-temperature-based indices 'SU' and 'TX90p,' calculated over the temporal aggregation of JJA for the future period spanning from 2015 to 2099. In particular, spatial patterns of climate projected trends are examined and displayed for each model individually and for the ensemble median (calculated for each grid cell), alongside the ensemble inter-model spread to account for projected uncertainty. Additionally, spatially-averaged trend values are analysed and presented using box plots to provide an overview of trend behavior across the distribution of the chosen subset of models when averaged across Europe.\n\nThe analysis and results follow the next outline:\n\n**[](section-1)**\n * [](section-1.1)\n * [](section-1.2)\n * [](section-1.3)\n * [](section-1.4)\n * [](section-1.5)\n * [](section-1.6)\n\n**[](section-2)**\n * [](section-2.1)\n * [](section-2.2)\n * [](section-2.3)\n\n**[](section-3)**\n * [](section-3.1)\n * [](section-3.2)\n * [](section-3.3)\n * [](section-3.4)\n * [](section-3.5)"} {"chunk_id": "climate_projections-cmip6_climate-and-weather-extremes_q02__7090310dd7cb", "report_id": "climate_projections-cmip6_climate-and-weather-extremes_q02", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q02", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector > Analysis and results > 1. Parameters, requests and functions definition > 1.2. Define Parameters", "title": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector", "chunk_index": 4, "token_count": 408, "text_raw": "In the \"Define Parameters\" section, various customisable options for the notebook are specified. Most of the parameters chosen are the same as those used in other assessments (\"CMIP6 Climate Projections: evaluating bias in extreme temperature indices for the reinsurance sector\"), being them:\n- The initial and ending year used for the future projections period can be specified by changing the parametes `future_slice` (2015-2099 is chosen for consistency between CMIP6 and CORDEX).\n- `historical_slice` determines the historical period used (1971 to 2000 is choosen to allow comparison to CORDEX models in other assessments).\n- The `timeseries` set the temporal aggregation. For instance, selecting \"JJA\" implies considering only the JJA season.\n- `collection_id` provides the choice between Global Climate Models CMIP6 or Regional Climate Models CORDEX. Although the code allows choosing between CMIP6 or CORDEX, the example provided in this notebook deals with CMIP6.\n- `area` allows specifying the geographical domain of interest.\n- The `interpolation_method` parameter allows selecting the interpolation method when regridding is performed over the indices.\n- The `chunk` selection allows the user to define if dividing into chunks when downloading the data on their local machine. Although it does not significantly affect the analysis, it is recommended to keep the default value for optimal performance.\n\nChoose annual or seasonal timeseries\nChoose CORDEX or CMIP6\nInterpolation method\nArea to show\nChunks for download\n\n(section-1.3)=", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-and-weather-extremes_q02 | Category: Climate_Projections\nSection: Uncertainty in projected changes in extreme temperature indices for the reinsurance sector > Analysis and results > 1. Parameters, requests and functions definition > 1.2. Define Parameters\n---\nIn the \"Define Parameters\" section, various customisable options for the notebook are specified. Most of the parameters chosen are the same as those used in other assessments (\"CMIP6 Climate Projections: evaluating bias in extreme temperature indices for the reinsurance sector\"), being them:\n- The initial and ending year used for the future projections period can be specified by changing the parametes `future_slice` (2015-2099 is chosen for consistency between CMIP6 and CORDEX).\n- `historical_slice` determines the historical period used (1971 to 2000 is choosen to allow comparison to CORDEX models in other assessments).\n- The `timeseries` set the temporal aggregation. For instance, selecting \"JJA\" implies considering only the JJA season.\n- `collection_id` provides the choice between Global Climate Models CMIP6 or Regional Climate Models CORDEX. Although the code allows choosing between CMIP6 or CORDEX, the example provided in this notebook deals with CMIP6.\n- `area` allows specifying the geographical domain of interest.\n- The `interpolation_method` parameter allows selecting the interpolation method when regridding is performed over the indices.\n- The `chunk` selection allows the user to define if dividing into chunks when downloading the data on their local machine. Although it does not significantly affect the analysis, it is recommended to keep the default value for optimal performance.\n\nChoose annual or seasonal timeseries\nChoose CORDEX or CMIP6\nInterpolation method\nArea to show\nChunks for download\n\n(section-1.3)="} {"chunk_id": "climate_projections-cmip6_climate-and-weather-extremes_q02__969f9d7cb9be", "report_id": "climate_projections-cmip6_climate-and-weather-extremes_q02", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q02", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector > Analysis and results > 1. Parameters, requests and functions definition > 1.3. Define models", "title": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector", "chunk_index": 5, "token_count": 252, "text_raw": "The following climate analyses are performed considering a subset of GCMs from CMIP6. Models names are listed in the parameters below. Some variable-dependent parameters are also selected, as the `index_names` parameter, which specifies the maximum-temperature-based indices ('SU' and 'TX90p' in our case) from the [icclim](https://icclim.readthedocs.io/en/stable/) Python package.\n\nThe selected CMIP6 models have available both the historical and SSP8.5 experiments, and they are the same as those used in other assessments (\"CMIP6 Climate Projections: evaluating bias in extreme temperature indices for the reinsurance sector\").\n\nDefine dictionaries to use in titles and caption\nDefine dictionaries to use in titles and caption\n\n(section-1.4)=", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-and-weather-extremes_q02 | Category: Climate_Projections\nSection: Uncertainty in projected changes in extreme temperature indices for the reinsurance sector > Analysis and results > 1. Parameters, requests and functions definition > 1.3. Define models\n---\nThe following climate analyses are performed considering a subset of GCMs from CMIP6. Models names are listed in the parameters below. Some variable-dependent parameters are also selected, as the `index_names` parameter, which specifies the maximum-temperature-based indices ('SU' and 'TX90p' in our case) from the [icclim](https://icclim.readthedocs.io/en/stable/) Python package.\n\nThe selected CMIP6 models have available both the historical and SSP8.5 experiments, and they are the same as those used in other assessments (\"CMIP6 Climate Projections: evaluating bias in extreme temperature indices for the reinsurance sector\").\n\nDefine dictionaries to use in titles and caption\nDefine dictionaries to use in titles and caption\n\n(section-1.4)="} {"chunk_id": "climate_projections-cmip6_climate-and-weather-extremes_q02__297c3d340e6c", "report_id": "climate_projections-cmip6_climate-and-weather-extremes_q02", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q02", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector > Analysis and results > 1. Parameters, requests and functions definition > 1.4. Define land-sea mask request", "title": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector", "chunk_index": 6, "token_count": 144, "text_raw": "Within this notebook, ERA5 will be used to download the land-sea mask when plotting. In this section, we set the required parameters for the cds-api data-request of ERA5 land-sea mask.\n\n(section-1.5)=", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-and-weather-extremes_q02 | Category: Climate_Projections\nSection: Uncertainty in projected changes in extreme temperature indices for the reinsurance sector > Analysis and results > 1. Parameters, requests and functions definition > 1.4. Define land-sea mask request\n---\nWithin this notebook, ERA5 will be used to download the land-sea mask when plotting. In this section, we set the required parameters for the cds-api data-request of ERA5 land-sea mask.\n\n(section-1.5)="} {"chunk_id": "climate_projections-cmip6_climate-and-weather-extremes_q02__aa09b0c08e9a", "report_id": "climate_projections-cmip6_climate-and-weather-extremes_q02", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q02", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector > Analysis and results > 1. Parameters, requests and functions definition > 1.5. Define model requests", "title": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector", "chunk_index": 7, "token_count": 198, "text_raw": "In this section we set the required parameters for the cds-api data-request.\n\nWhen `weights = True`, spatial weighting is applied for calculations requiring spatial data aggregation. This is particularly relevant for CMIP6 GCMs with regular lon-lat grids that do not consider varying surface extensions at different latitudes. In contrast, CORDEX RCMs, using rotated grids, inherently account for different cell surfaces based on latitude, eliminating the need for a latitude cosine multiplicative factor (`weights = False`).\n\n(section-1.6)=", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-and-weather-extremes_q02 | Category: Climate_Projections\nSection: Uncertainty in projected changes in extreme temperature indices for the reinsurance sector > Analysis and results > 1. Parameters, requests and functions definition > 1.5. Define model requests\n---\nIn this section we set the required parameters for the cds-api data-request.\n\nWhen `weights = True`, spatial weighting is applied for calculations requiring spatial data aggregation. This is particularly relevant for CMIP6 GCMs with regular lon-lat grids that do not consider varying surface extensions at different latitudes. In contrast, CORDEX RCMs, using rotated grids, inherently account for different cell surfaces based on latitude, eliminating the need for a latitude cosine multiplicative factor (`weights = False`).\n\n(section-1.6)="} {"chunk_id": "climate_projections-cmip6_climate-and-weather-extremes_q02__e0c4e82aabcf", "report_id": "climate_projections-cmip6_climate-and-weather-extremes_q02", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q02", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector > Analysis and results > 1. Parameters, requests and functions definition > 1.6. Functions to cache", "title": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector", "chunk_index": 8, "token_count": 300, "text_raw": "In this section, functions that will be executed in the caching phase are defined. Caching is the process of storing copies of files in a temporary storage location, so that they can be accessed more quickly. This process also checks if the user has already downloaded a file, avoiding redundant downloads.\n\nFunctions description:\n\n- The `select_timeseries` function subsets the dataset based on the chosen `timeseries` parameter.\n\n- The `compute_indices` function utilises the icclim package to calculate the maximum-temperature-based indices.\n\n- The `compute_trends` function employs the Mann-Kendall test for trend calculation.\n\n- Finally, the `compute_indices_and_trends` function calculates the maximum-temperature-based indices for the corresponding temporal aggregation using the `compute_indices` function, determines the indices mean for the future period (2015-2099), obtain the trends using the `compute_trends` function, and offers an option for regridding to `model_regrid`.\n\nOriginal bounds for conservative interpolation\n\n(section-2)=", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-and-weather-extremes_q02 | Category: Climate_Projections\nSection: Uncertainty in projected changes in extreme temperature indices for the reinsurance sector > Analysis and results > 1. Parameters, requests and functions definition > 1.6. Functions to cache\n---\nIn this section, functions that will be executed in the caching phase are defined. Caching is the process of storing copies of files in a temporary storage location, so that they can be accessed more quickly. This process also checks if the user has already downloaded a file, avoiding redundant downloads.\n\nFunctions description:\n\n- The `select_timeseries` function subsets the dataset based on the chosen `timeseries` parameter.\n\n- The `compute_indices` function utilises the icclim package to calculate the maximum-temperature-based indices.\n\n- The `compute_trends` function employs the Mann-Kendall test for trend calculation.\n\n- Finally, the `compute_indices_and_trends` function calculates the maximum-temperature-based indices for the corresponding temporal aggregation using the `compute_indices` function, determines the indices mean for the future period (2015-2099), obtain the trends using the `compute_trends` function, and offers an option for regridding to `model_regrid`.\n\nOriginal bounds for conservative interpolation\n\n(section-2)="} {"chunk_id": "climate_projections-cmip6_climate-and-weather-extremes_q02__bb37032a3d9a", "report_id": "climate_projections-cmip6_climate-and-weather-extremes_q02", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q02", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector > Analysis and results > 2. Downloading and processing > 2.1. Download and transform the regridding model", "title": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector", "chunk_index": 9, "token_count": 288, "text_raw": "In this section, the `download.download_and_transform` function from the 'c3s_eqc_automatic_quality_control' package is employed to download daily data from the selected CMIP6 regridding model, compute the maximum-temperature-based indices for the selected temporal aggregation, calculate the mean and trend over the future projections period (2015-2099), and cache the result (to avoid redundant downloads and processing).\n\nThe regridding model is intended here as the model whose grid will be used to interpolate the others. This ensures all models share a common grid, facilitating the calculation of median values for each cell point. The regridding model within this notebook is \"gfdl_esm4\" but a different one can be selected by just modifying the `model_regrid` parameter at [](section-1.3). It is key to highlight the importance of the chosen target grid depending on the specific application.\n\n(section-2.2)=", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-and-weather-extremes_q02 | Category: Climate_Projections\nSection: Uncertainty in projected changes in extreme temperature indices for the reinsurance sector > Analysis and results > 2. Downloading and processing > 2.1. Download and transform the regridding model\n---\nIn this section, the `download.download_and_transform` function from the 'c3s_eqc_automatic_quality_control' package is employed to download daily data from the selected CMIP6 regridding model, compute the maximum-temperature-based indices for the selected temporal aggregation, calculate the mean and trend over the future projections period (2015-2099), and cache the result (to avoid redundant downloads and processing).\n\nThe regridding model is intended here as the model whose grid will be used to interpolate the others. This ensures all models share a common grid, facilitating the calculation of median values for each cell point. The regridding model within this notebook is \"gfdl_esm4\" but a different one can be selected by just modifying the `model_regrid` parameter at [](section-1.3). It is key to highlight the importance of the chosen target grid depending on the specific application.\n\n(section-2.2)="} {"chunk_id": "climate_projections-cmip6_climate-and-weather-extremes_q02__a1ab5ca8f28f", "report_id": "climate_projections-cmip6_climate-and-weather-extremes_q02", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q02", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector > Analysis and results > 2. Downloading and processing > 2.2. Download and transform models", "title": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector", "chunk_index": 10, "token_count": 359, "text_raw": "In this section, the `download.download_and_transform` function from the 'c3s_eqc_automatic_quality_control' package is employed to download daily data from the CMIP6 models, compute the maximum-temperature-based indices for the selected temporal aggregation, calculate the mean and trend over the future period (2015-2099), interpolate to the regridding model's grid (only for the cases in which it is specified, in the other cases, the original model's grid is mantained), and cache the result (to avoid redundant downloads and processing).\n\nOriginal model\nInterpolated model\n\n```text\nmodel='access_cm2'\nmodel='awi_cm_1_1_mr'\nmodel='cmcc_esm2'\nmodel='cnrm_cm6_1_hr'\nmodel='cnrm_esm2_1'\nmodel='ec_earth3_cc'\nmodel='gfdl_esm4'\nmodel='inm_cm4_8'\nmodel='inm_cm5_0'\nmodel='kiost_esm'\nmodel='mpi_esm1_2_lr'\nmodel='miroc6'\nmodel='miroc_es2l'\nmodel='mri_esm2_0'\nmodel='noresm2_mm'\nmodel='nesm3'\n```\n\n(section-2.3)=", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-and-weather-extremes_q02 | Category: Climate_Projections\nSection: Uncertainty in projected changes in extreme temperature indices for the reinsurance sector > Analysis and results > 2. Downloading and processing > 2.2. Download and transform models\n---\nIn this section, the `download.download_and_transform` function from the 'c3s_eqc_automatic_quality_control' package is employed to download daily data from the CMIP6 models, compute the maximum-temperature-based indices for the selected temporal aggregation, calculate the mean and trend over the future period (2015-2099), interpolate to the regridding model's grid (only for the cases in which it is specified, in the other cases, the original model's grid is mantained), and cache the result (to avoid redundant downloads and processing).\n\nOriginal model\nInterpolated model\n\n```text\nmodel='access_cm2'\nmodel='awi_cm_1_1_mr'\nmodel='cmcc_esm2'\nmodel='cnrm_cm6_1_hr'\nmodel='cnrm_esm2_1'\nmodel='ec_earth3_cc'\nmodel='gfdl_esm4'\nmodel='inm_cm4_8'\nmodel='inm_cm5_0'\nmodel='kiost_esm'\nmodel='mpi_esm1_2_lr'\nmodel='miroc6'\nmodel='miroc_es2l'\nmodel='mri_esm2_0'\nmodel='noresm2_mm'\nmodel='nesm3'\n```\n\n(section-2.3)="} {"chunk_id": "climate_projections-cmip6_climate-and-weather-extremes_q02__62bbd87c6bff", "report_id": "climate_projections-cmip6_climate-and-weather-extremes_q02", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q02", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector > Analysis and results > 2. Downloading and processing > 2.3. Apply land-sea mask, change attributes and cut the region to show", "title": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector", "chunk_index": 11, "token_count": 272, "text_raw": "This section performs the following tasks:\n\n1. Cut the region of interest.\n2. Downloads the sea mask for ERA5.\n3. Regrids ERA5's mask to the `model_regrid` grid and applies it to the regridded data\n4. Regrids the ERA5 land-sea mask to the model's original grid and applies it to them.\n5. Change some variable attributes for plotting purposes.\n\n**Note:** `ds_interpolated` contains data from the models regridded to the regridding model's grid. `model_datasets` contain the same data but in the original grid of each model.\n\nEdit attributes\n\n```text\n100%|██████████| 1/1 [00:00<00:00, 16.00it/s]\n```\n\n(section-3)=", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-and-weather-extremes_q02 | Category: Climate_Projections\nSection: Uncertainty in projected changes in extreme temperature indices for the reinsurance sector > Analysis and results > 2. Downloading and processing > 2.3. Apply land-sea mask, change attributes and cut the region to show\n---\nThis section performs the following tasks:\n\n1. Cut the region of interest.\n2. Downloads the sea mask for ERA5.\n3. Regrids ERA5's mask to the `model_regrid` grid and applies it to the regridded data\n4. Regrids the ERA5 land-sea mask to the model's original grid and applies it to them.\n5. Change some variable attributes for plotting purposes.\n\n**Note:** `ds_interpolated` contains data from the models regridded to the regridding model's grid. `model_datasets` contain the same data but in the original grid of each model.\n\nEdit attributes\n\n```text\n100%|██████████| 1/1 [00:00<00:00, 16.00it/s]\n```\n\n(section-3)="} {"chunk_id": "climate_projections-cmip6_climate-and-weather-extremes_q02__479a0f10e821", "report_id": "climate_projections-cmip6_climate-and-weather-extremes_q02", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q02", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector > Analysis and results > 3. Plot and describe results", "title": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector", "chunk_index": 12, "token_count": 205, "text_raw": "This section will display the following results:\n\n- Maps representing the spatial distribution of the **future trends** (2015-2099) of the indices 'SU' and 'TX90p' for each model individually, the ensemble median (understood as the median of the trend values of the chosen subset of models calculated for each grid cell), and the ensemble spread (derived as the standard deviation of the distribution of the chosen subset of models).\n- **Boxplots** which represent statistical distributions (PDF) built on the the spatially-averaged future trend from each considered model.\n\n(section-3.1)=", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-and-weather-extremes_q02 | Category: Climate_Projections\nSection: Uncertainty in projected changes in extreme temperature indices for the reinsurance sector > Analysis and results > 3. Plot and describe results\n---\nThis section will display the following results:\n\n- Maps representing the spatial distribution of the **future trends** (2015-2099) of the indices 'SU' and 'TX90p' for each model individually, the ensemble median (understood as the median of the trend values of the chosen subset of models calculated for each grid cell), and the ensemble spread (derived as the standard deviation of the distribution of the chosen subset of models).\n- **Boxplots** which represent statistical distributions (PDF) built on the the spatially-averaged future trend from each considered model.\n\n(section-3.1)="} {"chunk_id": "climate_projections-cmip6_climate-and-weather-extremes_q02__79050a3cec57", "report_id": "climate_projections-cmip6_climate-and-weather-extremes_q02", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q02", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector > Analysis and results > 3. Plot and describe results > 3.1. Define plotting functions", "title": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector", "chunk_index": 13, "token_count": 636, "text_raw": "The functions presented here are used to plot the trends calculated over the future period (2015-2099) for each of the indices ('SU' and 'TX90p').\n\nFor a selected index, two layout types will be displayed, depending on the chosen function:\n\n1. Layout including the ensemble median and the ensemble spread for the trend: `plot_ensemble()` is used.\n2. Layout including every model trend: `plot_models()` is employed.\n\n`trend==True` allows displaying trend values over the future period, while `trend==False` show mean values. In this notebook, which focuses on the future period, only trend values will be shown, and, consequently, `trend==True`. When the `trend` argument is set to True, regions with no significance are hatched. For individual models, a grid point is considered to have a statistically significant trend when the p-value is lower than 0.05 (in such cases, no hatching is shown). However, for determining trend significance for the ensemble median (understood as the median of the trend values of the chosen subset of models calculated for each grid cell), reliance is placed on agreement categories, following the advanced approach proposed in AR6 [IPCC](https://www.ipcc.ch/report/ar6/wg1/downloads/report/IPCC_AR6_WGI_Atlas.pdf) on pages 1945-1950. The `hatch_p_value_ensemble()` function is used to distinguish, for each grid point, between three possible cases:\n\n1. If more than 66% of the models are statistically significant (p-value < 0.05) and more than 80% of the models share the same sign, we consider the ensemble median trend to be statistically significant, and there is agreement on the sign. To represent this, no hatching is used.\n2. If less than 66% of the models are statistically significant, regardless of agreement on the sign of the trend, hatching is applied (indicating that the ensemble median trend is not statistically significant).\n3. If more than 66% of the models are statistically significant but less than 80% of the models share the same sign, we consider the ensemble median trend to be statistically significant, but there is no agreement on the sign of the trend. This is represented using crosses.\n\nDefine function to plot the caption of the figures (for the ensemble case)\nAdd caption to the figure\nAdd each line to the figure\nend captioning\nDefine function to plot the caption of the figures (for the individual models case)\nAdd caption to the figure\nAdd each line to the figure\n\n(section-3.2)=", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-and-weather-extremes_q02 | Category: Climate_Projections\nSection: Uncertainty in projected changes in extreme temperature indices for the reinsurance sector > Analysis and results > 3. Plot and describe results > 3.1. Define plotting functions\n---\nThe functions presented here are used to plot the trends calculated over the future period (2015-2099) for each of the indices ('SU' and 'TX90p').\n\nFor a selected index, two layout types will be displayed, depending on the chosen function:\n\n1. Layout including the ensemble median and the ensemble spread for the trend: `plot_ensemble()` is used.\n2. Layout including every model trend: `plot_models()` is employed.\n\n`trend==True` allows displaying trend values over the future period, while `trend==False` show mean values. In this notebook, which focuses on the future period, only trend values will be shown, and, consequently, `trend==True`. When the `trend` argument is set to True, regions with no significance are hatched. For individual models, a grid point is considered to have a statistically significant trend when the p-value is lower than 0.05 (in such cases, no hatching is shown). However, for determining trend significance for the ensemble median (understood as the median of the trend values of the chosen subset of models calculated for each grid cell), reliance is placed on agreement categories, following the advanced approach proposed in AR6 [IPCC](https://www.ipcc.ch/report/ar6/wg1/downloads/report/IPCC_AR6_WGI_Atlas.pdf) on pages 1945-1950. The `hatch_p_value_ensemble()` function is used to distinguish, for each grid point, between three possible cases:\n\n1. If more than 66% of the models are statistically significant (p-value < 0.05) and more than 80% of the models share the same sign, we consider the ensemble median trend to be statistically significant, and there is agreement on the sign. To represent this, no hatching is used.\n2. If less than 66% of the models are statistically significant, regardless of agreement on the sign of the trend, hatching is applied (indicating that the ensemble median trend is not statistically significant).\n3. If more than 66% of the models are statistically significant but less than 80% of the models share the same sign, we consider the ensemble median trend to be statistically significant, but there is no agreement on the sign of the trend. This is represented using crosses.\n\nDefine function to plot the caption of the figures (for the ensemble case)\nAdd caption to the figure\nAdd each line to the figure\nend captioning\nDefine function to plot the caption of the figures (for the individual models case)\nAdd caption to the figure\nAdd each line to the figure\n\n(section-3.2)="} {"chunk_id": "climate_projections-cmip6_climate-and-weather-extremes_q02__a53bf0d70f34", "report_id": "climate_projections-cmip6_climate-and-weather-extremes_q02", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q02", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector > Analysis and results > 3. Plot and describe results > 3.2. Plot ensemble maps", "title": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector", "chunk_index": 14, "token_count": 271, "text_raw": "In this section, we invoke the `plot_ensemble()` function to visualise the trend calculated over the future period (2015-2099) for the model ensemble across Europe. Note that the model data used in this section has previously been interpolated to the \"regridding model\" grid (`\"gfdl_esm4\"` for this notebook).\n\nSpecifically, for each of the indices ('SU' and 'TX90p'), this section presents a single layout including trend values of the future period (2015-2099) for: (a) the ensemble median (understood as the median of the trend values of the chosen subset of models calculated for each grid cell) and (b) the ensemble spread (derived as the standard deviation of the distribution of the chosen subset of models).\n\nChange default colorbars\nFig number counter\nCommon title\n\n(section-3.3)=", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-and-weather-extremes_q02 | Category: Climate_Projections\nSection: Uncertainty in projected changes in extreme temperature indices for the reinsurance sector > Analysis and results > 3. Plot and describe results > 3.2. Plot ensemble maps\n---\nIn this section, we invoke the `plot_ensemble()` function to visualise the trend calculated over the future period (2015-2099) for the model ensemble across Europe. Note that the model data used in this section has previously been interpolated to the \"regridding model\" grid (`\"gfdl_esm4\"` for this notebook).\n\nSpecifically, for each of the indices ('SU' and 'TX90p'), this section presents a single layout including trend values of the future period (2015-2099) for: (a) the ensemble median (understood as the median of the trend values of the chosen subset of models calculated for each grid cell) and (b) the ensemble spread (derived as the standard deviation of the distribution of the chosen subset of models).\n\nChange default colorbars\nFig number counter\nCommon title\n\n(section-3.3)="} {"chunk_id": "climate_projections-cmip6_climate-and-weather-extremes_q02__1c649ab53e56", "report_id": "climate_projections-cmip6_climate-and-weather-extremes_q02", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q02", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector > Analysis and results > 3. Plot and describe results > 3.3. Plot model maps", "title": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector", "chunk_index": 15, "token_count": 185, "text_raw": "In this section, we invoke the `plot_models()` function to visualise the trend over the future period (2015-2099) for every model individually. Note that the model data used in this section maintains its original grid.\n\nSpecifically, for each of the indices ('SU' and 'TX90p'), this section presents a single layout including the trend for the future period (2015-2099) of every model.\n\n(section-3.4)=", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-and-weather-extremes_q02 | Category: Climate_Projections\nSection: Uncertainty in projected changes in extreme temperature indices for the reinsurance sector > Analysis and results > 3. Plot and describe results > 3.3. Plot model maps\n---\nIn this section, we invoke the `plot_models()` function to visualise the trend over the future period (2015-2099) for every model individually. Note that the model data used in this section maintains its original grid.\n\nSpecifically, for each of the indices ('SU' and 'TX90p'), this section presents a single layout including the trend for the future period (2015-2099) of every model.\n\n(section-3.4)="} {"chunk_id": "climate_projections-cmip6_climate-and-weather-extremes_q02__cc9e7a4bcb73", "report_id": "climate_projections-cmip6_climate-and-weather-extremes_q02", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q02", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector > Analysis and results > 3. Plot and describe results > 3.4. Boxplots of the future trend", "title": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector", "chunk_index": 16, "token_count": 399, "text_raw": "Finally, we present boxplots representing the ensemble distribution of each climate model trend calculated over the future period (2015-2099) across Europe.\n\nDots represent the spatially-averaged future trend over the selected region (change of the number of days per decade) for each model (grey) and the ensemble mean (blue). The ensemble median is shown as a green line. Note that the spatially averaged values are calculated for each model from its original grid (i.e., no interpolated data has been used here).\n\nThe boxplot visually illustrates the distribution of trends among the climate models, with the box covering the first quartile (Q1 = 25th percentile) to the third quartile (Q3 = 75th percentile), and a green line indicating the ensemble median (Q2 = 50th percentile). Whiskers extend from the edges of the box to show the full data range.\n\nEnsemble mean\n\n
\n
\n

Fig 5. Boxplots illustrating the future trends of the distribution of the chosen subset of models for: (a) the 'SU' index and (b) the 'TX90p' index. The distribution is created by considering spatially averaged trends across Europe. The ensemble mean and the ensemble median trends are both included. Outliers in the distribution are denoted by a grey circle with a black contour.

\n
\n\n(section-3.5)=", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-and-weather-extremes_q02 | Category: Climate_Projections\nSection: Uncertainty in projected changes in extreme temperature indices for the reinsurance sector > Analysis and results > 3. Plot and describe results > 3.4. Boxplots of the future trend\n---\nFinally, we present boxplots representing the ensemble distribution of each climate model trend calculated over the future period (2015-2099) across Europe.\n\nDots represent the spatially-averaged future trend over the selected region (change of the number of days per decade) for each model (grey) and the ensemble mean (blue). The ensemble median is shown as a green line. Note that the spatially averaged values are calculated for each model from its original grid (i.e., no interpolated data has been used here).\n\nThe boxplot visually illustrates the distribution of trends among the climate models, with the box covering the first quartile (Q1 = 25th percentile) to the third quartile (Q3 = 75th percentile), and a green line indicating the ensemble median (Q2 = 50th percentile). Whiskers extend from the edges of the box to show the full data range.\n\nEnsemble mean\n\n
\n
\n

Fig 5. Boxplots illustrating the future trends of the distribution of the chosen subset of models for: (a) the 'SU' index and (b) the 'TX90p' index. The distribution is created by considering spatially averaged trends across Europe. The ensemble mean and the ensemble median trends are both included. Outliers in the distribution are denoted by a grey circle with a black contour.

\n
\n\n(section-3.5)="} {"chunk_id": "climate_projections-cmip6_climate-and-weather-extremes_q02__1d7a48850e57", "report_id": "climate_projections-cmip6_climate-and-weather-extremes_q02", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q02", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector > Analysis and results > 3. Plot and describe results > 3.5. Results summary and discussion", "title": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector", "chunk_index": 17, "token_count": 613, "text_raw": "- What do the results mean for users? Are the biases from the historical assessment (\"CMIP6 Climate Projections: evaluating bias in extreme temperature indices for the reinsurance sector\") relevant?\n\n- The projected increase in summer days and days above the 90th percentile offers valuable insights for some applications like developing strategies to optimise reinsurance protections from 2015-2099. While all considered models show a positive trend for these indices, the magnitude of these trends and their uncertainty (quantified by the inter-model spread) vary spatially, emphasising the need for careful interpretation.\n\n- Additionally, biases identified in these models during the historical period (1971-2000), as reviewed in \"CMIP6 Climate Projections: evaluating bias in extreme temperature indices for the reinsurance sector\", should be accounted for before using these results. This evaluation highlighted an overall underestimation of trends and climatology for 'SU', along with challenges in accurately representing spatial distribution, particularly over regions with complex orography.\n\n- For the selected subset of models, the 'SU' index indicates near-zero trends in the southern Mediterranean Basin and northernmost Europe. This disparity may stem from northern Europe rarely reaching the 25°C threshold required for summer days, while the southern Mediterranean consistently exceeds this threshold throughout the JJA season. This underscores the need for a comprehensive assessment using both statistically ('SU' index) and physically-based extreme indices ('TX90p' index).\n\n- Regional differences in 'TX90p' reveal higher values in the southern half of Europe. This regional differences are better appreciated in another assessment using CORDEX Regional Climate Models data (\"CORDEX Climate projections: evaluating uncertainty in projected changes in extreme temperature indices for the reinsurance sector.\").\n\n- Boxplots display spatially averaged positive trends across Europe. The ensemble median trend for the 'SU' index is approximately 2 days per decade, similar to ERA5 data from 1971 to 2000. For the 'TX90p' index, the ensemble median trend is around 6.5 days per decade. The interquantile range spans from 1.7 to 2.25 days per decade for 'SU' and from 6 to 7 days per decade for 'TX90p'.\n\n\n
\nRESULTS NOTE:
\nIt is important to note that the results presented are specific to the 16 models chosen, and users should aim to assess as wide a range of models as possible before making a sub-selection.", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-and-weather-extremes_q02 | Category: Climate_Projections\nSection: Uncertainty in projected changes in extreme temperature indices for the reinsurance sector > Analysis and results > 3. Plot and describe results > 3.5. Results summary and discussion\n---\n- What do the results mean for users? Are the biases from the historical assessment (\"CMIP6 Climate Projections: evaluating bias in extreme temperature indices for the reinsurance sector\") relevant?\n\n- The projected increase in summer days and days above the 90th percentile offers valuable insights for some applications like developing strategies to optimise reinsurance protections from 2015-2099. While all considered models show a positive trend for these indices, the magnitude of these trends and their uncertainty (quantified by the inter-model spread) vary spatially, emphasising the need for careful interpretation.\n\n- Additionally, biases identified in these models during the historical period (1971-2000), as reviewed in \"CMIP6 Climate Projections: evaluating bias in extreme temperature indices for the reinsurance sector\", should be accounted for before using these results. This evaluation highlighted an overall underestimation of trends and climatology for 'SU', along with challenges in accurately representing spatial distribution, particularly over regions with complex orography.\n\n- For the selected subset of models, the 'SU' index indicates near-zero trends in the southern Mediterranean Basin and northernmost Europe. This disparity may stem from northern Europe rarely reaching the 25°C threshold required for summer days, while the southern Mediterranean consistently exceeds this threshold throughout the JJA season. This underscores the need for a comprehensive assessment using both statistically ('SU' index) and physically-based extreme indices ('TX90p' index).\n\n- Regional differences in 'TX90p' reveal higher values in the southern half of Europe. This regional differences are better appreciated in another assessment using CORDEX Regional Climate Models data (\"CORDEX Climate projections: evaluating uncertainty in projected changes in extreme temperature indices for the reinsurance sector.\").\n\n- Boxplots display spatially averaged positive trends across Europe. The ensemble median trend for the 'SU' index is approximately 2 days per decade, similar to ERA5 data from 1971 to 2000. For the 'TX90p' index, the ensemble median trend is around 6.5 days per decade. The interquantile range spans from 1.7 to 2.25 days per decade for 'SU' and from 6 to 7 days per decade for 'TX90p'.\n\n\n
\nRESULTS NOTE:
\nIt is important to note that the results presented are specific to the 16 models chosen, and users should aim to assess as wide a range of models as possible before making a sub-selection."} {"chunk_id": "climate_projections-cmip6_climate-and-weather-extremes_q02__869d14876599", "report_id": "climate_projections-cmip6_climate-and-weather-extremes_q02", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q02", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector > ℹ️ If you want to know more > Key resources", "title": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector", "chunk_index": 18, "token_count": 234, "text_raw": "Some key resources and further reading were linked throughout this assessment.\n\nThe CDS catalogue entries for the data used were:\n* CMIP6 climate projections (Daily - Daily maximum near-surface air temperature): https://cds.climate.copernicus.eu/datasets/projections-cmip6?tab=overview\n\nCode libraries used:\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)\n* [icclim](https://icclim.readthedocs.io/en/stable/) Python package", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-and-weather-extremes_q02 | Category: Climate_Projections\nSection: Uncertainty in projected changes in extreme temperature indices for the reinsurance sector > ℹ️ If you want to know more > Key resources\n---\nSome key resources and further reading were linked throughout this assessment.\n\nThe CDS catalogue entries for the data used were:\n* CMIP6 climate projections (Daily - Daily maximum near-surface air temperature): https://cds.climate.copernicus.eu/datasets/projections-cmip6?tab=overview\n\nCode libraries used:\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)\n* [icclim](https://icclim.readthedocs.io/en/stable/) Python package"} {"chunk_id": "climate_projections-cmip6_climate-and-weather-extremes_q02__94bc6bdc9e92", "report_id": "climate_projections-cmip6_climate-and-weather-extremes_q02", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q02", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector > ℹ️ If you want to know more > References", "title": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector", "chunk_index": 19, "token_count": 776, "text_raw": "[[1]](https://doi.org/10.3390/atmos11020146) Tesselaar, M., Botzen, W.J.W., Aerts, J.C.J.H. (2020). Impacts of Climate Change and Remote Natural Catastrophes on EU Flood Insurance Markets: An Analysis of Soft and Hard Reinsurance Markets for Flood Coverage. Atmosphere 2020, 11, 146. https://doi.org/10.3390/atmos11020146\n\n[[2]](https://doi.org/10.5194/nhess-22-659-2022) Rädler, A. T. (2022). Invited perspectives: how does climate change affect the risk of natural hazards? Challenges and step changes from the reinsurance perspective. Nat. Hazards Earth Syst. Sci., 22, 659–664. https://doi.org/10.5194/nhess-22-659-2022\n\n[[3]](https://www.ipcc.ch/site/assets/uploads/2018/03/wg2TARchap8.pdf) Vellinga, P., Mills, E., Bowers, L., Berz, G.A., Huq, S., Kozak, L.M., Paultikof, J., Schanzenbacker, B., Shida, S., Soler, G., Benson, C., Bidan, P., Bruce, J.W., Huyck, P.M., Lemcke, G., Peara, A., Radevsky, R., Schoubroeck, C.V., Dlugolecki, A.F. (2001). Insurance and other financial services. In J. J. McCarthy, O. F. Canziani, N. A. Leary, D. J. Dokken, & K. S. White (Eds.), Climate change 2001: impacts, adaptation, and vulnerability. Contribution of working group 2 to the third assessment report of the intergovernmental panel on climate change. (pp. 417-450). Cambridge University Press.\n\n[[4]](https://doi.org/10.1002/wcc.71) Lemos, M.C. and Rood, R.B. (2010). Climate projections and their impact on policy and practice. WIREs Clim Chg, 1: 670-682. https://doi.org/10.1002/wcc.71\n\n[[5]](https://doi.org/10.1002/wcc.579) Nissan H, Goddard L, de Perez EC, et al. (2019). On the use and misuse of climate change projections in international development. WIREs Clim Change, 10:e579. https://doi.org/10.1002/wcc.579\n\n[[6]](https://doi.org/10.5194/esd-13-321-2022) Cos, J., Doblas-Reyes, F., Jury, M., Marcos, R., Bretonnière, P.-A., and Samsó, M. (2022). The Mediterranean climate change hotspot in the CMIP5 and CMIP6 projections, Earth Syst. Dynam., 13, 321–340. https://doi.org/10.5194/esd-13-321-2022", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-and-weather-extremes_q02 | Category: Climate_Projections\nSection: Uncertainty in projected changes in extreme temperature indices for the reinsurance sector > ℹ️ If you want to know more > References\n---\n[[1]](https://doi.org/10.3390/atmos11020146) Tesselaar, M., Botzen, W.J.W., Aerts, J.C.J.H. (2020). Impacts of Climate Change and Remote Natural Catastrophes on EU Flood Insurance Markets: An Analysis of Soft and Hard Reinsurance Markets for Flood Coverage. Atmosphere 2020, 11, 146. https://doi.org/10.3390/atmos11020146\n\n[[2]](https://doi.org/10.5194/nhess-22-659-2022) Rädler, A. T. (2022). Invited perspectives: how does climate change affect the risk of natural hazards? Challenges and step changes from the reinsurance perspective. Nat. Hazards Earth Syst. Sci., 22, 659–664. https://doi.org/10.5194/nhess-22-659-2022\n\n[[3]](https://www.ipcc.ch/site/assets/uploads/2018/03/wg2TARchap8.pdf) Vellinga, P., Mills, E., Bowers, L., Berz, G.A., Huq, S., Kozak, L.M., Paultikof, J., Schanzenbacker, B., Shida, S., Soler, G., Benson, C., Bidan, P., Bruce, J.W., Huyck, P.M., Lemcke, G., Peara, A., Radevsky, R., Schoubroeck, C.V., Dlugolecki, A.F. (2001). Insurance and other financial services. In J. J. McCarthy, O. F. Canziani, N. A. Leary, D. J. Dokken, & K. S. White (Eds.), Climate change 2001: impacts, adaptation, and vulnerability. Contribution of working group 2 to the third assessment report of the intergovernmental panel on climate change. (pp. 417-450). Cambridge University Press.\n\n[[4]](https://doi.org/10.1002/wcc.71) Lemos, M.C. and Rood, R.B. (2010). Climate projections and their impact on policy and practice. WIREs Clim Chg, 1: 670-682. https://doi.org/10.1002/wcc.71\n\n[[5]](https://doi.org/10.1002/wcc.579) Nissan H, Goddard L, de Perez EC, et al. (2019). On the use and misuse of climate change projections in international development. WIREs Clim Change, 10:e579. https://doi.org/10.1002/wcc.579\n\n[[6]](https://doi.org/10.5194/esd-13-321-2022) Cos, J., Doblas-Reyes, F., Jury, M., Marcos, R., Bretonnière, P.-A., and Samsó, M. (2022). The Mediterranean climate change hotspot in the CMIP5 and CMIP6 projections, Earth Syst. Dynam., 13, 321–340. https://doi.org/10.5194/esd-13-321-2022"} {"chunk_id": "climate_projections-cmip6_climate-and-weather-extremes_q03__1a7a7daafffc", "report_id": "climate_projections-cmip6_climate-and-weather-extremes_q03", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q03", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector", "title": "Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector", "chunk_index": 0, "token_count": 122, "text_raw": "Production date: 10-07-2024\n\nProduced by: CMCC foundation - Euro-Mediterranean Center on Climate Change. Albert Martinez Boti.", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-and-weather-extremes_q03 | Category: Climate_Projections\nSection: Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector\n---\nProduction date: 10-07-2024\n\nProduced by: CMCC foundation - Euro-Mediterranean Center on Climate Change. Albert Martinez Boti."} {"chunk_id": "climate_projections-cmip6_climate-and-weather-extremes_q03__6d1eb6626d28", "report_id": "climate_projections-cmip6_climate-and-weather-extremes_q03", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q03", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector > Quality assessment question", "title": "Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector", "chunk_index": 1, "token_count": 796, "text_raw": "* **What are the projected changes for a global warming level of 2°C and associated uncertainties in air temperature extremes in Europe?**\n\nClimate change has a major impact on the reinsurance market [[1]](https://doi.org/10.3390/atmos11020146)[[2]](https://doi.org/10.5194/nhess-22-659-2022). In the third assessment report of the IPCC, hot temperature extremes were already presented as relevant to insurance and related services [[3]](https://www.ipcc.ch/site/assets/uploads/2018/03/wg2TARchap8.pdf). Consequently, the need for reliable regional and global climate projections has become paramount, offering valuable insights for optimising reinsurance strategies in the face of a changing climate landscape. Nonetheless, despite their pivotal role, uncertainties inherent in these projections can potentially lead to misuse [[4]](https://doi.org/10.1002/wcc.71)[[5]](https://doi.org/10.1002/wcc.579). This underscores the importance of accurately calculating and accounting for uncertainties to ensure their appropriate consideration. This notebook utilises data from [CMIP6](https://cds.climate.copernicus.eu/datasets/projections-cmip6?tab=overview) Global Climate Models (GCM) and explores the projected changes and uncertainties in future projections of maximum temperature-based extreme indices by considering the ensemble inter-model spread of a subset of CMIP6 models at a global mean warming level of 2°C. Two maximum temperature-based indices from [ECA&D](https://www.ecad.eu/indicesextremes/) indices (one of physical nature and the other of statistical nature) are computed using the [icclim](https://icclim.readthedocs.io/en/stable/) Python package. The first index, identified by the ETCCDI short name 'SU', quantifies the occurrence of summer days (i.e., with daily maximum temperatures exceeding 25°C) within a year or a season (JJA in this notebook). The second index, labeled 'TX90p', describes the number of days with daily maximum temperatures exceeding the daily 90th percentile of maximum temperature for a 5-day moving window. In this notebook, these calculations are performed over the historical period from 1971 to 2000 and compared to the global warming level of 2°C. The global warming level of 2°C can be defined as the first time the 30-year moving average (centre year) of global temperature is above 2°C compared to pre-industrial (Grigory Nikulin et al. (2018) [[6]](https://doi.org/10.1088/1748-9326/aab1b1)). The preindustrial period is defined here as the period spanning from 1861 to 1890 and the index calculations are performed for the Shared Socioeconomic Pathways SSP5-8.5. It is important to mention that the results presented here pertain to a specific subset of the CMIP6 ensemble and may not be generalisable to the entire dataset. Also note that a separate assessment examines the representation of climatology and trends of these indices for the same models during the historical period (1971-2000), while another assessment looks at the projected trends of these indices for the same models during a fixed future period (2015-2099).", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-and-weather-extremes_q03 | Category: Climate_Projections\nSection: Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector > Quality assessment question\n---\n* **What are the projected changes for a global warming level of 2°C and associated uncertainties in air temperature extremes in Europe?**\n\nClimate change has a major impact on the reinsurance market [[1]](https://doi.org/10.3390/atmos11020146)[[2]](https://doi.org/10.5194/nhess-22-659-2022). In the third assessment report of the IPCC, hot temperature extremes were already presented as relevant to insurance and related services [[3]](https://www.ipcc.ch/site/assets/uploads/2018/03/wg2TARchap8.pdf). Consequently, the need for reliable regional and global climate projections has become paramount, offering valuable insights for optimising reinsurance strategies in the face of a changing climate landscape. Nonetheless, despite their pivotal role, uncertainties inherent in these projections can potentially lead to misuse [[4]](https://doi.org/10.1002/wcc.71)[[5]](https://doi.org/10.1002/wcc.579). This underscores the importance of accurately calculating and accounting for uncertainties to ensure their appropriate consideration. This notebook utilises data from [CMIP6](https://cds.climate.copernicus.eu/datasets/projections-cmip6?tab=overview) Global Climate Models (GCM) and explores the projected changes and uncertainties in future projections of maximum temperature-based extreme indices by considering the ensemble inter-model spread of a subset of CMIP6 models at a global mean warming level of 2°C. Two maximum temperature-based indices from [ECA&D](https://www.ecad.eu/indicesextremes/) indices (one of physical nature and the other of statistical nature) are computed using the [icclim](https://icclim.readthedocs.io/en/stable/) Python package. The first index, identified by the ETCCDI short name 'SU', quantifies the occurrence of summer days (i.e., with daily maximum temperatures exceeding 25°C) within a year or a season (JJA in this notebook). The second index, labeled 'TX90p', describes the number of days with daily maximum temperatures exceeding the daily 90th percentile of maximum temperature for a 5-day moving window. In this notebook, these calculations are performed over the historical period from 1971 to 2000 and compared to the global warming level of 2°C. The global warming level of 2°C can be defined as the first time the 30-year moving average (centre year) of global temperature is above 2°C compared to pre-industrial (Grigory Nikulin et al. (2018) [[6]](https://doi.org/10.1088/1748-9326/aab1b1)). The preindustrial period is defined here as the period spanning from 1861 to 1890 and the index calculations are performed for the Shared Socioeconomic Pathways SSP5-8.5. It is important to mention that the results presented here pertain to a specific subset of the CMIP6 ensemble and may not be generalisable to the entire dataset. Also note that a separate assessment examines the representation of climatology and trends of these indices for the same models during the historical period (1971-2000), while another assessment looks at the projected trends of these indices for the same models during a fixed future period (2015-2099)."} {"chunk_id": "climate_projections-cmip6_climate-and-weather-extremes_q03__d130b807a058", "report_id": "climate_projections-cmip6_climate-and-weather-extremes_q03", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q03", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector > Quality assessment statement", "title": "Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector", "chunk_index": 2, "token_count": 888, "text_raw": "These are the key outcomes of this assessment\n\n* For the CMIP6 models analysed, there is clear agreement on the sign of the climate signal for the selected indices during the JJA season, underscoring the need for proactive measures to optimise reinsurance protections in the face of projected increases in air temperature extremes for a world 2°C warmer. The advantage of using Global Warming levels (GWL of 2°C for this notebook), as opposed to considering the trend for a fixed future period, is that it eliminates the systematic biases of the models by focusing on differences rather than absolute values. However, it is important to note that the timing of the thirty-year period when the global mean temperature reaches 2°C above the preindustrial baseline varies depending on the model, making it challenging to determine.\n\n* The considered subset of CMIP6 models shows that, in a world 2°C warmer than the preindustrial baseline, the average number of days with maximum daily temperatures surpassing the daily 90th percentile (calculated for the historical period) during the JJA season exhibits spatial differences across Europe. Notably, values are higher for the Mediterranean Basin, consistent with the findings of Josep Cos et al. (2022) [[7]](https://doi.org/10.5194/esd-13-321-2022), who assessed the Mediterranean climate change hotspot using CMIP6 projections. The boxplot analysis for this index displays an ensemble median spatially-averaged value over Europe larger than 32 days for the JJA season, indicating that under a 2°C global warming level, one out of every three days is projected to have a maximum daily temperature above the daily 90th percentile calculated for the historical period.\n\n* The frequency of summer days occurring during the JJA season, is projected to generally increase in a world 2°C warmer than the preindustrial baseline (1861-1890), compared to the average occurrences during the control period (1971-2000) across Europe, albeit with regional variations. Minor changes are expected in the northernmost and southern regions. In the north, where temperatures historically remain below the 25°C threshold even during the JJA season, the increase in the number of summer days is minimal (the threshold temperature of 25°C may be too high to be reached even in a world 2°C warmer than the preindustrial). In the southern Mediterranean Basin, the threshold is usually surpassed throughout the entire JJA season in the historical period, and thus, little change is observed (specially in northern Africa). This arise limitations of this index, indicating the potential need to select a higher threshold to capture the changes in these regions. Such regions, where the changes may be just as impactful or even more so than areas with a significant increase in days above 25 degrees, require careful consideration.\n\n* These findings emphasise the importance of integrating both statistical and physical extreme indices for a comprehensive assessment of climate impacts, particularly when developing adaptive strategies such as reinsurance protections. While percentile-based indices may better capture changes across all regions, they may pose challenges for users who are more used to fixed thresholds.\n```\n\nattachment:ffd3cf41-b222-4773-9815-fb507773368c.png\n---\nalt: GWL_TX90p\nwidth: 700px\n---\nBoxplot illustrating the climate signal (i.e., the mean values for the warming level of 2°C compared to the historical period from 1971 to 2000) for the ensemble distribution of the 'TX90p' index. For this index, the daily 90th percentile threshold is calculated based on the historical period (1971-2000). The distribution is created by considering spatially averaged values across Europe. The ensemble mean and the ensemble median are both included. Outliers in the distribution are denoted by a grey circle with a black contour.\n```", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-and-weather-extremes_q03 | Category: Climate_Projections\nSection: Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector > Quality assessment statement\n---\nThese are the key outcomes of this assessment\n\n* For the CMIP6 models analysed, there is clear agreement on the sign of the climate signal for the selected indices during the JJA season, underscoring the need for proactive measures to optimise reinsurance protections in the face of projected increases in air temperature extremes for a world 2°C warmer. The advantage of using Global Warming levels (GWL of 2°C for this notebook), as opposed to considering the trend for a fixed future period, is that it eliminates the systematic biases of the models by focusing on differences rather than absolute values. However, it is important to note that the timing of the thirty-year period when the global mean temperature reaches 2°C above the preindustrial baseline varies depending on the model, making it challenging to determine.\n\n* The considered subset of CMIP6 models shows that, in a world 2°C warmer than the preindustrial baseline, the average number of days with maximum daily temperatures surpassing the daily 90th percentile (calculated for the historical period) during the JJA season exhibits spatial differences across Europe. Notably, values are higher for the Mediterranean Basin, consistent with the findings of Josep Cos et al. (2022) [[7]](https://doi.org/10.5194/esd-13-321-2022), who assessed the Mediterranean climate change hotspot using CMIP6 projections. The boxplot analysis for this index displays an ensemble median spatially-averaged value over Europe larger than 32 days for the JJA season, indicating that under a 2°C global warming level, one out of every three days is projected to have a maximum daily temperature above the daily 90th percentile calculated for the historical period.\n\n* The frequency of summer days occurring during the JJA season, is projected to generally increase in a world 2°C warmer than the preindustrial baseline (1861-1890), compared to the average occurrences during the control period (1971-2000) across Europe, albeit with regional variations. Minor changes are expected in the northernmost and southern regions. In the north, where temperatures historically remain below the 25°C threshold even during the JJA season, the increase in the number of summer days is minimal (the threshold temperature of 25°C may be too high to be reached even in a world 2°C warmer than the preindustrial). In the southern Mediterranean Basin, the threshold is usually surpassed throughout the entire JJA season in the historical period, and thus, little change is observed (specially in northern Africa). This arise limitations of this index, indicating the potential need to select a higher threshold to capture the changes in these regions. Such regions, where the changes may be just as impactful or even more so than areas with a significant increase in days above 25 degrees, require careful consideration.\n\n* These findings emphasise the importance of integrating both statistical and physical extreme indices for a comprehensive assessment of climate impacts, particularly when developing adaptive strategies such as reinsurance protections. While percentile-based indices may better capture changes across all regions, they may pose challenges for users who are more used to fixed thresholds.\n```\n\nattachment:ffd3cf41-b222-4773-9815-fb507773368c.png\n---\nalt: GWL_TX90p\nwidth: 700px\n---\nBoxplot illustrating the climate signal (i.e., the mean values for the warming level of 2°C compared to the historical period from 1971 to 2000) for the ensemble distribution of the 'TX90p' index. For this index, the daily 90th percentile threshold is calculated based on the historical period (1971-2000). The distribution is created by considering spatially averaged values across Europe. The ensemble mean and the ensemble median are both included. Outliers in the distribution are denoted by a grey circle with a black contour.\n```"} {"chunk_id": "climate_projections-cmip6_climate-and-weather-extremes_q03__e7ca29e1a373", "report_id": "climate_projections-cmip6_climate-and-weather-extremes_q03", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q03", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector > Methodology", "title": "Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector", "chunk_index": 3, "token_count": 455, "text_raw": "This notebook provides an assessment of the projected changes and their associated uncertainties, utilising a subset of 16 models from [CMIP6](https://cds.climate.copernicus.eu/datasets/projections-cmip6?tab=overview) under a global warming level of 2°C. The uncertainty is explored by analysing the ensemble inter-model spread of projected changes for the maximum-temperature-based indices 'SU' and 'TX90p,' calculated over the temporal aggregation of JJA for the specific global warming level of 2°C. In particular, spatial patterns of the climate signal (i.e., mean values of the indices for the warming level of 2°C compared to the historical period from 1971 to 2000) are examined and displayed for each model individually and for the ensemble median (calculated for each grid cell), alongside the ensemble inter-model spread to account for projected uncertainty. Additionally, spatially-averaged values are analysed and presented using box plots to provide an overview of climate signal behavior across the distribution of the chosen subset of models when averaged across Europe.\n\nThe analysis and results follow the next outline:\n\n**[](section-1)**\n * [](section-1.1)\n * [](section-1.2)\n * [](section-1.3)\n * [](section-1.4)\n * [](section-1.5)\n * [](section-1.6)\n\n**[](section-2)**\n * [](section-2.1)\n * [](section-2.2)\n * [](section-2.3)\n\n**[](section-3)**\n * [](section-3.1)\n * [](section-3.2)\n * [](section-3.3)\n * [](section-3.4)\n * [](section-3.5)", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-and-weather-extremes_q03 | Category: Climate_Projections\nSection: Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector > Methodology\n---\nThis notebook provides an assessment of the projected changes and their associated uncertainties, utilising a subset of 16 models from [CMIP6](https://cds.climate.copernicus.eu/datasets/projections-cmip6?tab=overview) under a global warming level of 2°C. The uncertainty is explored by analysing the ensemble inter-model spread of projected changes for the maximum-temperature-based indices 'SU' and 'TX90p,' calculated over the temporal aggregation of JJA for the specific global warming level of 2°C. In particular, spatial patterns of the climate signal (i.e., mean values of the indices for the warming level of 2°C compared to the historical period from 1971 to 2000) are examined and displayed for each model individually and for the ensemble median (calculated for each grid cell), alongside the ensemble inter-model spread to account for projected uncertainty. Additionally, spatially-averaged values are analysed and presented using box plots to provide an overview of climate signal behavior across the distribution of the chosen subset of models when averaged across Europe.\n\nThe analysis and results follow the next outline:\n\n**[](section-1)**\n * [](section-1.1)\n * [](section-1.2)\n * [](section-1.3)\n * [](section-1.4)\n * [](section-1.5)\n * [](section-1.6)\n\n**[](section-2)**\n * [](section-2.1)\n * [](section-2.2)\n * [](section-2.3)\n\n**[](section-3)**\n * [](section-3.1)\n * [](section-3.2)\n * [](section-3.3)\n * [](section-3.4)\n * [](section-3.5)"} {"chunk_id": "climate_projections-cmip6_climate-and-weather-extremes_q03__0953fa7b9e2b", "report_id": "climate_projections-cmip6_climate-and-weather-extremes_q03", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q03", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector > Analysis and results > 1. Parameters, requests and functions definition > 1.2. Define Parameters", "title": "Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector", "chunk_index": 4, "token_count": 421, "text_raw": "In the \"Define Parameters\" section, various customisable options for the notebook are specified. Most of the parameters chosen are the same as those used in other assessments (\"CMIP6 Climate Projections: evaluating bias in extreme temperature indices for the reinsurance sector\"), being them:\n\n- `historical_slice` determines the historical (or control) period used (1971 to 2000 to be consistent with other user question of the same use case)\n- The `timeseries` set the temporal aggregation. For instance, selecting \"JJA\" implies considering only the JJA season.\n- The `index_names` parameter enables the selection of maximum-temperature-based indices ('SU' and 'TX90p' in our case) from the [icclim](https://icclim.readthedocs.io/en/stable/) Python package.\n- `collection_id` sets the family of models. Only CMIP6 is implemented for this notebook.\n- `area` allows specifying the geographical domain of interest.\n- The `interpolation_method` parameter allows selecting the interpolation method when regridding is performed over the indices.\n- The `chunk` selection allows the user to define if dividing into chunks when downloading the data on their local machine. Although it does not significantly affect the analysis, it is recommended to keep the default value for optimal performance.\n\nChoose annual or seasonal timeseries\nSelect the family of models\nDefine region for analysis\nDefine region for request\nDefine index names\nInterpolation method\nChunks for download\n\n(section-1.3)=", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-and-weather-extremes_q03 | Category: Climate_Projections\nSection: Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector > Analysis and results > 1. Parameters, requests and functions definition > 1.2. Define Parameters\n---\nIn the \"Define Parameters\" section, various customisable options for the notebook are specified. Most of the parameters chosen are the same as those used in other assessments (\"CMIP6 Climate Projections: evaluating bias in extreme temperature indices for the reinsurance sector\"), being them:\n\n- `historical_slice` determines the historical (or control) period used (1971 to 2000 to be consistent with other user question of the same use case)\n- The `timeseries` set the temporal aggregation. For instance, selecting \"JJA\" implies considering only the JJA season.\n- The `index_names` parameter enables the selection of maximum-temperature-based indices ('SU' and 'TX90p' in our case) from the [icclim](https://icclim.readthedocs.io/en/stable/) Python package.\n- `collection_id` sets the family of models. Only CMIP6 is implemented for this notebook.\n- `area` allows specifying the geographical domain of interest.\n- The `interpolation_method` parameter allows selecting the interpolation method when regridding is performed over the indices.\n- The `chunk` selection allows the user to define if dividing into chunks when downloading the data on their local machine. Although it does not significantly affect the analysis, it is recommended to keep the default value for optimal performance.\n\nChoose annual or seasonal timeseries\nSelect the family of models\nDefine region for analysis\nDefine region for request\nDefine index names\nInterpolation method\nChunks for download\n\n(section-1.3)="} {"chunk_id": "climate_projections-cmip6_climate-and-weather-extremes_q03__63433e234d99", "report_id": "climate_projections-cmip6_climate-and-weather-extremes_q03", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q03", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector > Analysis and results > 1. Parameters, requests and functions definition > 1.3. Define models", "title": "Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector", "chunk_index": 5, "token_count": 652, "text_raw": "The following climate analyses are performed considering a subset of GCMs from CMIP6. Models names are listed in the parameters below. The selected CMIP6 models have available both the historical and SSP8.5 experiments, and are the same as those used in another assessment (\"CMIP6 Climate Projections: evaluating bias in extreme temperature indices for the reinsurance sector\"). Additionally, for each model, the thirty-year period corresponding to the Global Warming Level of 2°C above the preindustrial period is specified.\n\nThe Global Warming Level of 2°C can be defined as the earliest point at which a 30-year moving average of global temperature exceeds by two degrees compared to the 1861–1890 baseline [[6]](https://doi.org/10.1088/1748-9326/aab1b1). This thirty-year period varies depending on the Global Climate Model being analysed. To determine this timeframe, the global mean surface temperature for the preindustrial baseline is first calculated for each model. Then, a rolling mean over 30 years is applied to the future period from 2015 to 2099, following the SSP5-8.5 scenario. Finally, the earliest 30-year period at which each model reaches the global warming level of 2°C above the preindustrial period is identified.\n\n\n
\nNOTE on the 30-year slice used for Global Warming Level calculations:
\nGiven the 30-year duration of the historical/control period, the period of mean global surface temperature reaching 2 degrees above preindustrial levels was defined as a 30-year interval (while some studies use 20-year slices for this purpose). The pre-industrial period was also calculated as a 30-year slice (as done in [6]), despite the standard practice of considering the period from 1850 to 1900. This approach not only saves computational time but also ensures consistency in working with 30-year slices.\n\n\n
\nNOTE on the Global Warming Level calculation:
\nTo streamline the current notebook and avoid excessive length and complexity, the calculation of the Global Warming Level of 2°C above the preindustrial period for each model has been conducted externally.\n\nDefine dictionaries to use in titles and caption\nDefine dictionaries to use in titles and caption\n\n(section-1.4)=", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-and-weather-extremes_q03 | Category: Climate_Projections\nSection: Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector > Analysis and results > 1. Parameters, requests and functions definition > 1.3. Define models\n---\nThe following climate analyses are performed considering a subset of GCMs from CMIP6. Models names are listed in the parameters below. The selected CMIP6 models have available both the historical and SSP8.5 experiments, and are the same as those used in another assessment (\"CMIP6 Climate Projections: evaluating bias in extreme temperature indices for the reinsurance sector\"). Additionally, for each model, the thirty-year period corresponding to the Global Warming Level of 2°C above the preindustrial period is specified.\n\nThe Global Warming Level of 2°C can be defined as the earliest point at which a 30-year moving average of global temperature exceeds by two degrees compared to the 1861–1890 baseline [[6]](https://doi.org/10.1088/1748-9326/aab1b1). This thirty-year period varies depending on the Global Climate Model being analysed. To determine this timeframe, the global mean surface temperature for the preindustrial baseline is first calculated for each model. Then, a rolling mean over 30 years is applied to the future period from 2015 to 2099, following the SSP5-8.5 scenario. Finally, the earliest 30-year period at which each model reaches the global warming level of 2°C above the preindustrial period is identified.\n\n\n
\nNOTE on the 30-year slice used for Global Warming Level calculations:
\nGiven the 30-year duration of the historical/control period, the period of mean global surface temperature reaching 2 degrees above preindustrial levels was defined as a 30-year interval (while some studies use 20-year slices for this purpose). The pre-industrial period was also calculated as a 30-year slice (as done in [6]), despite the standard practice of considering the period from 1850 to 1900. This approach not only saves computational time but also ensures consistency in working with 30-year slices.\n\n\n
\nNOTE on the Global Warming Level calculation:
\nTo streamline the current notebook and avoid excessive length and complexity, the calculation of the Global Warming Level of 2°C above the preindustrial period for each model has been conducted externally.\n\nDefine dictionaries to use in titles and caption\nDefine dictionaries to use in titles and caption\n\n(section-1.4)="} {"chunk_id": "climate_projections-cmip6_climate-and-weather-extremes_q03__297c3d340e6c", "report_id": "climate_projections-cmip6_climate-and-weather-extremes_q03", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q03", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector > Analysis and results > 1. Parameters, requests and functions definition > 1.4. Define land-sea mask request", "title": "Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector", "chunk_index": 6, "token_count": 162, "text_raw": "Within this notebook, ERA5 will be used to download the land-sea mask when plotting. In this section, we set the required parameters for the cds-api data-request of ERA5 land-sea mask.\n\n(section-1.5)=", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-and-weather-extremes_q03 | Category: Climate_Projections\nSection: Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector > Analysis and results > 1. Parameters, requests and functions definition > 1.4. Define land-sea mask request\n---\nWithin this notebook, ERA5 will be used to download the land-sea mask when plotting. In this section, we set the required parameters for the cds-api data-request of ERA5 land-sea mask.\n\n(section-1.5)="} {"chunk_id": "climate_projections-cmip6_climate-and-weather-extremes_q03__aa09b0c08e9a", "report_id": "climate_projections-cmip6_climate-and-weather-extremes_q03", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q03", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector > Analysis and results > 1. Parameters, requests and functions definition > 1.5. Define model requests", "title": "Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector", "chunk_index": 7, "token_count": 216, "text_raw": "In this section we set the required parameters for the cds-api data-request.\n\nWhen `weights = True`, spatial weighting is applied for calculations requiring spatial data aggregation. This is particularly relevant for CMIP6 GCMs with regular lon-lat grids that do not consider varying surface extensions at different latitudes. In contrast, CORDEX RCMs, using rotated grids, inherently account for different cell surfaces based on latitude, eliminating the need for a latitude cosine multiplicative factor (`weights = False`).\n\n(section-1.6)=", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-and-weather-extremes_q03 | Category: Climate_Projections\nSection: Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector > Analysis and results > 1. Parameters, requests and functions definition > 1.5. Define model requests\n---\nIn this section we set the required parameters for the cds-api data-request.\n\nWhen `weights = True`, spatial weighting is applied for calculations requiring spatial data aggregation. This is particularly relevant for CMIP6 GCMs with regular lon-lat grids that do not consider varying surface extensions at different latitudes. In contrast, CORDEX RCMs, using rotated grids, inherently account for different cell surfaces based on latitude, eliminating the need for a latitude cosine multiplicative factor (`weights = False`).\n\n(section-1.6)="} {"chunk_id": "climate_projections-cmip6_climate-and-weather-extremes_q03__864e3083851a", "report_id": "climate_projections-cmip6_climate-and-weather-extremes_q03", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q03", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector > Analysis and results > 1. Parameters, requests and functions definition > 1.6. Functions to cache", "title": "Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector", "chunk_index": 8, "token_count": 317, "text_raw": "In this section, functions that will be executed in the caching phase are defined. Caching is the process of storing copies of files in a temporary storage location, so that they can be accessed more quickly. This process also checks if the user has already downloaded a file, avoiding redundant downloads.\n\nFunctions description:\n\n- The `select_timeseries` function subsets the dataset based on the chosen `timeseries` parameter, which could be a specific season (e.g., \"JJA\") or \"annual.\" It tailors the dataset according to the specified criteria, providing either annual or seasonal data as the outcome.\n\n- The `compute_indices` function utilises the icclim package to calculate the selected maximum-temperature-based indices.\n\n- Finally, the `compute_indices_and_trends` function selects the temporal aggregation using the `select_timeseries` function. It then calculates maximum-temperature-based indices via the `compute_indices` function and determines the indices mean over the period of interest.\n\nOriginal bounds for conservative interpolation\n\n(section-2)=", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-and-weather-extremes_q03 | Category: Climate_Projections\nSection: Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector > Analysis and results > 1. Parameters, requests and functions definition > 1.6. Functions to cache\n---\nIn this section, functions that will be executed in the caching phase are defined. Caching is the process of storing copies of files in a temporary storage location, so that they can be accessed more quickly. This process also checks if the user has already downloaded a file, avoiding redundant downloads.\n\nFunctions description:\n\n- The `select_timeseries` function subsets the dataset based on the chosen `timeseries` parameter, which could be a specific season (e.g., \"JJA\") or \"annual.\" It tailors the dataset according to the specified criteria, providing either annual or seasonal data as the outcome.\n\n- The `compute_indices` function utilises the icclim package to calculate the selected maximum-temperature-based indices.\n\n- Finally, the `compute_indices_and_trends` function selects the temporal aggregation using the `select_timeseries` function. It then calculates maximum-temperature-based indices via the `compute_indices` function and determines the indices mean over the period of interest.\n\nOriginal bounds for conservative interpolation\n\n(section-2)="} {"chunk_id": "climate_projections-cmip6_climate-and-weather-extremes_q03__06338de47840", "report_id": "climate_projections-cmip6_climate-and-weather-extremes_q03", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q03", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector > Analysis and results > 2. Downloading and processing > 2.1. Download and transform the regridding model", "title": "Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector", "chunk_index": 9, "token_count": 384, "text_raw": "In this section, the `download.download_and_transform` function from the 'c3s_eqc_automatic_quality_control' package is employed to download daily data from the selected CMIP6 regridding model, compute the indices for the selected temporal aggregation (\"JJA\" in our example), calculate the mean for the historical period (1971-2000) and for the thirty-year period corresponding to the 2°C Global Warming Level, and cache the result (to avoid redundant downloads and processing).\n\nThe regridding model is intended here as the model whose grid will be used to interpolate the others. This ensures all models share a common grid, facilitating the calculation of median values for each cell point. The regridding model within this notebook is \"gfdl_esm4\" but a different one can be selected by just modifying the `model_regrid` parameter at [](section-1.3). It is key to highlight the importance of the chosen target grid depending on the specific application.\n\n\n
\nNOTE on the Global Warming Level calculation:
\nThe thirty year-period corresponding to the 2°C Global Warming Level is different for each model. Its calculation has been conducted externally.\n\n(section-2.2)=", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-and-weather-extremes_q03 | Category: Climate_Projections\nSection: Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector > Analysis and results > 2. Downloading and processing > 2.1. Download and transform the regridding model\n---\nIn this section, the `download.download_and_transform` function from the 'c3s_eqc_automatic_quality_control' package is employed to download daily data from the selected CMIP6 regridding model, compute the indices for the selected temporal aggregation (\"JJA\" in our example), calculate the mean for the historical period (1971-2000) and for the thirty-year period corresponding to the 2°C Global Warming Level, and cache the result (to avoid redundant downloads and processing).\n\nThe regridding model is intended here as the model whose grid will be used to interpolate the others. This ensures all models share a common grid, facilitating the calculation of median values for each cell point. The regridding model within this notebook is \"gfdl_esm4\" but a different one can be selected by just modifying the `model_regrid` parameter at [](section-1.3). It is key to highlight the importance of the chosen target grid depending on the specific application.\n\n\n
\nNOTE on the Global Warming Level calculation:
\nThe thirty year-period corresponding to the 2°C Global Warming Level is different for each model. Its calculation has been conducted externally.\n\n(section-2.2)="} {"chunk_id": "climate_projections-cmip6_climate-and-weather-extremes_q03__3470841bd29e", "report_id": "climate_projections-cmip6_climate-and-weather-extremes_q03", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q03", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector > Analysis and results > 2. Downloading and processing > 2.2. Download and transform models", "title": "Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector", "chunk_index": 10, "token_count": 564, "text_raw": "In this section, we utilise the `download_and_transform` function from the 'c3s_eqc_automatic_quality_control' package to download daily data from the CMIP6 models for the historical period (1971-2000) and for the thirty-year period corresponding to the 2°C warming level of each model. Subsequently, temporal aggregation is selected (in our example, 'JJA'). The next step involves calculating the indices and the climate signal. For the 'SU' index, as we are dealing with an index calculated using a physical threshold, the climate signal is obtained as the difference between the mean values for the thirty-year period corresponding to the 2°C Global Warming Level and the mean values calculated over the historical period (1971-2000). The climate signal for the 'TX90p' index, as we are dealing with an index based on a statistical threshold, is obtained by calculating - for the thirty-year period corresponding to the 2°C warming level - the mean number of days with daily maximum temperatures exceeding the daily 90th percentile. The daily 90th percentile threshold is calculated based on the historical period of 1971-2000.\n\n\n
\nNOTE on the Global Warming Level calculation:
\nThe thirty year-period corresponding to the 2°C Global Warming Level is different for each model. Its calculation has been conducted externally.\n\nOriginal model\nInterpolated model\n\n```text\nmodel='access_cm2'\nmodel='awi_cm_1_1_mr'\nmodel='cmcc_esm2'\nmodel='cnrm_cm6_1_hr'\nmodel='cnrm_esm2_1'\nmodel='ec_earth3_cc'\nmodel='gfdl_esm4'\nmodel='inm_cm4_8'\nmodel='inm_cm5_0'\nmodel='kiost_esm'\nmodel='miroc6'\nmodel='miroc_es2l'\nmodel='mpi_esm1_2_lr'\nmodel='mri_esm2_0'\nmodel='noresm2_mm'\nmodel='nesm3'\n```\n\n(section-2.3)=", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-and-weather-extremes_q03 | Category: Climate_Projections\nSection: Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector > Analysis and results > 2. Downloading and processing > 2.2. Download and transform models\n---\nIn this section, we utilise the `download_and_transform` function from the 'c3s_eqc_automatic_quality_control' package to download daily data from the CMIP6 models for the historical period (1971-2000) and for the thirty-year period corresponding to the 2°C warming level of each model. Subsequently, temporal aggregation is selected (in our example, 'JJA'). The next step involves calculating the indices and the climate signal. For the 'SU' index, as we are dealing with an index calculated using a physical threshold, the climate signal is obtained as the difference between the mean values for the thirty-year period corresponding to the 2°C Global Warming Level and the mean values calculated over the historical period (1971-2000). The climate signal for the 'TX90p' index, as we are dealing with an index based on a statistical threshold, is obtained by calculating - for the thirty-year period corresponding to the 2°C warming level - the mean number of days with daily maximum temperatures exceeding the daily 90th percentile. The daily 90th percentile threshold is calculated based on the historical period of 1971-2000.\n\n\n
\nNOTE on the Global Warming Level calculation:
\nThe thirty year-period corresponding to the 2°C Global Warming Level is different for each model. Its calculation has been conducted externally.\n\nOriginal model\nInterpolated model\n\n```text\nmodel='access_cm2'\nmodel='awi_cm_1_1_mr'\nmodel='cmcc_esm2'\nmodel='cnrm_cm6_1_hr'\nmodel='cnrm_esm2_1'\nmodel='ec_earth3_cc'\nmodel='gfdl_esm4'\nmodel='inm_cm4_8'\nmodel='inm_cm5_0'\nmodel='kiost_esm'\nmodel='miroc6'\nmodel='miroc_es2l'\nmodel='mpi_esm1_2_lr'\nmodel='mri_esm2_0'\nmodel='noresm2_mm'\nmodel='nesm3'\n```\n\n(section-2.3)="} {"chunk_id": "climate_projections-cmip6_climate-and-weather-extremes_q03__010e9ce99bfa", "report_id": "climate_projections-cmip6_climate-and-weather-extremes_q03", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q03", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector > Analysis and results > 2. Downloading and processing > 2.3. Apply land-sea mask, change attributes and cut the region to show", "title": "Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector", "chunk_index": 11, "token_count": 258, "text_raw": "This section performs the following tasks:\n\n1. Cut the region of interest.\n2. Downloads the sea mask for ERA5.\n3. Regrids ERA5's mask to the `model_regrid` grid and applies it to the regridded data\n4. Regrids the ERA5 land-sea mask to the model's original grid and applies it to them.\n5. Change some variable attributes for plotting purposes.\n\n**Note:** `ds_interpolated` contains data from the models regridded to the regridding model's grid. `model_datasets` contain the same data but in the original grid of each model.\n\nEdit attributes\n\n(section-3)=", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-and-weather-extremes_q03 | Category: Climate_Projections\nSection: Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector > Analysis and results > 2. Downloading and processing > 2.3. Apply land-sea mask, change attributes and cut the region to show\n---\nThis section performs the following tasks:\n\n1. Cut the region of interest.\n2. Downloads the sea mask for ERA5.\n3. Regrids ERA5's mask to the `model_regrid` grid and applies it to the regridded data\n4. Regrids the ERA5 land-sea mask to the model's original grid and applies it to them.\n5. Change some variable attributes for plotting purposes.\n\n**Note:** `ds_interpolated` contains data from the models regridded to the regridding model's grid. `model_datasets` contain the same data but in the original grid of each model.\n\nEdit attributes\n\n(section-3)="} {"chunk_id": "climate_projections-cmip6_climate-and-weather-extremes_q03__78ab996d93ab", "report_id": "climate_projections-cmip6_climate-and-weather-extremes_q03", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q03", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector > Analysis and results > 3. Plot and describe results", "title": "Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector", "chunk_index": 12, "token_count": 239, "text_raw": "This section will display the following results:\n\n- Maps representing the spatial distribution of the climate change signal (i.e., mean values for the warming level of 2°C compared to the historical period 1971-2000) of the indices 'SU' and 'TX90p' for each model individually, the ensemble median (understood as the median of the climate signal values of the chosen subset of models calculated for each grid cell), and the ensemble spread (derived as the standard deviation of the distribution of the chosen subset of models).\n\n- Boxplots representing statistical distributions (PDFs) constructed from the spatially-averaged climate signal of each considered model.\n\n(section-3.1)=", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-and-weather-extremes_q03 | Category: Climate_Projections\nSection: Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector > Analysis and results > 3. Plot and describe results\n---\nThis section will display the following results:\n\n- Maps representing the spatial distribution of the climate change signal (i.e., mean values for the warming level of 2°C compared to the historical period 1971-2000) of the indices 'SU' and 'TX90p' for each model individually, the ensemble median (understood as the median of the climate signal values of the chosen subset of models calculated for each grid cell), and the ensemble spread (derived as the standard deviation of the distribution of the chosen subset of models).\n\n- Boxplots representing statistical distributions (PDFs) constructed from the spatially-averaged climate signal of each considered model.\n\n(section-3.1)="} {"chunk_id": "climate_projections-cmip6_climate-and-weather-extremes_q03__3d15667dd9f5", "report_id": "climate_projections-cmip6_climate-and-weather-extremes_q03", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q03", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector > Analysis and results > 3. Plot and describe results > 3.1. Define plotting functions", "title": "Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector", "chunk_index": 13, "token_count": 276, "text_raw": "The functions presented here are used to plot the climate signal (i.e., the mean values for the warming level of 2°C compared to the historical period) for each of the indices ('SU' and 'TX90p').\n\nFor a selected index, two layout types will be displayed, depending on the chosen function:\n\n1. Layout including the ensemble median and the ensemble spread: `plot_ensemble()` is used.\n2. Layout including every model: `plot_models()` is employed.\n\nDefine function to plot the caption of the figures (for the ensemble case)\nAdd caption to the figure\nAdd each line to the figure\nend captioning\nDefine function to plot the cation of the figures (for the individual models case)\nAdd caption to the figure\nAdd each line to the figure\n\n(section-3.2)=", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-and-weather-extremes_q03 | Category: Climate_Projections\nSection: Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector > Analysis and results > 3. Plot and describe results > 3.1. Define plotting functions\n---\nThe functions presented here are used to plot the climate signal (i.e., the mean values for the warming level of 2°C compared to the historical period) for each of the indices ('SU' and 'TX90p').\n\nFor a selected index, two layout types will be displayed, depending on the chosen function:\n\n1. Layout including the ensemble median and the ensemble spread: `plot_ensemble()` is used.\n2. Layout including every model: `plot_models()` is employed.\n\nDefine function to plot the caption of the figures (for the ensemble case)\nAdd caption to the figure\nAdd each line to the figure\nend captioning\nDefine function to plot the cation of the figures (for the individual models case)\nAdd caption to the figure\nAdd each line to the figure\n\n(section-3.2)="} {"chunk_id": "climate_projections-cmip6_climate-and-weather-extremes_q03__c4a835604d85", "report_id": "climate_projections-cmip6_climate-and-weather-extremes_q03", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q03", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector > Analysis and results > 3. Plot and describe results > 3.2. Plot ensemble maps", "title": "Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector", "chunk_index": 14, "token_count": 269, "text_raw": "In this section, we invoke the `plot_ensemble()` function to visualise the climate signal (i.e., the mean values for the warming level of 2°C compared to the historical period) across Europe for each of the indices ('SU' and 'TX90p'). The layout includes: (a) the ensemble median (understood as the median of the climate signal values of the chosen subset of models calculated for each grid cell) and (b) the ensemble spread (derived as the standard deviation of the distribution of the chosen subset of models).\n\nNote that the model data used in this section has previously been interpolated to the \"regridding model\" grid (`\"gfdl_esm4\"` for this notebook).\n\nFig number counter\nCommon title\n\n(section-3.3)=", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-and-weather-extremes_q03 | Category: Climate_Projections\nSection: Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector > Analysis and results > 3. Plot and describe results > 3.2. Plot ensemble maps\n---\nIn this section, we invoke the `plot_ensemble()` function to visualise the climate signal (i.e., the mean values for the warming level of 2°C compared to the historical period) across Europe for each of the indices ('SU' and 'TX90p'). The layout includes: (a) the ensemble median (understood as the median of the climate signal values of the chosen subset of models calculated for each grid cell) and (b) the ensemble spread (derived as the standard deviation of the distribution of the chosen subset of models).\n\nNote that the model data used in this section has previously been interpolated to the \"regridding model\" grid (`\"gfdl_esm4\"` for this notebook).\n\nFig number counter\nCommon title\n\n(section-3.3)="} {"chunk_id": "climate_projections-cmip6_climate-and-weather-extremes_q03__685851bcad03", "report_id": "climate_projections-cmip6_climate-and-weather-extremes_q03", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q03", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector > Analysis and results > 3. Plot and describe results > 3.3. Plot model maps", "title": "Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector", "chunk_index": 15, "token_count": 185, "text_raw": "In this section, we invoke the `plot_models()` function to visualise the climate signal for every model individually across Europe. Note that the model data used in this section maintains its original grid.\n\nSpecifically, for each of the indices ('SU' and 'TX90p'), this section presents a single layout including the climate signal of every model.\n\n(section-3.4)=", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-and-weather-extremes_q03 | Category: Climate_Projections\nSection: Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector > Analysis and results > 3. Plot and describe results > 3.3. Plot model maps\n---\nIn this section, we invoke the `plot_models()` function to visualise the climate signal for every model individually across Europe. Note that the model data used in this section maintains its original grid.\n\nSpecifically, for each of the indices ('SU' and 'TX90p'), this section presents a single layout including the climate signal of every model.\n\n(section-3.4)="} {"chunk_id": "climate_projections-cmip6_climate-and-weather-extremes_q03__fa0cc626ce14", "report_id": "climate_projections-cmip6_climate-and-weather-extremes_q03", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q03", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector > Analysis and results > 3. Plot and describe results > 3.4. Boxplots of the climate change signal", "title": "Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector", "chunk_index": 16, "token_count": 432, "text_raw": "Finally, we present boxplots representing the ensemble distribution of each climate model signal for the 2°C global warming level.\n\nDots represent the spatially-averaged climate signal over the selected region for each model (grey) and the ensemble mean (blue). The ensemble median is shown as a green line. Note that the spatially averaged values are calculated for each model from its original grid (i.e., no interpolated data has been used here).\n\nThe boxplot visually illustrates the distribution of the climate signal among the climate models, with the box covering the first quartile (Q1 = 25th percentile) to the third quartile (Q3 = 75th percentile), and a green line indicating the ensemble median (Q2 = 50th percentile). Whiskers extend from the edges of the box to show the full data range.\n\nEnsemble mean\n\n
\n
\n

Fig 5. Boxplots illustrating the climate signal (i.e., mean values for the warming level of 2°C compared to the historical period from 1971 to 2000) of the distribution of the chosen subset of models for: (a) the 'SU' index, and (b) the 'TX90p' index. The distribution is created by considering spatially averaged values across Europe. The ensemble mean and the ensemble median are both included. Outliers in the distribution are denoted by a grey circle with a black contour.

\n
\n\n(section-3.5)=", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-and-weather-extremes_q03 | Category: Climate_Projections\nSection: Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector > Analysis and results > 3. Plot and describe results > 3.4. Boxplots of the climate change signal\n---\nFinally, we present boxplots representing the ensemble distribution of each climate model signal for the 2°C global warming level.\n\nDots represent the spatially-averaged climate signal over the selected region for each model (grey) and the ensemble mean (blue). The ensemble median is shown as a green line. Note that the spatially averaged values are calculated for each model from its original grid (i.e., no interpolated data has been used here).\n\nThe boxplot visually illustrates the distribution of the climate signal among the climate models, with the box covering the first quartile (Q1 = 25th percentile) to the third quartile (Q3 = 75th percentile), and a green line indicating the ensemble median (Q2 = 50th percentile). Whiskers extend from the edges of the box to show the full data range.\n\nEnsemble mean\n\n
\n
\n

Fig 5. Boxplots illustrating the climate signal (i.e., mean values for the warming level of 2°C compared to the historical period from 1971 to 2000) of the distribution of the chosen subset of models for: (a) the 'SU' index, and (b) the 'TX90p' index. The distribution is created by considering spatially averaged values across Europe. The ensemble mean and the ensemble median are both included. Outliers in the distribution are denoted by a grey circle with a black contour.

\n
\n\n(section-3.5)="} {"chunk_id": "climate_projections-cmip6_climate-and-weather-extremes_q03__71160dbafb0f", "report_id": "climate_projections-cmip6_climate-and-weather-extremes_q03", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q03", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector > Analysis and results > 3. Plot and describe results > 3.5. Results summary and discussion", "title": "Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector", "chunk_index": 17, "token_count": 967, "text_raw": "- The frequency of mean summer days occurring during the JJA season, defined as those with a daily maximum temperature exceeding 25°C, is projected to generally increase in a world 2°C warmer than the preindustrial baseline (1861-1890), compared to the average occurrences during the control period (1971-2000) across Europe, albeit with regional variations. Minor changes are expected in the northernmost and southern regions. In the north, where temperatures historically remain below the 25°C threshold even during the JJA season, the increase in the number of summer days is minimal (the threshold temperature of 25°C may be too high to be reached even in a world 2°C warmer than the preindustrial). In the southern Mediterranean Basin, the threshold is already surpassed throughout the entire JJA season in the historical period, and thus, no change is observed. This underscores the importance of incorporating both statistical and physical extreme indices for a comprehensive assessment of climate impacts.\n\n- The ensemble median spatially-averaged value shown through the boxplot indicates that with a 2°C global warming, the count of summer days ('SU') is projected to rise by approximately 10 days compared to the historical average for the JJA season. The interquartile range for the 'SU' index extends from around 6 to almost 11 days.\n\n- In a world 2°C warmer than the preindustrial baseline, the average number of days with maximum daily temperatures surpassing the reference 90th percentile (calculated for the historical period) during the JJA season exhibits spatial differences across Europe. Notably, values are higher for the Mediterranean Basin, consistent with the findings of Josep Cos et al. (2022) [[7]](https://doi.org/10.5194/esd-13-321-2022), who assessed the Mediterranean climate change hotspot using CMIP6 projections. The boxplot analysis for this index displays an ensemble median spatially-averaged value over Europe larger than 32 days for the JJA season, indicating that under a 2°C global warming level, one out of every three days will have a maximum daily temperature above the 90th percentile reference calculated for the historical period. The interquartile range spans approximately 28 to 33 days for the 'TX90p' index.\n\n- It is important to highlight that the boxplots reflect spatially-averaged values, and their outputs should be interpreted with caution, as the results may vary substantially when considering different regions across Europe.\n\n- What do the results mean for users? Are the biases from the historical notebook (\"CMIP6 Climate Projections: evaluating bias in extreme temperature indices for the reinsurance sector\") relevant? What are the differences compared to looking at the trend for a fixed future period?\n - For the CMIP6 models analysed, there is clear agreement on the sign of the projected climate signal, defined as the differences between the mean values of the indices calculated for a centered thirty-year period when the global mean temperature is 2°C higher than the preindustrial baseline and the mean values for the historical period from 1971 to 2000. This agreement suggests that it may be safe to include this information when designing strategies to optimise reinsurance protections in a world 2°C warmer than the preindustrial baseline.\n\n- The advantage of using this approach, instead of considering the trend for a fixed future period, is that some of the systematic biases of the models may be removed [[7]](https://doi.org/10.5194/esd-13-321-2022), as we are dealing with differences rather than absolute values. However, it is important to note that the centered thirty-year period when the global mean temperature reaches 2°C above the preindustrial baseline varies depending on the model considered, and determining this timing can be challenging.\n\n\n
\nRESULTS NOTE:
\nIt is important to note that the results presented are specific to the 16 models chosen, and users should aim to assess as wide a range of models as possible before making a sub-selection.", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-and-weather-extremes_q03 | Category: Climate_Projections\nSection: Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector > Analysis and results > 3. Plot and describe results > 3.5. Results summary and discussion\n---\n- The frequency of mean summer days occurring during the JJA season, defined as those with a daily maximum temperature exceeding 25°C, is projected to generally increase in a world 2°C warmer than the preindustrial baseline (1861-1890), compared to the average occurrences during the control period (1971-2000) across Europe, albeit with regional variations. Minor changes are expected in the northernmost and southern regions. In the north, where temperatures historically remain below the 25°C threshold even during the JJA season, the increase in the number of summer days is minimal (the threshold temperature of 25°C may be too high to be reached even in a world 2°C warmer than the preindustrial). In the southern Mediterranean Basin, the threshold is already surpassed throughout the entire JJA season in the historical period, and thus, no change is observed. This underscores the importance of incorporating both statistical and physical extreme indices for a comprehensive assessment of climate impacts.\n\n- The ensemble median spatially-averaged value shown through the boxplot indicates that with a 2°C global warming, the count of summer days ('SU') is projected to rise by approximately 10 days compared to the historical average for the JJA season. The interquartile range for the 'SU' index extends from around 6 to almost 11 days.\n\n- In a world 2°C warmer than the preindustrial baseline, the average number of days with maximum daily temperatures surpassing the reference 90th percentile (calculated for the historical period) during the JJA season exhibits spatial differences across Europe. Notably, values are higher for the Mediterranean Basin, consistent with the findings of Josep Cos et al. (2022) [[7]](https://doi.org/10.5194/esd-13-321-2022), who assessed the Mediterranean climate change hotspot using CMIP6 projections. The boxplot analysis for this index displays an ensemble median spatially-averaged value over Europe larger than 32 days for the JJA season, indicating that under a 2°C global warming level, one out of every three days will have a maximum daily temperature above the 90th percentile reference calculated for the historical period. The interquartile range spans approximately 28 to 33 days for the 'TX90p' index.\n\n- It is important to highlight that the boxplots reflect spatially-averaged values, and their outputs should be interpreted with caution, as the results may vary substantially when considering different regions across Europe.\n\n- What do the results mean for users? Are the biases from the historical notebook (\"CMIP6 Climate Projections: evaluating bias in extreme temperature indices for the reinsurance sector\") relevant? What are the differences compared to looking at the trend for a fixed future period?\n - For the CMIP6 models analysed, there is clear agreement on the sign of the projected climate signal, defined as the differences between the mean values of the indices calculated for a centered thirty-year period when the global mean temperature is 2°C higher than the preindustrial baseline and the mean values for the historical period from 1971 to 2000. This agreement suggests that it may be safe to include this information when designing strategies to optimise reinsurance protections in a world 2°C warmer than the preindustrial baseline.\n\n- The advantage of using this approach, instead of considering the trend for a fixed future period, is that some of the systematic biases of the models may be removed [[7]](https://doi.org/10.5194/esd-13-321-2022), as we are dealing with differences rather than absolute values. However, it is important to note that the centered thirty-year period when the global mean temperature reaches 2°C above the preindustrial baseline varies depending on the model considered, and determining this timing can be challenging.\n\n\n
\nRESULTS NOTE:
\nIt is important to note that the results presented are specific to the 16 models chosen, and users should aim to assess as wide a range of models as possible before making a sub-selection."} {"chunk_id": "climate_projections-cmip6_climate-and-weather-extremes_q03__869d14876599", "report_id": "climate_projections-cmip6_climate-and-weather-extremes_q03", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q03", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector > ℹ️ If you want to know more > Key resources", "title": "Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector", "chunk_index": 18, "token_count": 252, "text_raw": "Some key resources and further reading were linked throughout this assessment.\n\nThe CDS catalogue entries for the data used were:\n* CMIP6 climate projections (Daily - Daily maximum near-surface air temperature): https://cds.climate.copernicus.eu/datasets/projections-cmip6?tab=overview\n\nCode libraries used:\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)\n* [icclim](https://icclim.readthedocs.io/en/stable/) Python package", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-and-weather-extremes_q03 | Category: Climate_Projections\nSection: Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector > ℹ️ If you want to know more > Key resources\n---\nSome key resources and further reading were linked throughout this assessment.\n\nThe CDS catalogue entries for the data used were:\n* CMIP6 climate projections (Daily - Daily maximum near-surface air temperature): https://cds.climate.copernicus.eu/datasets/projections-cmip6?tab=overview\n\nCode libraries used:\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)\n* [icclim](https://icclim.readthedocs.io/en/stable/) Python package"} {"chunk_id": "climate_projections-cmip6_climate-and-weather-extremes_q03__b92362e20742", "report_id": "climate_projections-cmip6_climate-and-weather-extremes_q03", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q03", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector > ℹ️ If you want to know more > References", "title": "Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector", "chunk_index": 19, "token_count": 940, "text_raw": "[[1]](https://doi.org/10.3390/atmos11020146) Tesselaar, M., Botzen, W.J.W., Aerts, J.C.J.H. (2020). Impacts of Climate Change and Remote Natural Catastrophes on EU Flood Insurance Markets: An Analysis of Soft and Hard Reinsurance Markets for Flood Coverage. Atmosphere 2020, 11, 146. https://doi.org/10.3390/atmos11020146\n\n[[2]](https://doi.org/10.5194/nhess-22-659-2022) Rädler, A. T. (2022). Invited perspectives: how does climate change affect the risk of natural hazards? Challenges and step changes from the reinsurance perspective. Nat. Hazards Earth Syst. Sci., 22, 659–664. https://doi.org/10.5194/nhess-22-659-2022\n\n[[3]](https://www.ipcc.ch/site/assets/uploads/2018/03/wg2TARchap8.pdf) Vellinga, P., Mills, E., Bowers, L., Berz, G.A., Huq, S., Kozak, L.M., Paultikof, J., Schanzenbacker, B., Shida, S., Soler, G., Benson, C., Bidan, P., Bruce, J.W., Huyck, P.M., Lemcke, G., Peara, A., Radevsky, R., Schoubroeck, C.V., Dlugolecki, A.F. (2001). Insurance and other financial services. In J. J. McCarthy, O. F. Canziani, N. A. Leary, D. J. Dokken, & K. S. White (Eds.), Climate change 2001: impacts, adaptation, and vulnerability. Contribution of working group 2 to the third assessment report of the intergovernmental panel on climate change. (pp. 417-450). Cambridge University Press.\n\n[[4]](https://doi.org/10.1002/wcc.71) Lemos, M.C. and Rood, R.B. (2010). Climate projections and their impact on policy and practice. WIREs Clim Chg, 1: 670-682. https://doi.org/10.1002/wcc.71\n\n[[5]](https://doi.org/10.1002/wcc.579) Nissan H, Goddard L, de Perez EC, et al. (2019). On the use and misuse of climate change projections in international development. WIREs Clim Change, 10:e579. https://doi.org/10.1002/wcc.579\n\n[[6]](https://doi.org/10.1088/1748-9326/aab1b1) Nikulin, G., Lennard, C., Dosio, A., Kjellström, E., Chen, Y., Hänsler, A., Kupiainen, M., Laprise, R., Laura Mariotti, L., Maule C.F., et al. (2018). The effects of 1.5 and 2 degrees of global warming on Africa in the CORDEX ensemble. Environ. Res. Lett. 13 065003. https://doi.org/10.1088/1748-9326/aab1b1\n\n[[7]](https://doi.org/10.5194/esd-13-321-2022) Cos, J., Doblas-Reyes, F., Jury, M., Marcos, R., Bretonnière, P.-A., and Samsó, M. (2022). The Mediterranean climate change hotspot in the CMIP5 and CMIP6 projections, Earth Syst. Dynam., 13, 321–340. https://doi.org/10.5194/esd-13-321-2022", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-and-weather-extremes_q03 | Category: Climate_Projections\nSection: Uncertainty in projected changes in extreme temperature indices at a 2°C Global Warming Level for the reinsurance sector > ℹ️ If you want to know more > References\n---\n[[1]](https://doi.org/10.3390/atmos11020146) Tesselaar, M., Botzen, W.J.W., Aerts, J.C.J.H. (2020). Impacts of Climate Change and Remote Natural Catastrophes on EU Flood Insurance Markets: An Analysis of Soft and Hard Reinsurance Markets for Flood Coverage. Atmosphere 2020, 11, 146. https://doi.org/10.3390/atmos11020146\n\n[[2]](https://doi.org/10.5194/nhess-22-659-2022) Rädler, A. T. (2022). Invited perspectives: how does climate change affect the risk of natural hazards? Challenges and step changes from the reinsurance perspective. Nat. Hazards Earth Syst. Sci., 22, 659–664. https://doi.org/10.5194/nhess-22-659-2022\n\n[[3]](https://www.ipcc.ch/site/assets/uploads/2018/03/wg2TARchap8.pdf) Vellinga, P., Mills, E., Bowers, L., Berz, G.A., Huq, S., Kozak, L.M., Paultikof, J., Schanzenbacker, B., Shida, S., Soler, G., Benson, C., Bidan, P., Bruce, J.W., Huyck, P.M., Lemcke, G., Peara, A., Radevsky, R., Schoubroeck, C.V., Dlugolecki, A.F. (2001). Insurance and other financial services. In J. J. McCarthy, O. F. Canziani, N. A. Leary, D. J. Dokken, & K. S. White (Eds.), Climate change 2001: impacts, adaptation, and vulnerability. Contribution of working group 2 to the third assessment report of the intergovernmental panel on climate change. (pp. 417-450). Cambridge University Press.\n\n[[4]](https://doi.org/10.1002/wcc.71) Lemos, M.C. and Rood, R.B. (2010). Climate projections and their impact on policy and practice. WIREs Clim Chg, 1: 670-682. https://doi.org/10.1002/wcc.71\n\n[[5]](https://doi.org/10.1002/wcc.579) Nissan H, Goddard L, de Perez EC, et al. (2019). On the use and misuse of climate change projections in international development. WIREs Clim Change, 10:e579. https://doi.org/10.1002/wcc.579\n\n[[6]](https://doi.org/10.1088/1748-9326/aab1b1) Nikulin, G., Lennard, C., Dosio, A., Kjellström, E., Chen, Y., Hänsler, A., Kupiainen, M., Laprise, R., Laura Mariotti, L., Maule C.F., et al. (2018). The effects of 1.5 and 2 degrees of global warming on Africa in the CORDEX ensemble. Environ. Res. Lett. 13 065003. https://doi.org/10.1088/1748-9326/aab1b1\n\n[[7]](https://doi.org/10.5194/esd-13-321-2022) Cos, J., Doblas-Reyes, F., Jury, M., Marcos, R., Bretonnière, P.-A., and Samsó, M. (2022). The Mediterranean climate change hotspot in the CMIP5 and CMIP6 projections, Earth Syst. Dynam., 13, 321–340. https://doi.org/10.5194/esd-13-321-2022"} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q01__bc557a705f78", "report_id": "climate_projections-cmip6_climate-impact-indicators_q01", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q01", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Biases in energy-consumption-related indices in Europe", "title": "Biases in energy-consumption-related indices in Europe", "chunk_index": 0, "token_count": 96, "text_raw": "Production date: 22-05-2024\n\nProduced by: CMCC foundation - Euro-Mediterranean Center on Climate Change. Albert Martinez Boti.", "text_with_prefix": "EQC Quality Assessment: \"Biases in energy-consumption-related indices in Europe\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q01 | Category: Climate_Projections\nSection: Biases in energy-consumption-related indices in Europe\n---\nProduction date: 22-05-2024\n\nProduced by: CMCC foundation - Euro-Mediterranean Center on Climate Change. Albert Martinez Boti."} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q01__543102ab6d46", "report_id": "climate_projections-cmip6_climate-impact-indicators_q01", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q01", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Biases in energy-consumption-related indices in Europe > Quality assessment question", "title": "Biases in energy-consumption-related indices in Europe", "chunk_index": 1, "token_count": 641, "text_raw": "* **How well do the CMIP6 projections represent Energy Degree Days climatology and trend over a historic period?**\n\nSectors affected by climate change are varied including agriculture [[1]](https://doi.org/10.1007/s10113-010-0173-x), forest ecosystems [[2]](https://doi.org/10.1016/j.foreco.2009.09.023), and energy consumption [[3]](https://doi.org/10.1016/j.enbuild.2014.09.052). Under projected future global warming over Europe [[4]](https://doi.org/10.1007/s10113-013-0499-2)[[5]](http://hdl.handle.net/10013/epic.45156.d001), the current increase in energy demand is expected to persist until the end of this century and beyond [[6]](https://doi.org/10.1002/joc.5362). Identifying which climate-change-related impacts are likely to increase, by how much, and inherent regional patterns, is important for any effective strategy for managing future climate risks. This notebook utilises data from a subset of models from [CMIP6](https://cds.climate.copernicus.eu/datasets/projections-cmip6?tab=overview) Global Climate Models (GCMs) and compares them with [ERA5](https://cds.climate.copernicus.eu/datasets/reanalysis-era5-complete?tab=doc) reanalysis, serving as the reference product. Two energy-consumption-related indices are calculated from daily mean temperatures using the [icclim](https://icclim.readthedocs.io/en/stable/) Python package: Cooling Degree Days (CDDs) and Heating Degree Days (HDDs). Degree days measure how much warmer or colder it is compared to standard temperatures (usually 15.5°C for heating and 22°C for cooling). Higher degree day numbers indicate more extreme temperatures, which typically lead to increased energy use for heating or cooling buildings. In the presented code, CDD calculations use summer aggregation (CDD22), while HDD calculations focus on winter (HDD15.5), presenting results as daily averages rather than cumulative values. Within this notebook, these calculations are performed over the historical period spanning from 1971 to 2000. It is important to note that the results presented here pertain to a specific subset of the CMIP6 ensemble and may not be generalisable to the entire dataset. Also note that a separate assessment examines the representation of trends of these indices for the same models during a fixed future period (2015-2099), while another assessment looks at the projected climate signal of these indices for the same models at a 2°C Global Warming Level.", "text_with_prefix": "EQC Quality Assessment: \"Biases in energy-consumption-related indices in Europe\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q01 | Category: Climate_Projections\nSection: Biases in energy-consumption-related indices in Europe > Quality assessment question\n---\n* **How well do the CMIP6 projections represent Energy Degree Days climatology and trend over a historic period?**\n\nSectors affected by climate change are varied including agriculture [[1]](https://doi.org/10.1007/s10113-010-0173-x), forest ecosystems [[2]](https://doi.org/10.1016/j.foreco.2009.09.023), and energy consumption [[3]](https://doi.org/10.1016/j.enbuild.2014.09.052). Under projected future global warming over Europe [[4]](https://doi.org/10.1007/s10113-013-0499-2)[[5]](http://hdl.handle.net/10013/epic.45156.d001), the current increase in energy demand is expected to persist until the end of this century and beyond [[6]](https://doi.org/10.1002/joc.5362). Identifying which climate-change-related impacts are likely to increase, by how much, and inherent regional patterns, is important for any effective strategy for managing future climate risks. This notebook utilises data from a subset of models from [CMIP6](https://cds.climate.copernicus.eu/datasets/projections-cmip6?tab=overview) Global Climate Models (GCMs) and compares them with [ERA5](https://cds.climate.copernicus.eu/datasets/reanalysis-era5-complete?tab=doc) reanalysis, serving as the reference product. Two energy-consumption-related indices are calculated from daily mean temperatures using the [icclim](https://icclim.readthedocs.io/en/stable/) Python package: Cooling Degree Days (CDDs) and Heating Degree Days (HDDs). Degree days measure how much warmer or colder it is compared to standard temperatures (usually 15.5°C for heating and 22°C for cooling). Higher degree day numbers indicate more extreme temperatures, which typically lead to increased energy use for heating or cooling buildings. In the presented code, CDD calculations use summer aggregation (CDD22), while HDD calculations focus on winter (HDD15.5), presenting results as daily averages rather than cumulative values. Within this notebook, these calculations are performed over the historical period spanning from 1971 to 2000. It is important to note that the results presented here pertain to a specific subset of the CMIP6 ensemble and may not be generalisable to the entire dataset. Also note that a separate assessment examines the representation of trends of these indices for the same models during a fixed future period (2015-2099), while another assessment looks at the projected climate signal of these indices for the same models at a 2°C Global Warming Level."} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q01__d2ab4cff8735", "report_id": "climate_projections-cmip6_climate-impact-indicators_q01", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q01", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Biases in energy-consumption-related indices in Europe > Quality assessment statement", "title": "Biases in energy-consumption-related indices in Europe", "chunk_index": 2, "token_count": 693, "text_raw": "These are the key outcomes of this assessment\n\n* CMIP6 projections provide valuable insights into trends and climatology of Energy Degree Days across Europe, though with inherent uncertainties, and differences compared to the reference dataset (ERA5). During the historical period (1971-2000), GCMs exhibited biases in capturing both mean values and trends for energy-consumption-related indices.\n\n* For the 16 models considered, biases were observed in mean values of Heating Degree Days ('HDD15.5') over DJF and Cooling Degree Days ('CDD22') over JJA. Specifically, for most of the models, there is an overestimation of HDD15.5 mean values for DJF (except in regions with complex orography where there was underestimation) and CDD22 mean values for JJA across most of the Mediterranean Basin.\n\n* The subset of CMIP6 models show a decrease in the energy needed for heating during winter within the historical period across Europe, especially in the eastern and northern regions. This aligns with results obtained using the ERA5 reference product, except in the Balkans where there is an increase in required heating energy during this period. The ERA5 reference product also shows an increase of the amount of energy needed for cooling buildings in summer during this period, particularly across the Mediterranean Basin. This trend is well captured by the subset of CMIP6 models. Depending on the region, these results may enhance confidence in using these models for analysing future trends, although they do not guarantee accuracy.\n\n* Despite the biases and high inter-model spread in some regions, the outcomes of this notebook offer valuable insights for decisions sensitive to future energy demand. These results may enhance confidence in using these models to analyse future trends, although their accuracy is not assured. To improve the accuracy of these insights, biases should be considered and corrected [[6]](https://doi.org/10.1002/joc.5362)[[7]](https://doi.org/10.1038/s41467-021-25504-8). \n```\n\nattachment:83f3529b-f7ec-456a-aab1-9dee68ba895b.png \n---\nalt: Mean Bias HDD\nwidth: 900px\n---\nHeating Degree Days daily average calculated using the winter comfort threshold of 15.5°C ('HDD15.5') for the temporal aggreggtion of 'DJF'. Mean bias for the historical period (1971 - 2000) of each individual CMIP6 model. The colorbar chosen for representing biases for the HDD15.5 index ranges from red to blue. In this color scheme, negative bias values (red colors) indicate that the Heating Degree Days displayed by the models are lower than those shown by ERA5. Instead, blueish colors denote more Heating Degree Days, indicating a positive bias compared to ERA5. This selection is based on the rationale that more Heating Degree Days are associated with colder conditions, typically represented by blueish colors, while fewer Heating Degree Days imply warmer conditions, depicted by reddish colors.\n```", "text_with_prefix": "EQC Quality Assessment: \"Biases in energy-consumption-related indices in Europe\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q01 | Category: Climate_Projections\nSection: Biases in energy-consumption-related indices in Europe > Quality assessment statement\n---\nThese are the key outcomes of this assessment\n\n* CMIP6 projections provide valuable insights into trends and climatology of Energy Degree Days across Europe, though with inherent uncertainties, and differences compared to the reference dataset (ERA5). During the historical period (1971-2000), GCMs exhibited biases in capturing both mean values and trends for energy-consumption-related indices.\n\n* For the 16 models considered, biases were observed in mean values of Heating Degree Days ('HDD15.5') over DJF and Cooling Degree Days ('CDD22') over JJA. Specifically, for most of the models, there is an overestimation of HDD15.5 mean values for DJF (except in regions with complex orography where there was underestimation) and CDD22 mean values for JJA across most of the Mediterranean Basin.\n\n* The subset of CMIP6 models show a decrease in the energy needed for heating during winter within the historical period across Europe, especially in the eastern and northern regions. This aligns with results obtained using the ERA5 reference product, except in the Balkans where there is an increase in required heating energy during this period. The ERA5 reference product also shows an increase of the amount of energy needed for cooling buildings in summer during this period, particularly across the Mediterranean Basin. This trend is well captured by the subset of CMIP6 models. Depending on the region, these results may enhance confidence in using these models for analysing future trends, although they do not guarantee accuracy.\n\n* Despite the biases and high inter-model spread in some regions, the outcomes of this notebook offer valuable insights for decisions sensitive to future energy demand. These results may enhance confidence in using these models to analyse future trends, although their accuracy is not assured. To improve the accuracy of these insights, biases should be considered and corrected [[6]](https://doi.org/10.1002/joc.5362)[[7]](https://doi.org/10.1038/s41467-021-25504-8). \n```\n\nattachment:83f3529b-f7ec-456a-aab1-9dee68ba895b.png \n---\nalt: Mean Bias HDD\nwidth: 900px\n---\nHeating Degree Days daily average calculated using the winter comfort threshold of 15.5°C ('HDD15.5') for the temporal aggreggtion of 'DJF'. Mean bias for the historical period (1971 - 2000) of each individual CMIP6 model. The colorbar chosen for representing biases for the HDD15.5 index ranges from red to blue. In this color scheme, negative bias values (red colors) indicate that the Heating Degree Days displayed by the models are lower than those shown by ERA5. Instead, blueish colors denote more Heating Degree Days, indicating a positive bias compared to ERA5. This selection is based on the rationale that more Heating Degree Days are associated with colder conditions, typically represented by blueish colors, while fewer Heating Degree Days imply warmer conditions, depicted by reddish colors.\n```"} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q01__cad9b099184e", "report_id": "climate_projections-cmip6_climate-impact-indicators_q01", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q01", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Biases in energy-consumption-related indices in Europe > Methodology", "title": "Biases in energy-consumption-related indices in Europe", "chunk_index": 3, "token_count": 975, "text_raw": "The reference methodology used here for the indices calculation is similar to the one followed by Scoccimarro et al., (2023) [[8]](https://doi.org/10.1038/s43247-023-00878-3). However the thermal comfort thresholds used in this notebook are slightly different. A winter comfort temperature of 15.5°C and a summer comfort temperature of 22.0°C are used here (as in the [CDS application](https://dast.copernicus-climate.eu/documents/app-heating-cooling-degree-days/C3S_EEA_HDD_CDD_application_user_guide_v0.9.pdf)). In the presented code, the CDD calculations are based on the JJA aggregation, with a comfort temperature of 22°C (CDD22), while HDD calculations focus on winter (DJF) with a comfort temperature of 15.5°C (HDD15.5). More specifically, to calculate CDD22, the sum of the differences between the daily mean temperature and the thermal comfort temperature of 22°C is computed. This calculation occurs only when the mean temperature is above the thermal comfort level; otherwise, the CDD22 for that day is set to 0. For example, a day with a mean temperature of 28°C would result in 6°C. Two consecutive hot days like this would total 12°C over the two-day period. Similarly, to calculate HDD15.5, the sum of the differences between the thermal comfort temperature of 15.5°C and the daily mean temperature is determined. This happens only when the mean temperature is below the thermal comfort level; otherwise, the HDD15.5 for that day is set to 0. Finally, to obtain more intuitive values, the sum is averaged over the number of days in the season to produce daily average values. This approach differs from the [CDS application](https://dast.copernicus-climate.eu/documents/app-heating-cooling-degree-days/C3S_EEA_HDD_CDD_application_user_guide_v0.9.pdf), where both the sum over a period and the daily average values can be displayed. In Spinoni et al., (2018) [[6]](https://doi.org/10.1002/joc.5362), as well as in the [CDS application](https://dast.copernicus-climate.eu/documents/app-heating-cooling-degree-days/C3S_EEA_HDD_CDD_application_user_guide_v0.9.pdf), more advanced methods for calculating CDD and HDD involve considering maximum, minimum, and mean temperatures. However, to prevent overloading the notebook and maintain simplicity while ensuring compatibility with the [icclim](https://icclim.readthedocs.io/en/stable/) Python package, we opted to utilise a single variable (2m mean temperature).\n\nThis notebook provides an assessment of the systematic errors (trend and climate mean) in a subset of 16 models from [CMIP6](https://cds.climate.copernicus.eu/datasets/projections-cmip6?tab=overview). It achieves this by comparing the model predictions with the ERA5 reanalysis for the energy-consumption-related indices 'HDD15.5' and 'CDD22' for the historical period spanning from 1971 to 2000. In particular, spatial patterns of climate mean and trend, along with biases, are examined and displayed for each model and the ensemble median (calculated for each grid cell). Additionally, spatially-averaged trend values are analysed and presented using box plots to provide an overview of trend behavior across the distribution of the chosen subset of models when averaged across Europe.\n\nThe analysis and results follow the next outline:\n\n**[](section-1)**\n * [](section-1.1)\n * [](section-1.2)\n * [](section-1.3)\n * [](section-1.4)\n * [](section-1.5)\n * [](section-1.6)\n\n**[](section-2)**\n * [](section-2.1)\n * [](section-2.2)\n * [](section-2.3)\n\n**[](section-3)**\n * [](section-3.1)\n * [](section-3.2)\n * [](section-3.3)\n * [](section-3.4)\n * [](section-3.5)\n * [](section-3.6)", "text_with_prefix": "EQC Quality Assessment: \"Biases in energy-consumption-related indices in Europe\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q01 | Category: Climate_Projections\nSection: Biases in energy-consumption-related indices in Europe > Methodology\n---\nThe reference methodology used here for the indices calculation is similar to the one followed by Scoccimarro et al., (2023) [[8]](https://doi.org/10.1038/s43247-023-00878-3). However the thermal comfort thresholds used in this notebook are slightly different. A winter comfort temperature of 15.5°C and a summer comfort temperature of 22.0°C are used here (as in the [CDS application](https://dast.copernicus-climate.eu/documents/app-heating-cooling-degree-days/C3S_EEA_HDD_CDD_application_user_guide_v0.9.pdf)). In the presented code, the CDD calculations are based on the JJA aggregation, with a comfort temperature of 22°C (CDD22), while HDD calculations focus on winter (DJF) with a comfort temperature of 15.5°C (HDD15.5). More specifically, to calculate CDD22, the sum of the differences between the daily mean temperature and the thermal comfort temperature of 22°C is computed. This calculation occurs only when the mean temperature is above the thermal comfort level; otherwise, the CDD22 for that day is set to 0. For example, a day with a mean temperature of 28°C would result in 6°C. Two consecutive hot days like this would total 12°C over the two-day period. Similarly, to calculate HDD15.5, the sum of the differences between the thermal comfort temperature of 15.5°C and the daily mean temperature is determined. This happens only when the mean temperature is below the thermal comfort level; otherwise, the HDD15.5 for that day is set to 0. Finally, to obtain more intuitive values, the sum is averaged over the number of days in the season to produce daily average values. This approach differs from the [CDS application](https://dast.copernicus-climate.eu/documents/app-heating-cooling-degree-days/C3S_EEA_HDD_CDD_application_user_guide_v0.9.pdf), where both the sum over a period and the daily average values can be displayed. In Spinoni et al., (2018) [[6]](https://doi.org/10.1002/joc.5362), as well as in the [CDS application](https://dast.copernicus-climate.eu/documents/app-heating-cooling-degree-days/C3S_EEA_HDD_CDD_application_user_guide_v0.9.pdf), more advanced methods for calculating CDD and HDD involve considering maximum, minimum, and mean temperatures. However, to prevent overloading the notebook and maintain simplicity while ensuring compatibility with the [icclim](https://icclim.readthedocs.io/en/stable/) Python package, we opted to utilise a single variable (2m mean temperature).\n\nThis notebook provides an assessment of the systematic errors (trend and climate mean) in a subset of 16 models from [CMIP6](https://cds.climate.copernicus.eu/datasets/projections-cmip6?tab=overview). It achieves this by comparing the model predictions with the ERA5 reanalysis for the energy-consumption-related indices 'HDD15.5' and 'CDD22' for the historical period spanning from 1971 to 2000. In particular, spatial patterns of climate mean and trend, along with biases, are examined and displayed for each model and the ensemble median (calculated for each grid cell). Additionally, spatially-averaged trend values are analysed and presented using box plots to provide an overview of trend behavior across the distribution of the chosen subset of models when averaged across Europe.\n\nThe analysis and results follow the next outline:\n\n**[](section-1)**\n * [](section-1.1)\n * [](section-1.2)\n * [](section-1.3)\n * [](section-1.4)\n * [](section-1.5)\n * [](section-1.6)\n\n**[](section-2)**\n * [](section-2.1)\n * [](section-2.2)\n * [](section-2.3)\n\n**[](section-3)**\n * [](section-3.1)\n * [](section-3.2)\n * [](section-3.3)\n * [](section-3.4)\n * [](section-3.5)\n * [](section-3.6)"} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q01__82ca9d2fcc21", "report_id": "climate_projections-cmip6_climate-impact-indicators_q01", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q01", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Biases in energy-consumption-related indices in Europe > Analysis and results > 1. Parameters, requests and functions definition > 1.2. Define Parameters", "title": "Biases in energy-consumption-related indices in Europe", "chunk_index": 4, "token_count": 371, "text_raw": "In the \"Define Parameters\" section, various customisable options for the notebook are specified:\n\n- The initial and ending year used for the historical period can be specified by changing the parameters `year_start` and `year_stop` (1971-2000 is chosen for consistency between CORDEX and CMIP6).\n- `index_timeseries` is a dictionary that set the temporal aggregation for every index considered within this notebook ('HDD15.5' and 'CDD22'). In the presented code, the CDD calculations are always based on the JJA aggregation, with a comfort temperature of 22°C (CDD22), while HDD calculations focus on winter (DJF) with a comfort temperature of 15.5°C (HDD15.5).\n- `collection_id` sets the family of models. Only CMIP6 is implemented for this notebook.\n- `area` allows specifying the geographical domain of interest.\n- The `interpolation_method` parameter allows selecting the interpolation method when regridding is performed over the indices.\n- The `chunk` selection allows the user to define if dividing into chunks when downloading the data on their local machine. Although it does not significantly affect the analysis, it is recommended to keep the default value for optimal performance.\n\nChoose annual or seasonal timeseries\nInterpolation method\nSelect the family of models\nArea to show\nChunks for download\n\n(section-1.3)=", "text_with_prefix": "EQC Quality Assessment: \"Biases in energy-consumption-related indices in Europe\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q01 | Category: Climate_Projections\nSection: Biases in energy-consumption-related indices in Europe > Analysis and results > 1. Parameters, requests and functions definition > 1.2. Define Parameters\n---\nIn the \"Define Parameters\" section, various customisable options for the notebook are specified:\n\n- The initial and ending year used for the historical period can be specified by changing the parameters `year_start` and `year_stop` (1971-2000 is chosen for consistency between CORDEX and CMIP6).\n- `index_timeseries` is a dictionary that set the temporal aggregation for every index considered within this notebook ('HDD15.5' and 'CDD22'). In the presented code, the CDD calculations are always based on the JJA aggregation, with a comfort temperature of 22°C (CDD22), while HDD calculations focus on winter (DJF) with a comfort temperature of 15.5°C (HDD15.5).\n- `collection_id` sets the family of models. Only CMIP6 is implemented for this notebook.\n- `area` allows specifying the geographical domain of interest.\n- The `interpolation_method` parameter allows selecting the interpolation method when regridding is performed over the indices.\n- The `chunk` selection allows the user to define if dividing into chunks when downloading the data on their local machine. Although it does not significantly affect the analysis, it is recommended to keep the default value for optimal performance.\n\nChoose annual or seasonal timeseries\nInterpolation method\nSelect the family of models\nArea to show\nChunks for download\n\n(section-1.3)="} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q01__9d498880dc31", "report_id": "climate_projections-cmip6_climate-impact-indicators_q01", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q01", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Biases in energy-consumption-related indices in Europe > Analysis and results > 1. Parameters, requests and functions definition > 1.3. Define models", "title": "Biases in energy-consumption-related indices in Europe", "chunk_index": 5, "token_count": 156, "text_raw": "The following climate analyses are performed considering a subset of GCMs from CMIP6. Models names are listed in the parameters below. Some variable-dependent parameters are also selected.\n\nThe selected CMIP6 models have available both experiments the historical and the SSP8.5.\n\nDefine models\nDefine dictionaries to use in titles and caption\n\n(section-1.4)=", "text_with_prefix": "EQC Quality Assessment: \"Biases in energy-consumption-related indices in Europe\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q01 | Category: Climate_Projections\nSection: Biases in energy-consumption-related indices in Europe > Analysis and results > 1. Parameters, requests and functions definition > 1.3. Define models\n---\nThe following climate analyses are performed considering a subset of GCMs from CMIP6. Models names are listed in the parameters below. Some variable-dependent parameters are also selected.\n\nThe selected CMIP6 models have available both experiments the historical and the SSP8.5.\n\nDefine models\nDefine dictionaries to use in titles and caption\n\n(section-1.4)="} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q01__4382fcb98ca8", "report_id": "climate_projections-cmip6_climate-impact-indicators_q01", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q01", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Biases in energy-consumption-related indices in Europe > Analysis and results > 1. Parameters, requests and functions definition > 1.4. Define ERA5 request", "title": "Biases in energy-consumption-related indices in Europe", "chunk_index": 6, "token_count": 123, "text_raw": "Within this notebook, ERA5 serves as the reference product. In this section, we set the required parameters for the cds-api data-request of ERA5.\n\n(section-1.5)=", "text_with_prefix": "EQC Quality Assessment: \"Biases in energy-consumption-related indices in Europe\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q01 | Category: Climate_Projections\nSection: Biases in energy-consumption-related indices in Europe > Analysis and results > 1. Parameters, requests and functions definition > 1.4. Define ERA5 request\n---\nWithin this notebook, ERA5 serves as the reference product. In this section, we set the required parameters for the cds-api data-request of ERA5.\n\n(section-1.5)="} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q01__fbd34ac5bc13", "report_id": "climate_projections-cmip6_climate-impact-indicators_q01", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q01", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Biases in energy-consumption-related indices in Europe > Analysis and results > 1. Parameters, requests and functions definition > 1.5. Define model requests", "title": "Biases in energy-consumption-related indices in Europe", "chunk_index": 7, "token_count": 151, "text_raw": "In this section we set the required parameters for the cds-api data-request.\n\nWhen `weights = True`, spatial weighting is applied for calculations requiring spatial data aggregation. This is particularly relevant for CMIP6 GCMs with regular lon-lat grids that do not consider varying surface extensions at different latitudes.\n\n(section-1.6)=", "text_with_prefix": "EQC Quality Assessment: \"Biases in energy-consumption-related indices in Europe\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q01 | Category: Climate_Projections\nSection: Biases in energy-consumption-related indices in Europe > Analysis and results > 1. Parameters, requests and functions definition > 1.5. Define model requests\n---\nIn this section we set the required parameters for the cds-api data-request.\n\nWhen `weights = True`, spatial weighting is applied for calculations requiring spatial data aggregation. This is particularly relevant for CMIP6 GCMs with regular lon-lat grids that do not consider varying surface extensions at different latitudes.\n\n(section-1.6)="} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q01__5da7fbbca294", "report_id": "climate_projections-cmip6_climate-impact-indicators_q01", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q01", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Biases in energy-consumption-related indices in Europe > Analysis and results > 1. Parameters, requests and functions definition > 1.6. Functions to cache", "title": "Biases in energy-consumption-related indices in Europe", "chunk_index": 8, "token_count": 306, "text_raw": "In this section, functions that will be executed in the caching phase are defined. Caching is the process of storing copies of files in a temporary storage location, so that they can be accessed more quickly. This process also checks if the user has already downloaded a file, avoiding redundant downloads.\n\nFunctions description:\n\n- The `select_timeseries` function subsets the dataset based on the chosen `timeseries` parameter.\n\n- The `compute_indices` function utilises the icclim package to calculate the energy-consumption-related indices.\n\n- The `compute_trends` function employs the Mann-Kendall test for trend calculation.\n\n- Finally, the `compute_indices_and_trends` computes the daily mean temperature (only if we are dealing with ERA5), calculates the energy consumption-related indices for the corresponding temporal aggregation using the `compute_indices` function, determines the indices mean over the historical period (1971-2000), obtain the trends using the `compute_trends` function, and offers an option for regridding to ERA5 if required.\n\nOriginal bounds for conservative interpolation\n\n(section-2)=", "text_with_prefix": "EQC Quality Assessment: \"Biases in energy-consumption-related indices in Europe\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q01 | Category: Climate_Projections\nSection: Biases in energy-consumption-related indices in Europe > Analysis and results > 1. Parameters, requests and functions definition > 1.6. Functions to cache\n---\nIn this section, functions that will be executed in the caching phase are defined. Caching is the process of storing copies of files in a temporary storage location, so that they can be accessed more quickly. This process also checks if the user has already downloaded a file, avoiding redundant downloads.\n\nFunctions description:\n\n- The `select_timeseries` function subsets the dataset based on the chosen `timeseries` parameter.\n\n- The `compute_indices` function utilises the icclim package to calculate the energy-consumption-related indices.\n\n- The `compute_trends` function employs the Mann-Kendall test for trend calculation.\n\n- Finally, the `compute_indices_and_trends` computes the daily mean temperature (only if we are dealing with ERA5), calculates the energy consumption-related indices for the corresponding temporal aggregation using the `compute_indices` function, determines the indices mean over the historical period (1971-2000), obtain the trends using the `compute_trends` function, and offers an option for regridding to ERA5 if required.\n\nOriginal bounds for conservative interpolation\n\n(section-2)="} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q01__7696e94de3d0", "report_id": "climate_projections-cmip6_climate-impact-indicators_q01", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q01", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Biases in energy-consumption-related indices in Europe > Analysis and results > 2. Downloading and processing > 2.1. Download and transform ERA5", "title": "Biases in energy-consumption-related indices in Europe", "chunk_index": 9, "token_count": 191, "text_raw": "In this section, the `download.download_and_transform` function from the 'c3s_eqc_automatic_quality_control' package is employed to download ERA5 reference data, obtain the daily mean temperature from hourly data, compute the energy-consumption-related indices for the selected temporal aggregation (\"DJF\" for HDD and \"JJA\" for CDD), calculate the mean and trend over the historical period (1971-2000) and cache the result (to avoid redundant downloads and processing).\n\n(section-2.2)=", "text_with_prefix": "EQC Quality Assessment: \"Biases in energy-consumption-related indices in Europe\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q01 | Category: Climate_Projections\nSection: Biases in energy-consumption-related indices in Europe > Analysis and results > 2. Downloading and processing > 2.1. Download and transform ERA5\n---\nIn this section, the `download.download_and_transform` function from the 'c3s_eqc_automatic_quality_control' package is employed to download ERA5 reference data, obtain the daily mean temperature from hourly data, compute the energy-consumption-related indices for the selected temporal aggregation (\"DJF\" for HDD and \"JJA\" for CDD), calculate the mean and trend over the historical period (1971-2000) and cache the result (to avoid redundant downloads and processing).\n\n(section-2.2)="} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q01__9d2e6f87ea11", "report_id": "climate_projections-cmip6_climate-impact-indicators_q01", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q01", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Biases in energy-consumption-related indices in Europe > Analysis and results > 2. Downloading and processing > 2.2. Download and transform models", "title": "Biases in energy-consumption-related indices in Europe", "chunk_index": 10, "token_count": 363, "text_raw": "In this section, the `download.download_and_transform` function from the 'c3s_eqc_automatic_quality_control' package is employed to download daily data from the CMIP6 models, compute the energy-consumption-related indices for the selected temporal aggregation (\"DJF\" for HDD and \"JJA\" for CDD), calculate the mean and trend over the historical period (1971-2000), interpolate to ERA5's grid (only for the cases in which it is specified, in the other cases, the original model's grid is mantained), and cache the result (to avoid redundant downloads and processing).\n\nOriginal model\nInterpolated model\n\n```text\nmodel='access_cm2'\nmodel='awi_cm_1_1_mr'\nmodel='cmcc_esm2'\nmodel='cnrm_cm6_1_hr'\nmodel='cnrm_esm2_1'\nmodel='ec_earth3_cc'\nmodel='gfdl_esm4'\nmodel='inm_cm4_8'\nmodel='inm_cm5_0'\nmodel='kiost_esm'\nmodel='mpi_esm1_2_lr'\nmodel='miroc6'\nmodel='miroc_es2l'\nmodel='mri_esm2_0'\nmodel='noresm2_mm'\nmodel='nesm3'\n```\n\n(section-2.3)=", "text_with_prefix": "EQC Quality Assessment: \"Biases in energy-consumption-related indices in Europe\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q01 | Category: Climate_Projections\nSection: Biases in energy-consumption-related indices in Europe > Analysis and results > 2. Downloading and processing > 2.2. Download and transform models\n---\nIn this section, the `download.download_and_transform` function from the 'c3s_eqc_automatic_quality_control' package is employed to download daily data from the CMIP6 models, compute the energy-consumption-related indices for the selected temporal aggregation (\"DJF\" for HDD and \"JJA\" for CDD), calculate the mean and trend over the historical period (1971-2000), interpolate to ERA5's grid (only for the cases in which it is specified, in the other cases, the original model's grid is mantained), and cache the result (to avoid redundant downloads and processing).\n\nOriginal model\nInterpolated model\n\n```text\nmodel='access_cm2'\nmodel='awi_cm_1_1_mr'\nmodel='cmcc_esm2'\nmodel='cnrm_cm6_1_hr'\nmodel='cnrm_esm2_1'\nmodel='ec_earth3_cc'\nmodel='gfdl_esm4'\nmodel='inm_cm4_8'\nmodel='inm_cm5_0'\nmodel='kiost_esm'\nmodel='mpi_esm1_2_lr'\nmodel='miroc6'\nmodel='miroc_es2l'\nmodel='mri_esm2_0'\nmodel='noresm2_mm'\nmodel='nesm3'\n```\n\n(section-2.3)="} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q01__7f450810a8f3", "report_id": "climate_projections-cmip6_climate-impact-indicators_q01", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q01", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Biases in energy-consumption-related indices in Europe > Analysis and results > 2. Downloading and processing > 2.3. Apply land-sea mask, change attributes and cut the region to show", "title": "Biases in energy-consumption-related indices in Europe", "chunk_index": 11, "token_count": 260, "text_raw": "This section performs the following tasks:\n\n1. Cut the region of interest.\n2. Downloads the sea mask for ERA5.\n3. Applies the sea mask to both ERA5 data and the model data, which were previously regridded to ERA5's grid (i.e., it applies the ERA5 sea mask to `ds_interpolated`).\n4. Regrids the ERA5 land-sea mask to the model's grid and applies it to them.\n5. Change some variable attributes for plotting purposes.\n\n**Note:** `ds_interpolated` contains data from the models (mean and trend over the historical period, p-value of the trends...) regridded to ERA5. `model_datasets` contain the same data but in the original grid of each model.\n\nEdit attributes\n\n(section-3)=", "text_with_prefix": "EQC Quality Assessment: \"Biases in energy-consumption-related indices in Europe\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q01 | Category: Climate_Projections\nSection: Biases in energy-consumption-related indices in Europe > Analysis and results > 2. Downloading and processing > 2.3. Apply land-sea mask, change attributes and cut the region to show\n---\nThis section performs the following tasks:\n\n1. Cut the region of interest.\n2. Downloads the sea mask for ERA5.\n3. Applies the sea mask to both ERA5 data and the model data, which were previously regridded to ERA5's grid (i.e., it applies the ERA5 sea mask to `ds_interpolated`).\n4. Regrids the ERA5 land-sea mask to the model's grid and applies it to them.\n5. Change some variable attributes for plotting purposes.\n\n**Note:** `ds_interpolated` contains data from the models (mean and trend over the historical period, p-value of the trends...) regridded to ERA5. `model_datasets` contain the same data but in the original grid of each model.\n\nEdit attributes\n\n(section-3)="} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q01__a2e98db1f973", "report_id": "climate_projections-cmip6_climate-impact-indicators_q01", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q01", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Biases in energy-consumption-related indices in Europe > Analysis and results > 3. Plot and describe results", "title": "Biases in energy-consumption-related indices in Europe", "chunk_index": 12, "token_count": 315, "text_raw": "This section will display the following results:\n\n- Maps representing the spatial distribution of the **historical mean values** (1971-2000) of the indices 'HDD15.5' and 'CDD22' for ERA5, each model individually, the ensemble median (understood as the median of the mean values of the chosen subset of models calculated for each grid cell), and the ensemble spread (derived as the standard deviation of the distribution of the chosen subset of models).\n- Maps representing the spatial distribution of the **historical trends** (1971-2000) of the considered indices. Similar to the first analysis, this includes ERA5, each model individually, the ensemble median (understood as the median of the trend values of the chosen subset of models calculated for each grid cell), and the ensemble spread (derived as the standard deviation of the distribution of the chosen subset of models).\n- **Bias maps of the historical mean values**.\n- **Trend bias maps**. \n- **Boxplots** representing statistical distributions (PDF) built on the spatially-averaged historical trends from each considered model, displayed together with ERA5.\n\n(section-3.1)=", "text_with_prefix": "EQC Quality Assessment: \"Biases in energy-consumption-related indices in Europe\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q01 | Category: Climate_Projections\nSection: Biases in energy-consumption-related indices in Europe > Analysis and results > 3. Plot and describe results\n---\nThis section will display the following results:\n\n- Maps representing the spatial distribution of the **historical mean values** (1971-2000) of the indices 'HDD15.5' and 'CDD22' for ERA5, each model individually, the ensemble median (understood as the median of the mean values of the chosen subset of models calculated for each grid cell), and the ensemble spread (derived as the standard deviation of the distribution of the chosen subset of models).\n- Maps representing the spatial distribution of the **historical trends** (1971-2000) of the considered indices. Similar to the first analysis, this includes ERA5, each model individually, the ensemble median (understood as the median of the trend values of the chosen subset of models calculated for each grid cell), and the ensemble spread (derived as the standard deviation of the distribution of the chosen subset of models).\n- **Bias maps of the historical mean values**.\n- **Trend bias maps**. \n- **Boxplots** representing statistical distributions (PDF) built on the spatially-averaged historical trends from each considered model, displayed together with ERA5.\n\n(section-3.1)="} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q01__0f2ea2e0aa06", "report_id": "climate_projections-cmip6_climate-impact-indicators_q01", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q01", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Biases in energy-consumption-related indices in Europe > Analysis and results > 3. Plot and describe results > 3.1. Define plotting functions", "title": "Biases in energy-consumption-related indices in Europe", "chunk_index": 13, "token_count": 828, "text_raw": "The functions presented here are used to plot the mean values and trends calculated over the historical period (1971-2000) for each of the indices ('HDD15.5' and 'CDD22').\n\nFor a selected index, three layout types can be displayed, depending on the chosen function:\n\n1. Layout including the reference ERA5 product, the ensemble median, the bias of the ensemble median, and the ensemble spread: `plot_ensemble()` is used.\n2. Layout including every model (for the trend and mean values): `plot_models()` is employed.\n3. Layout including the bias of every model (for the trend and mean values): `plot_models()` is used again.\n\n`trend==True` argument allows displaying trend values over the historical period, while `trend==False` will show mean values. When the `trend` argument is set to `True`, regions with no significance are hatched. For individual models and ERA5, a grid point is considered to have a statistically significant trend when the p-value is lower than 0.05 (in such cases, no hatching is shown). However, for determining trend significance for the ensemble median (understood as the median of the trend values of the chosen subset of models calculated for each grid cell), reliance is placed on agreement categories, following the advanced approach proposed in AR6 [IPCC](https://www.ipcc.ch/report/ar6/wg1/downloads/report/IPCC_AR6_WGI_Atlas.pdf) on pages 1945-1950. The `hatch_p_value_ensemble()` function is used to distinguish, for each grid point, between three possible cases:\n\n1. If more than 66% of the models are statistically significant (p-value < 0.05) and more than 80% of the models share the same sign, we consider the ensemble median trend to be statistically significant, and there is agreement on the sign. To represent this, no hatching is used.\n2. If less than 66% of the models are statistically significant, regardless of agreement on the sign of the trend, hatching is applied (indicating that the ensemble median trend is not statistically significant).\n3. If more than 66% of the models are statistically significant but less than 80% of the models share the same sign, we consider the ensemble median trend to be statistically significant, but there is no agreement on the sign of the trend. This is represented using crosses.\n\nDefine function to plot the caption of the figures (for the ensemble case)\nAdd caption to the figure\nAdd each line to the figure\nend captioning\nDefine function to plot the caption of the figures (for the individual models case)\nAdd caption to the figure\nAdd each line to the figure\n\n\n
\nCOLORBAR NOTE:
\nThe colorbar chosen for representing biases and trends for the HDD15.5 index ranges from red to blue. In this color scheme, negative bias values (red colors) indicate that the Heating Degree Days displayed by the models are lower than those shown by ERA5. Similarly, negative trend values (red colors) indicate a decrease in Heating Degree Days over time. Instead, blueish colors denote more Heating Degree Days, either indicating a positive bias compared to ERA5 or an increase in Heating Degree Days over time. This selection is based on the rationale that more Heating Degree Days are associated with colder conditions, typically represented by blueish colors, while fewer Heating Degree Days imply warmer conditions, depicted by reddish colors.\n\n(section-3.2)=", "text_with_prefix": "EQC Quality Assessment: \"Biases in energy-consumption-related indices in Europe\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q01 | Category: Climate_Projections\nSection: Biases in energy-consumption-related indices in Europe > Analysis and results > 3. Plot and describe results > 3.1. Define plotting functions\n---\nThe functions presented here are used to plot the mean values and trends calculated over the historical period (1971-2000) for each of the indices ('HDD15.5' and 'CDD22').\n\nFor a selected index, three layout types can be displayed, depending on the chosen function:\n\n1. Layout including the reference ERA5 product, the ensemble median, the bias of the ensemble median, and the ensemble spread: `plot_ensemble()` is used.\n2. Layout including every model (for the trend and mean values): `plot_models()` is employed.\n3. Layout including the bias of every model (for the trend and mean values): `plot_models()` is used again.\n\n`trend==True` argument allows displaying trend values over the historical period, while `trend==False` will show mean values. When the `trend` argument is set to `True`, regions with no significance are hatched. For individual models and ERA5, a grid point is considered to have a statistically significant trend when the p-value is lower than 0.05 (in such cases, no hatching is shown). However, for determining trend significance for the ensemble median (understood as the median of the trend values of the chosen subset of models calculated for each grid cell), reliance is placed on agreement categories, following the advanced approach proposed in AR6 [IPCC](https://www.ipcc.ch/report/ar6/wg1/downloads/report/IPCC_AR6_WGI_Atlas.pdf) on pages 1945-1950. The `hatch_p_value_ensemble()` function is used to distinguish, for each grid point, between three possible cases:\n\n1. If more than 66% of the models are statistically significant (p-value < 0.05) and more than 80% of the models share the same sign, we consider the ensemble median trend to be statistically significant, and there is agreement on the sign. To represent this, no hatching is used.\n2. If less than 66% of the models are statistically significant, regardless of agreement on the sign of the trend, hatching is applied (indicating that the ensemble median trend is not statistically significant).\n3. If more than 66% of the models are statistically significant but less than 80% of the models share the same sign, we consider the ensemble median trend to be statistically significant, but there is no agreement on the sign of the trend. This is represented using crosses.\n\nDefine function to plot the caption of the figures (for the ensemble case)\nAdd caption to the figure\nAdd each line to the figure\nend captioning\nDefine function to plot the caption of the figures (for the individual models case)\nAdd caption to the figure\nAdd each line to the figure\n\n\n
\nCOLORBAR NOTE:
\nThe colorbar chosen for representing biases and trends for the HDD15.5 index ranges from red to blue. In this color scheme, negative bias values (red colors) indicate that the Heating Degree Days displayed by the models are lower than those shown by ERA5. Similarly, negative trend values (red colors) indicate a decrease in Heating Degree Days over time. Instead, blueish colors denote more Heating Degree Days, either indicating a positive bias compared to ERA5 or an increase in Heating Degree Days over time. This selection is based on the rationale that more Heating Degree Days are associated with colder conditions, typically represented by blueish colors, while fewer Heating Degree Days imply warmer conditions, depicted by reddish colors.\n\n(section-3.2)="} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q01__e804cc64ea39", "report_id": "climate_projections-cmip6_climate-impact-indicators_q01", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q01", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Biases in energy-consumption-related indices in Europe > Analysis and results > 3. Plot and describe results > 3.2. Plot ensemble maps", "title": "Biases in energy-consumption-related indices in Europe", "chunk_index": 14, "token_count": 361, "text_raw": "In this section, we invoke the `plot_ensemble()` function to visualise the mean values and trends calculated over the historical period (1971-2000) for the model ensemble and ERA5 reference product across Europe. Note that the model data used in this section has previously been interpolated to the ERA5 grid.\n\nSpecifically, for each of the indices ('HDD15.5' and 'CDD22'), this section presents two layouts:\n\n1. Mean values of the historical period (1971-2000) for: (a) the reference ERA5 product, (b) the ensemble median (understood as the median of the mean values of the chosen subset of models calculated for each grid cell), (c) the bias of the ensemble median, and (d) the ensemble spread (derived as the standard deviation of the distribution of the chosen subset of models).\n\n2. Trend values of the historical period (1971-2000) for: (a) the reference ERA5 product, (b) the ensemble median (understood as the median of the trend values of the chosen subset of models calculated for each grid cell), (c) the bias of the ensemble median, and (d) the ensemble spread (derived as the standard deviation of the distribution of the chosen subset of models).\n\nFig number counter\nCommon title\n\n(section-3.3)=", "text_with_prefix": "EQC Quality Assessment: \"Biases in energy-consumption-related indices in Europe\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q01 | Category: Climate_Projections\nSection: Biases in energy-consumption-related indices in Europe > Analysis and results > 3. Plot and describe results > 3.2. Plot ensemble maps\n---\nIn this section, we invoke the `plot_ensemble()` function to visualise the mean values and trends calculated over the historical period (1971-2000) for the model ensemble and ERA5 reference product across Europe. Note that the model data used in this section has previously been interpolated to the ERA5 grid.\n\nSpecifically, for each of the indices ('HDD15.5' and 'CDD22'), this section presents two layouts:\n\n1. Mean values of the historical period (1971-2000) for: (a) the reference ERA5 product, (b) the ensemble median (understood as the median of the mean values of the chosen subset of models calculated for each grid cell), (c) the bias of the ensemble median, and (d) the ensemble spread (derived as the standard deviation of the distribution of the chosen subset of models).\n\n2. Trend values of the historical period (1971-2000) for: (a) the reference ERA5 product, (b) the ensemble median (understood as the median of the trend values of the chosen subset of models calculated for each grid cell), (c) the bias of the ensemble median, and (d) the ensemble spread (derived as the standard deviation of the distribution of the chosen subset of models).\n\nFig number counter\nCommon title\n\n(section-3.3)="} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q01__10971f6c447e", "report_id": "climate_projections-cmip6_climate-impact-indicators_q01", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q01", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Biases in energy-consumption-related indices in Europe > Analysis and results > 3. Plot and describe results > 3.3. Plot model maps", "title": "Biases in energy-consumption-related indices in Europe", "chunk_index": 15, "token_count": 207, "text_raw": "In this section, we invoke the `plot_models()` function to visualise the mean values and trends calculated over the historical period (1971-2000) for every model individually across Europe. Note that the model data used in this section maintains its original grid.\n\nSpecifically, for each of the indices ('HDD15.5' and 'CDD22'), this section presents two layouts:\n\n1. A layout including the historical mean (1971-2000) of every model.\n\n2. A layout including the historical trend (1971-2000) of every model.\n\n(section-3.4)=", "text_with_prefix": "EQC Quality Assessment: \"Biases in energy-consumption-related indices in Europe\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q01 | Category: Climate_Projections\nSection: Biases in energy-consumption-related indices in Europe > Analysis and results > 3. Plot and describe results > 3.3. Plot model maps\n---\nIn this section, we invoke the `plot_models()` function to visualise the mean values and trends calculated over the historical period (1971-2000) for every model individually across Europe. Note that the model data used in this section maintains its original grid.\n\nSpecifically, for each of the indices ('HDD15.5' and 'CDD22'), this section presents two layouts:\n\n1. A layout including the historical mean (1971-2000) of every model.\n\n2. A layout including the historical trend (1971-2000) of every model.\n\n(section-3.4)="} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q01__1387a0ebe82c", "report_id": "climate_projections-cmip6_climate-impact-indicators_q01", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q01", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Biases in energy-consumption-related indices in Europe > Analysis and results > 3. Plot and describe results > 3.4. Plot bias maps", "title": "Biases in energy-consumption-related indices in Europe", "chunk_index": 16, "token_count": 221, "text_raw": "In this section, we invoke the `plot_models()` function to visualise the bias for the mean values and trends calculated over the historical period (1971-2000) for every model individually across Europe. Note that the model data used in this section has previously been interpolated to the ERA5 grid.\n\nSpecifically, for each of the indices ('HDD15.5' and 'CDD22'), this section presents two layouts:\n\n1. A layout including the bias for the historical mean (1971-2000) of every model.\n\n2. A layout including the bias for the historical trend (1971-2000) of every model.\n\n(section-3.5)=", "text_with_prefix": "EQC Quality Assessment: \"Biases in energy-consumption-related indices in Europe\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q01 | Category: Climate_Projections\nSection: Biases in energy-consumption-related indices in Europe > Analysis and results > 3. Plot and describe results > 3.4. Plot bias maps\n---\nIn this section, we invoke the `plot_models()` function to visualise the bias for the mean values and trends calculated over the historical period (1971-2000) for every model individually across Europe. Note that the model data used in this section has previously been interpolated to the ERA5 grid.\n\nSpecifically, for each of the indices ('HDD15.5' and 'CDD22'), this section presents two layouts:\n\n1. A layout including the bias for the historical mean (1971-2000) of every model.\n\n2. A layout including the bias for the historical trend (1971-2000) of every model.\n\n(section-3.5)="} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q01__7f01695d6420", "report_id": "climate_projections-cmip6_climate-impact-indicators_q01", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q01", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Biases in energy-consumption-related indices in Europe > Analysis and results > 3. Plot and describe results > 3.5. Boxplots of the historical trend", "title": "Biases in energy-consumption-related indices in Europe", "chunk_index": 17, "token_count": 398, "text_raw": "In this last section, we compare the trends of the climate models with the reference trend from ERA5.\n\nDots represent the spatially-averaged historical trend over the selected region (change of the number of days per decade) for each model (grey), the ensemble mean (blue), and the reference product (orange). The ensemble median is shown as a green line. Note that the spatially averaged values are calculated for each model from its original grid (i.e., no interpolated data has been used here).\n\nThe boxplot visually illustrates the distribution of trends among the climate models, with the box covering the first quartile (Q1 = 25th percentile) to the third quartile (Q3 = 75th percentile), and a green line indicating the ensemble median (Q2 = 50th percentile). Whiskers extend from the edges of the box to show the full data range.\n\nEnsemble mean\n\n
\n
\n

Fig 13. Boxplots illustrating the historical trends of the distribution of the chosen subset of models and ERA5 for the Energy Degree Days indices daily averaged: (a) 'CDD22', and (b) 'HDD15.5'. The distribution is created by considering spatially averaged trends across Europe. The ensemble mean and the ensemble median trends are both included. Outliers in the distribution are denoted by a grey circle with a black contour.

\n

\n\n(section-3.6)=", "text_with_prefix": "EQC Quality Assessment: \"Biases in energy-consumption-related indices in Europe\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q01 | Category: Climate_Projections\nSection: Biases in energy-consumption-related indices in Europe > Analysis and results > 3. Plot and describe results > 3.5. Boxplots of the historical trend\n---\nIn this last section, we compare the trends of the climate models with the reference trend from ERA5.\n\nDots represent the spatially-averaged historical trend over the selected region (change of the number of days per decade) for each model (grey), the ensemble mean (blue), and the reference product (orange). The ensemble median is shown as a green line. Note that the spatially averaged values are calculated for each model from its original grid (i.e., no interpolated data has been used here).\n\nThe boxplot visually illustrates the distribution of trends among the climate models, with the box covering the first quartile (Q1 = 25th percentile) to the third quartile (Q3 = 75th percentile), and a green line indicating the ensemble median (Q2 = 50th percentile). Whiskers extend from the edges of the box to show the full data range.\n\nEnsemble mean\n\n
\n
\n

Fig 13. Boxplots illustrating the historical trends of the distribution of the chosen subset of models and ERA5 for the Energy Degree Days indices daily averaged: (a) 'CDD22', and (b) 'HDD15.5'. The distribution is created by considering spatially averaged trends across Europe. The ensemble mean and the ensemble median trends are both included. Outliers in the distribution are denoted by a grey circle with a black contour.

\n

\n\n(section-3.6)="} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q01__5c57cc271bf5", "report_id": "climate_projections-cmip6_climate-impact-indicators_q01", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q01", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Biases in energy-consumption-related indices in Europe > Analysis and results > 3. Plot and describe results > 3.6. Results summary and discussion", "title": "Biases in energy-consumption-related indices in Europe", "chunk_index": 18, "token_count": 820, "text_raw": "- GCMs struggle to accurately represent the climatology of these indices in regions with complex orography, specially for the mean values of the HDD15.5 index. There is a common overestimation of HDD15.5 mean values for DJF, especially in continental and northern areas. Conversely, underestimation is evident in mountainous regions and in some western zones. For the climatological values of CDD22, there is overestimation for JJA across most of the Mediterranean Basin (except some Saharan regions).\n\n- The ERA5 historical data shows a reduction in Heating Degree Days (HDD) for DJF across most of Europe. However, while the ensemble median of GCMs (calculated for each grid cell) captures the trend behavior for most regions, it does not accurately reflect their magnitude. Indeed, a significant inter-model spread is evident, with differences in trend direction among individual models observed in some regions. This variability is particularly pronounced in the northern and northeastern parts of Europe.\n\n- The ERA5 historical trend shows an increase in Cooling Degree Days during the summer (JJA) across Europe, especially in the Mediterranean Basin. GCMs agree on the direction of the CDD2 trend for summer and replicate the historical trends (1971-2000) of the selected index, although they underestimate its magnitude.\n\n- Boxplots reveal different signs for the spatial-averaged historical trend characterising the distribution of the chosen subset of models for the HDD15.5 and DJF. The CMIP6 ensemble median trend displays a daily average value around -0.15 °C/10 years, similar to ERA5's -0.15 °C/10 years. The interquantile range of the ensemble ranges from below -0.2 to near -0.05 °C per decade. When assessing the biases of the trend, it can be seen that the interquantile range of the ensemble for the trend bias goes approximately from -0.1 to 0.1 °C per decade.\n\n- Boxplots exhibit a consistent positive spatial-averaged historical trend characterising the different models for the CDD22 for JJA. Magnitudes, however, are slightly underestimated when compared to ERA5. The CMIP6 ensemble median displays near 0.08°C/10 years, contrasting with ERA5's ~0.11 °C/10 years. The interquantile range of the ensemble ranges from 0.06 to 0.11 °C per decade. The interquantile range of the ensemble for the trend bias goes approximately from -0.05 to slightly above 0°C per decade.\n\n- What do the results mean for users? Are the biases relevant?\n - Despite the biases and high inter-model spread in some regions, the outcomes of this notebook may offer valuable insights for decisions sensitive to future energy demand. These results, particularly for CDD22 calculated for summer, may increase confidence in using these models to analyse future trends, although their accuracy is not guaranteed. To improve the accuracy of these insights, biases should be considered and corrected [[6]](https://doi.org/10.1002/joc.5362)[[7]](https://doi.org/10.1038/s41467-021-25504-8).\n\n\n
\nRESULTS NOTE:
\nIt is important to note that the results presented are specific to the 16 models chosen, and users should aim to assess as wide a range of models as possible before making a sub-selection.", "text_with_prefix": "EQC Quality Assessment: \"Biases in energy-consumption-related indices in Europe\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q01 | Category: Climate_Projections\nSection: Biases in energy-consumption-related indices in Europe > Analysis and results > 3. Plot and describe results > 3.6. Results summary and discussion\n---\n- GCMs struggle to accurately represent the climatology of these indices in regions with complex orography, specially for the mean values of the HDD15.5 index. There is a common overestimation of HDD15.5 mean values for DJF, especially in continental and northern areas. Conversely, underestimation is evident in mountainous regions and in some western zones. For the climatological values of CDD22, there is overestimation for JJA across most of the Mediterranean Basin (except some Saharan regions).\n\n- The ERA5 historical data shows a reduction in Heating Degree Days (HDD) for DJF across most of Europe. However, while the ensemble median of GCMs (calculated for each grid cell) captures the trend behavior for most regions, it does not accurately reflect their magnitude. Indeed, a significant inter-model spread is evident, with differences in trend direction among individual models observed in some regions. This variability is particularly pronounced in the northern and northeastern parts of Europe.\n\n- The ERA5 historical trend shows an increase in Cooling Degree Days during the summer (JJA) across Europe, especially in the Mediterranean Basin. GCMs agree on the direction of the CDD2 trend for summer and replicate the historical trends (1971-2000) of the selected index, although they underestimate its magnitude.\n\n- Boxplots reveal different signs for the spatial-averaged historical trend characterising the distribution of the chosen subset of models for the HDD15.5 and DJF. The CMIP6 ensemble median trend displays a daily average value around -0.15 °C/10 years, similar to ERA5's -0.15 °C/10 years. The interquantile range of the ensemble ranges from below -0.2 to near -0.05 °C per decade. When assessing the biases of the trend, it can be seen that the interquantile range of the ensemble for the trend bias goes approximately from -0.1 to 0.1 °C per decade.\n\n- Boxplots exhibit a consistent positive spatial-averaged historical trend characterising the different models for the CDD22 for JJA. Magnitudes, however, are slightly underestimated when compared to ERA5. The CMIP6 ensemble median displays near 0.08°C/10 years, contrasting with ERA5's ~0.11 °C/10 years. The interquantile range of the ensemble ranges from 0.06 to 0.11 °C per decade. The interquantile range of the ensemble for the trend bias goes approximately from -0.05 to slightly above 0°C per decade.\n\n- What do the results mean for users? Are the biases relevant?\n - Despite the biases and high inter-model spread in some regions, the outcomes of this notebook may offer valuable insights for decisions sensitive to future energy demand. These results, particularly for CDD22 calculated for summer, may increase confidence in using these models to analyse future trends, although their accuracy is not guaranteed. To improve the accuracy of these insights, biases should be considered and corrected [[6]](https://doi.org/10.1002/joc.5362)[[7]](https://doi.org/10.1038/s41467-021-25504-8).\n\n\n
\nRESULTS NOTE:
\nIt is important to note that the results presented are specific to the 16 models chosen, and users should aim to assess as wide a range of models as possible before making a sub-selection."} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q01__2095dd406429", "report_id": "climate_projections-cmip6_climate-impact-indicators_q01", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q01", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Biases in energy-consumption-related indices in Europe > ℹ️ If you want to know more > Key resources", "title": "Biases in energy-consumption-related indices in Europe", "chunk_index": 19, "token_count": 265, "text_raw": "Some key resources and further reading were linked throughout this assessment.\n\nThe CDS catalogue entries for the data used were:\n* CMIP6 climate projections (Daily - air temperature): https://cds.climate.copernicus.eu/datasets/projections-cmip6?tab=overview\n* ERA5 hourly data on single levels from 1940 to present (2m temperature): https://cds.climate.copernicus.eu/datasets/reanalysis-era5-single-levels?tab=overview\n\nCode libraries used:\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)\n* [icclim](https://icclim.readthedocs.io/en/stable/) Python package", "text_with_prefix": "EQC Quality Assessment: \"Biases in energy-consumption-related indices in Europe\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q01 | Category: Climate_Projections\nSection: Biases in energy-consumption-related indices in Europe > ℹ️ If you want to know more > Key resources\n---\nSome key resources and further reading were linked throughout this assessment.\n\nThe CDS catalogue entries for the data used were:\n* CMIP6 climate projections (Daily - air temperature): https://cds.climate.copernicus.eu/datasets/projections-cmip6?tab=overview\n* ERA5 hourly data on single levels from 1940 to present (2m temperature): https://cds.climate.copernicus.eu/datasets/reanalysis-era5-single-levels?tab=overview\n\nCode libraries used:\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)\n* [icclim](https://icclim.readthedocs.io/en/stable/) Python package"} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q01__83933d9c2112", "report_id": "climate_projections-cmip6_climate-impact-indicators_q01", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q01", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Biases in energy-consumption-related indices in Europe > ℹ️ If you want to know more > References", "title": "Biases in energy-consumption-related indices in Europe", "chunk_index": 20, "token_count": 886, "text_raw": "[[1]](https://doi.org/10.1007/s10113-010-0173-x) Bindi, M., Olesen, J.E. The responses of agriculture in Europe to climate change (2011). Reg Environ Change 11 (Suppl 1), 151–158. https://doi.org/10.1007/s10113-010-0173-x\n\n[[2]](https://doi.org/10.1016/j.foreco.2009.09.023) Lindner, M., Maroschek, M., Netherer, S., Kremer, A., Barbati, A., Garcia-Gonzalo, J., Seidi, R., Delzon, S., Corona, P., Kolstrom, M., Lexer, M.J., Marchetti, M. (2010). Climate change impacts, adaptive capacity, and vulnerability of European forest ecosystems. For. Ecol. Manage. 259(4): 698–709. https://doi.org/10.1016/j.foreco.2009.09.023\n\n[[3]](https://doi.org/10.1016/j.enbuild.2014.09.052) Santamouris, M., Cartalis, C., Synnefa, A., Kolokotsa, D. (2015). On the impact of urban heat island and global warming on the power demand and electricity consumption of buildings – a review. Energy Build. 98: 119–124. https://doi.org/10.1016/j.enbuild.2014.09.052\n\n[[4]](https://doi.org/10.1007/s10113-013-0499-2) Jacob, D., Petersen, J., Eggert, B. et al. (2014). EURO-CORDEX: new high-resolution climate change projections for European impact research. Reg Environ Change 14, 563–578. https://doi.org/10.1007/s10113-013-0499-2\n\n[[5]](http://hdl.handle.net/10013/epic.45156.d001) IPCC. 2014. In Climate Change 2014: Synthesis Report. Contribution of Working Groups I, II and III to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change, Core Writing Team, RK Pachauri, LA Meyer (eds). IPCC: Geneva, Switzerland 151 pp.\n\n[[6]](https://doi.org/10.1002/joc.5362) Spinoni, J., Vogt, J.V., Barbosa, P., Dosio, A., McCormick, N., Bigano, A. and Füssel, H.-M. (2018). Changes of heating and cooling degree-days in Europe from 1981 to 2100. Int. J. Climatol, 38: e191-e208. https://doi.org/10.1002/joc.5362\n\n[[7]](https://doi.org/10.1038/s41467-021-25504-8) Deroubaix, A., Labuhn, I., Camredon, M. et al. (2021). Large uncertainties in trends of energy demand for heating and cooling under climate change. Nat Commun 12, 5197. https://doi.org/10.1038/s41467-021-25504-8\n\n[[8]](https://doi.org/10.1038/s43247-023-00878-3) Scoccimarro, E., Cattaneo, O., Gualdi, S. et al. (2023). Country-level energy demand for cooling has increased over the past two decades. Commun Earth Environ 4, 208. https://doi.org/10.1038/s43247-023-00878-3", "text_with_prefix": "EQC Quality Assessment: \"Biases in energy-consumption-related indices in Europe\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q01 | Category: Climate_Projections\nSection: Biases in energy-consumption-related indices in Europe > ℹ️ If you want to know more > References\n---\n[[1]](https://doi.org/10.1007/s10113-010-0173-x) Bindi, M., Olesen, J.E. The responses of agriculture in Europe to climate change (2011). Reg Environ Change 11 (Suppl 1), 151–158. https://doi.org/10.1007/s10113-010-0173-x\n\n[[2]](https://doi.org/10.1016/j.foreco.2009.09.023) Lindner, M., Maroschek, M., Netherer, S., Kremer, A., Barbati, A., Garcia-Gonzalo, J., Seidi, R., Delzon, S., Corona, P., Kolstrom, M., Lexer, M.J., Marchetti, M. (2010). Climate change impacts, adaptive capacity, and vulnerability of European forest ecosystems. For. Ecol. Manage. 259(4): 698–709. https://doi.org/10.1016/j.foreco.2009.09.023\n\n[[3]](https://doi.org/10.1016/j.enbuild.2014.09.052) Santamouris, M., Cartalis, C., Synnefa, A., Kolokotsa, D. (2015). On the impact of urban heat island and global warming on the power demand and electricity consumption of buildings – a review. Energy Build. 98: 119–124. https://doi.org/10.1016/j.enbuild.2014.09.052\n\n[[4]](https://doi.org/10.1007/s10113-013-0499-2) Jacob, D., Petersen, J., Eggert, B. et al. (2014). EURO-CORDEX: new high-resolution climate change projections for European impact research. Reg Environ Change 14, 563–578. https://doi.org/10.1007/s10113-013-0499-2\n\n[[5]](http://hdl.handle.net/10013/epic.45156.d001) IPCC. 2014. In Climate Change 2014: Synthesis Report. Contribution of Working Groups I, II and III to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change, Core Writing Team, RK Pachauri, LA Meyer (eds). IPCC: Geneva, Switzerland 151 pp.\n\n[[6]](https://doi.org/10.1002/joc.5362) Spinoni, J., Vogt, J.V., Barbosa, P., Dosio, A., McCormick, N., Bigano, A. and Füssel, H.-M. (2018). Changes of heating and cooling degree-days in Europe from 1981 to 2100. Int. J. Climatol, 38: e191-e208. https://doi.org/10.1002/joc.5362\n\n[[7]](https://doi.org/10.1038/s41467-021-25504-8) Deroubaix, A., Labuhn, I., Camredon, M. et al. (2021). Large uncertainties in trends of energy demand for heating and cooling under climate change. Nat Commun 12, 5197. https://doi.org/10.1038/s41467-021-25504-8\n\n[[8]](https://doi.org/10.1038/s43247-023-00878-3) Scoccimarro, E., Cattaneo, O., Gualdi, S. et al. (2023). Country-level energy demand for cooling has increased over the past two decades. Commun Earth Environ 4, 208. https://doi.org/10.1038/s43247-023-00878-3"} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q02__d3a5fe8864e9", "report_id": "climate_projections-cmip6_climate-impact-indicators_q02", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q02", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes of energy consumption in Europe", "title": "Uncertainty in projected changes of energy consumption in Europe", "chunk_index": 0, "token_count": 94, "text_raw": "Production date: 23-05-2024\n\nProduced by: CMCC foundation - Euro-Mediterranean Center on Climate Change. Albert Martinez Boti.", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes of energy consumption in Europe\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q02 | Category: Climate_Projections\nSection: Uncertainty in projected changes of energy consumption in Europe\n---\nProduction date: 23-05-2024\n\nProduced by: CMCC foundation - Euro-Mediterranean Center on Climate Change. Albert Martinez Boti."} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q02__4c98ea14a4f1", "report_id": "climate_projections-cmip6_climate-impact-indicators_q02", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q02", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes of energy consumption in Europe > Quality assessment question", "title": "Uncertainty in projected changes of energy consumption in Europe", "chunk_index": 1, "token_count": 635, "text_raw": "* **What are the projected future changes and associated uncertainties of Energy Degree Days in Europe?**\n\nSectors affected by climate change are varied including agriculture [[1]](https://doi.org/10.1007/s10113-010-0173-x), forest ecosystems [[2]](https://doi.org/10.1016/j.foreco.2009.09.023), and energy consumption [[3]](https://doi.org/10.1016/j.enbuild.2014.09.052). Under projected future global warming over Europe [[4]](https://doi.org/10.1007/s10113-013-0499-2)[[5]](http://hdl.handle.net/10013/epic.45156.d001), the current increase in energy demand is expected to persist until the end of this century and beyond [[6]](https://doi.org/10.1002/joc.5362). Identifying which climate-change-related impacts are likely to increase, by how much, and inherent regional patterns, is important for any effective strategy for managing future climate risks. This notebook utilises data from a subset of models from [CMIP6](https://cds.climate.copernicus.eu/datasets/projections-cmip6?tab=overview) Global Climate Models (GCMs) and explores the uncertainty in future projections of energy-consumption-related indices by considering the ensemble inter-model spread of projected changes. Two energy-consumption-related indices are calculated from daily mean temperatures using the [icclim](https://icclim.readthedocs.io/en/stable/) Python package: Cooling Degree Days (CDDs) and Heating Degree Days (HDDs). Degree days measure how much warmer or colder it is compared to standard temperatures (usually 15.5°C for heating and 22°C for cooling). Higher degree day numbers indicate more extreme temperatures, which typically lead to increased energy use for heating or cooling buildings. In the presented code, CDD calculations use summer aggregation (CDD22), while HDD calculations focus on winter (HDD15.5), presenting results as daily averages rather than cumulative values. Within this notebook, these calculations are performed over the future period from 2016 to 2099, following the Shared Socioeconomic Pathways SSP5-8.5. It is important to note that the results presented here pertain to a specific subset of the CMIP6 ensemble and may not be generalisable to the entire dataset. Also note that a separate assessment examines the representation of climatology and trends of these indices for the same models during the historical period (1971-2000), while another assessment looks at the projected climate signal of these indices for the same models at a 2°C Global Warming Level.", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes of energy consumption in Europe\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q02 | Category: Climate_Projections\nSection: Uncertainty in projected changes of energy consumption in Europe > Quality assessment question\n---\n* **What are the projected future changes and associated uncertainties of Energy Degree Days in Europe?**\n\nSectors affected by climate change are varied including agriculture [[1]](https://doi.org/10.1007/s10113-010-0173-x), forest ecosystems [[2]](https://doi.org/10.1016/j.foreco.2009.09.023), and energy consumption [[3]](https://doi.org/10.1016/j.enbuild.2014.09.052). Under projected future global warming over Europe [[4]](https://doi.org/10.1007/s10113-013-0499-2)[[5]](http://hdl.handle.net/10013/epic.45156.d001), the current increase in energy demand is expected to persist until the end of this century and beyond [[6]](https://doi.org/10.1002/joc.5362). Identifying which climate-change-related impacts are likely to increase, by how much, and inherent regional patterns, is important for any effective strategy for managing future climate risks. This notebook utilises data from a subset of models from [CMIP6](https://cds.climate.copernicus.eu/datasets/projections-cmip6?tab=overview) Global Climate Models (GCMs) and explores the uncertainty in future projections of energy-consumption-related indices by considering the ensemble inter-model spread of projected changes. Two energy-consumption-related indices are calculated from daily mean temperatures using the [icclim](https://icclim.readthedocs.io/en/stable/) Python package: Cooling Degree Days (CDDs) and Heating Degree Days (HDDs). Degree days measure how much warmer or colder it is compared to standard temperatures (usually 15.5°C for heating and 22°C for cooling). Higher degree day numbers indicate more extreme temperatures, which typically lead to increased energy use for heating or cooling buildings. In the presented code, CDD calculations use summer aggregation (CDD22), while HDD calculations focus on winter (HDD15.5), presenting results as daily averages rather than cumulative values. Within this notebook, these calculations are performed over the future period from 2016 to 2099, following the Shared Socioeconomic Pathways SSP5-8.5. It is important to note that the results presented here pertain to a specific subset of the CMIP6 ensemble and may not be generalisable to the entire dataset. Also note that a separate assessment examines the representation of climatology and trends of these indices for the same models during the historical period (1971-2000), while another assessment looks at the projected climate signal of these indices for the same models at a 2°C Global Warming Level."} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q02__7ca3a397150b", "report_id": "climate_projections-cmip6_climate-impact-indicators_q02", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q02", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes of energy consumption in Europe > Quality assessment statement", "title": "Uncertainty in projected changes of energy consumption in Europe", "chunk_index": 2, "token_count": 440, "text_raw": "These are the key outcomes of this assessment\n\n* The subset of considered CMIP6 models agree on a general decrease in future trends (2016-2099) for HDD15.5 across Europe during DJF and an increase in CDD22 during JJA. However, regional variations exist.\n\n* The northern and eastern parts of Europe are projected to experience the largest decrease in HDD15.5, accompanied by higher inter-model variability. Some areas show no trend for CDD2 due to threshold temperatures not being reached. The Mediterranean Basin will see the greatest increase in CDD22, with higher inter-model variability.\n\n* The findings of this notebook could support decisions sensitive to future energy demand. Despite regional variations and some inter-model spread (calculated to account for projected uncertainty), the subset of 16 models from CMIP6 agree on a significant decrease in the energy required for heating spaces during winter. This decrease is particularly notable in regions with high HDD (northern and eastern regions that experience substantial heating energy consumption in winter). Conversely, more energy will be needed in the future to cool buildings during summer, especially in the Mediterranean Basin.\n```\n\nattachment:7245f9d8-4fce-41a6-9b6b-a4a3519f59e0.png \n---\nalt: trend_future_CDD\nwidth: 850px\n---\nCooling Degree Days daily average calculated using the summer comfort threshold of 22°C ('CDD22) for the temporal aggregation of 'JA'. Trend for the future period (2016-2099). The layout includes data corresponding to: (a) the ensemble median (understood as the median of the trend values of the chosen subset of models calculated for each grid cell) and (b) the ensemble spread (derived as the standard deviation of the distribution of the chosen subset of models).\n```", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes of energy consumption in Europe\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q02 | Category: Climate_Projections\nSection: Uncertainty in projected changes of energy consumption in Europe > Quality assessment statement\n---\nThese are the key outcomes of this assessment\n\n* The subset of considered CMIP6 models agree on a general decrease in future trends (2016-2099) for HDD15.5 across Europe during DJF and an increase in CDD22 during JJA. However, regional variations exist.\n\n* The northern and eastern parts of Europe are projected to experience the largest decrease in HDD15.5, accompanied by higher inter-model variability. Some areas show no trend for CDD2 due to threshold temperatures not being reached. The Mediterranean Basin will see the greatest increase in CDD22, with higher inter-model variability.\n\n* The findings of this notebook could support decisions sensitive to future energy demand. Despite regional variations and some inter-model spread (calculated to account for projected uncertainty), the subset of 16 models from CMIP6 agree on a significant decrease in the energy required for heating spaces during winter. This decrease is particularly notable in regions with high HDD (northern and eastern regions that experience substantial heating energy consumption in winter). Conversely, more energy will be needed in the future to cool buildings during summer, especially in the Mediterranean Basin.\n```\n\nattachment:7245f9d8-4fce-41a6-9b6b-a4a3519f59e0.png \n---\nalt: trend_future_CDD\nwidth: 850px\n---\nCooling Degree Days daily average calculated using the summer comfort threshold of 22°C ('CDD22) for the temporal aggregation of 'JA'. Trend for the future period (2016-2099). The layout includes data corresponding to: (a) the ensemble median (understood as the median of the trend values of the chosen subset of models calculated for each grid cell) and (b) the ensemble spread (derived as the standard deviation of the distribution of the chosen subset of models).\n```"} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q02__d6374b30604f", "report_id": "climate_projections-cmip6_climate-impact-indicators_q02", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q02", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes of energy consumption in Europe > Methodology", "title": "Uncertainty in projected changes of energy consumption in Europe", "chunk_index": 3, "token_count": 972, "text_raw": "The reference methodology used here for the indices calculation is similar to the one followed by Scoccimarro et al., (2023) [[7]](https://doi.org/10.1038/s43247-023-00878-3). However the thermal comfort thresholds used in this notebook are slightly different. A winter comfort temperature of 15.5°C and a summer comfort temperature of 22.0°C are used here (as in the [CDS application](https://dast.copernicus-climate.eu/documents/app-heating-cooling-degree-days/C3S_EEA_HDD_CDD_application_user_guide_v0.9.pdf)). In the presented code, the CDD calculations are based on the JJA aggregation, with a comfort temperature of 22°C (CDD22), while HDD calculations focus on winter (DJF) with a comfort temperature of 15.5°C (HDD15.5). More specifically, to calculate CDD22, the sum of the differences between the daily mean temperature and the thermal comfort temperature of 22°C is computed. This calculation occurs only when the mean temperature is above the thermal comfort level; otherwise, the CDD22 for that day is set to 0. For example, a day with a mean temperature of 28°C would result in 6°C. Two consecutive hot days like this would total 12°C over the two-day period. Similarly, to calculate HDD15.5, the sum of the differences between the thermal comfort temperature of 15.5°C and the daily mean temperature is determined. This happens only when the mean temperature is below the thermal comfort level; otherwise, the HDD15.5 for that day is set to 0. Finally, to obtain more intuitive values, the sum is averaged over the number of days in the season to produce daily average values. This approach differs from the [CDS application](https://dast.copernicus-climate.eu/documents/app-heating-cooling-degree-days/C3S_EEA_HDD_CDD_application_user_guide_v0.9.pdf), where both the sum over a period and the daily average values can be displayed. In Spinoni et al., (2018) [[6]](https://doi.org/10.1002/joc.5362), as well as in the [CDS application](https://dast.copernicus-climate.eu/documents/app-heating-cooling-degree-days/C3S_EEA_HDD_CDD_application_user_guide_v0.9.pdf), more advanced methods for calculating CDD and HDD involve considering maximum, minimum, and mean temperatures. However, to prevent overloading the notebook and maintain simplicity while ensuring compatibility with the [icclim](https://icclim.readthedocs.io/en/stable/) Python package, we opted to utilise a single variable (2m mean temperature).\n\nThis notebook offers an assessment of the projected changes and their associated uncertainties using a subset of 16 models from [CMIP6](https://cds.climate.copernicus.eu/datasets/projections-cmip6?tab=overview). The uncertainty is examined by analysing the ensemble inter-model spread of projected changes for the energy-consumption-related indices 'HDD15.5' and 'CDD22', calculated for the future period spanning from 2016 to 2099. In particular, spatial patterns of climate projected trends are examined and displayed for each model individually and for the ensemble median (calculated for each grid cell), alongside the ensemble inter-model spread to account for projected uncertainty. Additionally, spatially-averaged trend values are analysed and presented using box plots to provide an overview of trend behavior across the distribution of the chosen subset of models when averaged across Europe.\n\nThe analysis and results follow the next outline:\n\n**[](section-1)**\n * [](section-1.1)\n * [](section-1.2)\n * [](section-1.3)\n * [](section-1.4)\n * [](section-1.5)\n * [](section-1.6)\n\n**[](section-2)**\n * [](section-2.1)\n * [](section-2.2)\n * [](section-2.3)\n\n**[](section-3)**\n * [](section-3.1)\n * [](section-3.2)\n * [](section-3.3)\n * [](section-3.4)\n * [](section-3.5)", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes of energy consumption in Europe\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q02 | Category: Climate_Projections\nSection: Uncertainty in projected changes of energy consumption in Europe > Methodology\n---\nThe reference methodology used here for the indices calculation is similar to the one followed by Scoccimarro et al., (2023) [[7]](https://doi.org/10.1038/s43247-023-00878-3). However the thermal comfort thresholds used in this notebook are slightly different. A winter comfort temperature of 15.5°C and a summer comfort temperature of 22.0°C are used here (as in the [CDS application](https://dast.copernicus-climate.eu/documents/app-heating-cooling-degree-days/C3S_EEA_HDD_CDD_application_user_guide_v0.9.pdf)). In the presented code, the CDD calculations are based on the JJA aggregation, with a comfort temperature of 22°C (CDD22), while HDD calculations focus on winter (DJF) with a comfort temperature of 15.5°C (HDD15.5). More specifically, to calculate CDD22, the sum of the differences between the daily mean temperature and the thermal comfort temperature of 22°C is computed. This calculation occurs only when the mean temperature is above the thermal comfort level; otherwise, the CDD22 for that day is set to 0. For example, a day with a mean temperature of 28°C would result in 6°C. Two consecutive hot days like this would total 12°C over the two-day period. Similarly, to calculate HDD15.5, the sum of the differences between the thermal comfort temperature of 15.5°C and the daily mean temperature is determined. This happens only when the mean temperature is below the thermal comfort level; otherwise, the HDD15.5 for that day is set to 0. Finally, to obtain more intuitive values, the sum is averaged over the number of days in the season to produce daily average values. This approach differs from the [CDS application](https://dast.copernicus-climate.eu/documents/app-heating-cooling-degree-days/C3S_EEA_HDD_CDD_application_user_guide_v0.9.pdf), where both the sum over a period and the daily average values can be displayed. In Spinoni et al., (2018) [[6]](https://doi.org/10.1002/joc.5362), as well as in the [CDS application](https://dast.copernicus-climate.eu/documents/app-heating-cooling-degree-days/C3S_EEA_HDD_CDD_application_user_guide_v0.9.pdf), more advanced methods for calculating CDD and HDD involve considering maximum, minimum, and mean temperatures. However, to prevent overloading the notebook and maintain simplicity while ensuring compatibility with the [icclim](https://icclim.readthedocs.io/en/stable/) Python package, we opted to utilise a single variable (2m mean temperature).\n\nThis notebook offers an assessment of the projected changes and their associated uncertainties using a subset of 16 models from [CMIP6](https://cds.climate.copernicus.eu/datasets/projections-cmip6?tab=overview). The uncertainty is examined by analysing the ensemble inter-model spread of projected changes for the energy-consumption-related indices 'HDD15.5' and 'CDD22', calculated for the future period spanning from 2016 to 2099. In particular, spatial patterns of climate projected trends are examined and displayed for each model individually and for the ensemble median (calculated for each grid cell), alongside the ensemble inter-model spread to account for projected uncertainty. Additionally, spatially-averaged trend values are analysed and presented using box plots to provide an overview of trend behavior across the distribution of the chosen subset of models when averaged across Europe.\n\nThe analysis and results follow the next outline:\n\n**[](section-1)**\n * [](section-1.1)\n * [](section-1.2)\n * [](section-1.3)\n * [](section-1.4)\n * [](section-1.5)\n * [](section-1.6)\n\n**[](section-2)**\n * [](section-2.1)\n * [](section-2.2)\n * [](section-2.3)\n\n**[](section-3)**\n * [](section-3.1)\n * [](section-3.2)\n * [](section-3.3)\n * [](section-3.4)\n * [](section-3.5)"} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q02__06d9356a4a33", "report_id": "climate_projections-cmip6_climate-impact-indicators_q02", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q02", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes of energy consumption in Europe > Analysis and results > 1. Parameters, requests and functions definition > 1.2. Define Parameters", "title": "Uncertainty in projected changes of energy consumption in Europe", "chunk_index": 4, "token_count": 400, "text_raw": "In the \"Define Parameters\" section, various customisable options for the notebook are specified. Most of the parameters chosen are the same as those used in another assessment (\"CMIP6 Climate Projections: evaluating biases in energy-consumption-related indices in Europe\"), being them:\n\n- The initial and ending year used for the future projections period can be specified by changing the parameters `year_start` and `year_stop` (2016-2099 is chosen).\n- `index_timeseries` is a dictionary that set the temporal aggregation for every index considered within this notebook ('HDD15.5' and 'CDD22'). In the presented code, the CDD calculations are always based on the JJA aggregation, with a comfort temperature of 22°C (CDD22), while HDD calculations focus on winter (DJF) with a comfort temperature of 15.5°C (HDD15.5).\n- `collection_id` set the family of models. Only CMIP6 is implemented for this sub-notebook.\n- `area` allows specifying the geographical domain of interest.\n- The `interpolation_method` parameter allows selecting the interpolation method when regridding is performed over the indices.\n- The `chunk` selection allows the user to define if dividing into chunks when downloading the data on their local machine. Although it does not significantly affect the analysis, it is recommended to keep the default value for optimal performance.\n\nChoose annual or seasonal timeseries\nSelect the family of models\nInterpolation method\nArea to show\nChunks for download\n\n(section-1.3)=", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes of energy consumption in Europe\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q02 | Category: Climate_Projections\nSection: Uncertainty in projected changes of energy consumption in Europe > Analysis and results > 1. Parameters, requests and functions definition > 1.2. Define Parameters\n---\nIn the \"Define Parameters\" section, various customisable options for the notebook are specified. Most of the parameters chosen are the same as those used in another assessment (\"CMIP6 Climate Projections: evaluating biases in energy-consumption-related indices in Europe\"), being them:\n\n- The initial and ending year used for the future projections period can be specified by changing the parameters `year_start` and `year_stop` (2016-2099 is chosen).\n- `index_timeseries` is a dictionary that set the temporal aggregation for every index considered within this notebook ('HDD15.5' and 'CDD22'). In the presented code, the CDD calculations are always based on the JJA aggregation, with a comfort temperature of 22°C (CDD22), while HDD calculations focus on winter (DJF) with a comfort temperature of 15.5°C (HDD15.5).\n- `collection_id` set the family of models. Only CMIP6 is implemented for this sub-notebook.\n- `area` allows specifying the geographical domain of interest.\n- The `interpolation_method` parameter allows selecting the interpolation method when regridding is performed over the indices.\n- The `chunk` selection allows the user to define if dividing into chunks when downloading the data on their local machine. Although it does not significantly affect the analysis, it is recommended to keep the default value for optimal performance.\n\nChoose annual or seasonal timeseries\nSelect the family of models\nInterpolation method\nArea to show\nChunks for download\n\n(section-1.3)="} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q02__2df89fd6cfca", "report_id": "climate_projections-cmip6_climate-impact-indicators_q02", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q02", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes of energy consumption in Europe > Analysis and results > 1. Parameters, requests and functions definition > 1.3. Define models", "title": "Uncertainty in projected changes of energy consumption in Europe", "chunk_index": 5, "token_count": 184, "text_raw": "The following climate analyses are performed considering a subset of GCMs from CMIP6. Models names are listed in the parameters below. Some variable-dependent parameters are also selected.\n\nThe selected CMIP6 models have available both the historical and SSP8.5 experiments, and they are the same as those used in other assessments (\"CMIP6 Climate Projections: evaluating biases in energy-consumption-related indices in Europe\").\n\nDefine models\nDefine dictionaries to use in titles and caption\n\n(section-1.4)=", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes of energy consumption in Europe\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q02 | Category: Climate_Projections\nSection: Uncertainty in projected changes of energy consumption in Europe > Analysis and results > 1. Parameters, requests and functions definition > 1.3. Define models\n---\nThe following climate analyses are performed considering a subset of GCMs from CMIP6. Models names are listed in the parameters below. Some variable-dependent parameters are also selected.\n\nThe selected CMIP6 models have available both the historical and SSP8.5 experiments, and they are the same as those used in other assessments (\"CMIP6 Climate Projections: evaluating biases in energy-consumption-related indices in Europe\").\n\nDefine models\nDefine dictionaries to use in titles and caption\n\n(section-1.4)="} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q02__ee3c58eb54a1", "report_id": "climate_projections-cmip6_climate-impact-indicators_q02", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q02", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes of energy consumption in Europe > Analysis and results > 1. Parameters, requests and functions definition > 1.4. Define land-sea mask request", "title": "Uncertainty in projected changes of energy consumption in Europe", "chunk_index": 6, "token_count": 136, "text_raw": "Within this sub-notebook, ERA5 will be used to download the land-sea mask when plotting. In this section, we set the required parameters for the cds-api data-request of ERA5 land-sea mask.\n\n(section-1.5)=", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes of energy consumption in Europe\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q02 | Category: Climate_Projections\nSection: Uncertainty in projected changes of energy consumption in Europe > Analysis and results > 1. Parameters, requests and functions definition > 1.4. Define land-sea mask request\n---\nWithin this sub-notebook, ERA5 will be used to download the land-sea mask when plotting. In this section, we set the required parameters for the cds-api data-request of ERA5 land-sea mask.\n\n(section-1.5)="} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q02__fbd34ac5bc13", "report_id": "climate_projections-cmip6_climate-impact-indicators_q02", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q02", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes of energy consumption in Europe > Analysis and results > 1. Parameters, requests and functions definition > 1.5. Define model requests", "title": "Uncertainty in projected changes of energy consumption in Europe", "chunk_index": 7, "token_count": 149, "text_raw": "In this section we set the required parameters for the cds-api data-request.\n\nWhen `weights = True`, spatial weighting is applied for calculations requiring spatial data aggregation. This is particularly relevant for CMIP6 GCMs with regular lon-lat grids that do not consider varying surface extensions at different latitudes.\n\n(section-1.6)=", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes of energy consumption in Europe\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q02 | Category: Climate_Projections\nSection: Uncertainty in projected changes of energy consumption in Europe > Analysis and results > 1. Parameters, requests and functions definition > 1.5. Define model requests\n---\nIn this section we set the required parameters for the cds-api data-request.\n\nWhen `weights = True`, spatial weighting is applied for calculations requiring spatial data aggregation. This is particularly relevant for CMIP6 GCMs with regular lon-lat grids that do not consider varying surface extensions at different latitudes.\n\n(section-1.6)="} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q02__955a0d4964db", "report_id": "climate_projections-cmip6_climate-impact-indicators_q02", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q02", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes of energy consumption in Europe > Analysis and results > 1. Parameters, requests and functions definition > 1.6. Functions to cache", "title": "Uncertainty in projected changes of energy consumption in Europe", "chunk_index": 8, "token_count": 290, "text_raw": "In this section, functions that will be executed in the caching phase are defined. Caching is the process of storing copies of files in a temporary storage location, so that they can be accessed more quickly. This process also checks if the user has already downloaded a file, avoiding redundant downloads.\n\nFunctions description:\n\n- The `select_timeseries` function subsets the dataset based on the chosen `timeseries` parameter.\n\n- The `compute_indices` function utilises the icclim package to calculate the energy-consumption-related indices.\n\n- The `compute_trends` function employs the Mann-Kendall test for trend calculation.\n\n- Finally, the `compute_indices_and_trends` function calculates the energy consumption-related indices for the corresponding temporal aggregation using the `compute_indices` function, determines the indices mean for the future period (2016-2099), obtain the trends using the `compute_trends` function, and offers an option for regridding to `model_regrid`.\n\nOriginal bounds for conservative interpolation\n\n(section-2)=", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes of energy consumption in Europe\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q02 | Category: Climate_Projections\nSection: Uncertainty in projected changes of energy consumption in Europe > Analysis and results > 1. Parameters, requests and functions definition > 1.6. Functions to cache\n---\nIn this section, functions that will be executed in the caching phase are defined. Caching is the process of storing copies of files in a temporary storage location, so that they can be accessed more quickly. This process also checks if the user has already downloaded a file, avoiding redundant downloads.\n\nFunctions description:\n\n- The `select_timeseries` function subsets the dataset based on the chosen `timeseries` parameter.\n\n- The `compute_indices` function utilises the icclim package to calculate the energy-consumption-related indices.\n\n- The `compute_trends` function employs the Mann-Kendall test for trend calculation.\n\n- Finally, the `compute_indices_and_trends` function calculates the energy consumption-related indices for the corresponding temporal aggregation using the `compute_indices` function, determines the indices mean for the future period (2016-2099), obtain the trends using the `compute_trends` function, and offers an option for regridding to `model_regrid`.\n\nOriginal bounds for conservative interpolation\n\n(section-2)="} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q02__20e7c2edd90d", "report_id": "climate_projections-cmip6_climate-impact-indicators_q02", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q02", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes of energy consumption in Europe > Analysis and results > 2. Downloading and processing > 2.1. Download and transform the regridding model", "title": "Uncertainty in projected changes of energy consumption in Europe", "chunk_index": 9, "token_count": 293, "text_raw": "In this section, the `download.download_and_transform` function from the 'c3s_eqc_automatic_quality_control' package is employed to download daily data from the selected CMIP6 regridding model, compute the energy-consumption-related indices for the selected temporal aggregation (\"DJF\" for HDD and \"JJA\" for CDD), calculate the mean and trend over the future projections period (2016-2099), and cache the result (to avoid redundant downloads and processing).\n\nThe regridding model is intended here as the model whose grid will be used to interpolate the others. This ensures all models share a common grid, facilitating the calculation of median values for each cell point. The regridding model within this notebook is \"gfdl_esm4\" but a different one can be selected by just modifying the `model_regrid` parameter at [](section-1.3). It is key to highlight the importance of the chosen target grid depending on the specific application.\n\n(section-2.2)=", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes of energy consumption in Europe\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q02 | Category: Climate_Projections\nSection: Uncertainty in projected changes of energy consumption in Europe > Analysis and results > 2. Downloading and processing > 2.1. Download and transform the regridding model\n---\nIn this section, the `download.download_and_transform` function from the 'c3s_eqc_automatic_quality_control' package is employed to download daily data from the selected CMIP6 regridding model, compute the energy-consumption-related indices for the selected temporal aggregation (\"DJF\" for HDD and \"JJA\" for CDD), calculate the mean and trend over the future projections period (2016-2099), and cache the result (to avoid redundant downloads and processing).\n\nThe regridding model is intended here as the model whose grid will be used to interpolate the others. This ensures all models share a common grid, facilitating the calculation of median values for each cell point. The regridding model within this notebook is \"gfdl_esm4\" but a different one can be selected by just modifying the `model_regrid` parameter at [](section-1.3). It is key to highlight the importance of the chosen target grid depending on the specific application.\n\n(section-2.2)="} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q02__fef6b4406a23", "report_id": "climate_projections-cmip6_climate-impact-indicators_q02", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q02", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes of energy consumption in Europe > Analysis and results > 2. Downloading and processing > 2.2. Download and transform models", "title": "Uncertainty in projected changes of energy consumption in Europe", "chunk_index": 10, "token_count": 364, "text_raw": "In this section, the `download.download_and_transform` function from the 'c3s_eqc_automatic_quality_control' package is employed to download daily data from the CMIP6 models, compute the energy-consumption-related indices for the selected temporal aggregation (\"DJF\" for HDD and \"JJA\" for CDD), calculate the mean and trend over the future period (2016-2099), interpolate to the regridding model's grid (only for the cases in which it is specified, in the other cases, the original model's grid is mantained), and cache the result (to avoid redundant downloads and processing).\n\nOriginal model\nInterpolated model\n\n```text\nmodel='access_cm2'\nmodel='awi_cm_1_1_mr'\nmodel='cmcc_esm2'\nmodel='cnrm_cm6_1_hr'\nmodel='cnrm_esm2_1'\nmodel='ec_earth3_cc'\nmodel='gfdl_esm4'\nmodel='inm_cm4_8'\nmodel='inm_cm5_0'\nmodel='kiost_esm'\nmodel='mpi_esm1_2_lr'\nmodel='miroc6'\nmodel='miroc_es2l'\nmodel='mri_esm2_0'\nmodel='noresm2_mm'\nmodel='nesm3'\n```\n\n(section-2.3)=", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes of energy consumption in Europe\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q02 | Category: Climate_Projections\nSection: Uncertainty in projected changes of energy consumption in Europe > Analysis and results > 2. Downloading and processing > 2.2. Download and transform models\n---\nIn this section, the `download.download_and_transform` function from the 'c3s_eqc_automatic_quality_control' package is employed to download daily data from the CMIP6 models, compute the energy-consumption-related indices for the selected temporal aggregation (\"DJF\" for HDD and \"JJA\" for CDD), calculate the mean and trend over the future period (2016-2099), interpolate to the regridding model's grid (only for the cases in which it is specified, in the other cases, the original model's grid is mantained), and cache the result (to avoid redundant downloads and processing).\n\nOriginal model\nInterpolated model\n\n```text\nmodel='access_cm2'\nmodel='awi_cm_1_1_mr'\nmodel='cmcc_esm2'\nmodel='cnrm_cm6_1_hr'\nmodel='cnrm_esm2_1'\nmodel='ec_earth3_cc'\nmodel='gfdl_esm4'\nmodel='inm_cm4_8'\nmodel='inm_cm5_0'\nmodel='kiost_esm'\nmodel='mpi_esm1_2_lr'\nmodel='miroc6'\nmodel='miroc_es2l'\nmodel='mri_esm2_0'\nmodel='noresm2_mm'\nmodel='nesm3'\n```\n\n(section-2.3)="} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q02__010e9ce99bfa", "report_id": "climate_projections-cmip6_climate-impact-indicators_q02", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q02", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes of energy consumption in Europe > Analysis and results > 2. Downloading and processing > 2.3. Apply land-sea mask, change attributes and cut the region to show", "title": "Uncertainty in projected changes of energy consumption in Europe", "chunk_index": 11, "token_count": 230, "text_raw": "This section performs the following tasks:\n\n1. Cut the region of interest.\n2. Downloads the sea mask for ERA5.\n3. Regrids ERA5's mask to the `model_regrid` grid and applies it to the regridded data\n4. Regrids the ERA5 land-sea mask to the model's original grid and applies it to them.\n5. Change some variable attributes for plotting purposes.\n\n**Note:** `ds_interpolated` contains data from the models regridded to the regridding model's grid. `model_datasets` contain the same data but in the original grid of each model.\n\nEdit attributes\n\n(section-3)=", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes of energy consumption in Europe\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q02 | Category: Climate_Projections\nSection: Uncertainty in projected changes of energy consumption in Europe > Analysis and results > 2. Downloading and processing > 2.3. Apply land-sea mask, change attributes and cut the region to show\n---\nThis section performs the following tasks:\n\n1. Cut the region of interest.\n2. Downloads the sea mask for ERA5.\n3. Regrids ERA5's mask to the `model_regrid` grid and applies it to the regridded data\n4. Regrids the ERA5 land-sea mask to the model's original grid and applies it to them.\n5. Change some variable attributes for plotting purposes.\n\n**Note:** `ds_interpolated` contains data from the models regridded to the regridding model's grid. `model_datasets` contain the same data but in the original grid of each model.\n\nEdit attributes\n\n(section-3)="} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q02__f95f58f95f1f", "report_id": "climate_projections-cmip6_climate-impact-indicators_q02", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q02", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes of energy consumption in Europe > Analysis and results > 3. Plot and describe results", "title": "Uncertainty in projected changes of energy consumption in Europe", "chunk_index": 12, "token_count": 198, "text_raw": "This section will display the following results:\n\n- Maps representing the spatial distribution of the **future trends** (2016-2099) of the indices 'HDD15.5' and 'CDD22' for each model individually, the ensemble median (understood as the median of the trend values of the chosen subset of models calculated for each grid cell), and the ensemble spread (derived as the standard deviation of the distribution of the chosen subset of models).\n- **Boxplots** which represent statistical distributions (PDF) built on the spatially-averaged future trend from each considered model.\n\n(section-3.1)=", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes of energy consumption in Europe\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q02 | Category: Climate_Projections\nSection: Uncertainty in projected changes of energy consumption in Europe > Analysis and results > 3. Plot and describe results\n---\nThis section will display the following results:\n\n- Maps representing the spatial distribution of the **future trends** (2016-2099) of the indices 'HDD15.5' and 'CDD22' for each model individually, the ensemble median (understood as the median of the trend values of the chosen subset of models calculated for each grid cell), and the ensemble spread (derived as the standard deviation of the distribution of the chosen subset of models).\n- **Boxplots** which represent statistical distributions (PDF) built on the spatially-averaged future trend from each considered model.\n\n(section-3.1)="} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q02__5dc79186ca11", "report_id": "climate_projections-cmip6_climate-impact-indicators_q02", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q02", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes of energy consumption in Europe > Analysis and results > 3. Plot and describe results > 3.1. Define plotting functions", "title": "Uncertainty in projected changes of energy consumption in Europe", "chunk_index": 13, "token_count": 759, "text_raw": "The functions presented here are used to plot the trends calculated over the future period (2016-2099) for each of the indices ('HDD15.5' and 'CDD22').\n\nFor a selected index, two layout types will be displayed, depending on the chosen function:\n\n1. Layout including the ensemble median and the ensemble spread for the trend: `plot_ensemble()` is used.\n2. Layout including every model trend: `plot_models()` is employed.\n\n`trend==True` allows displaying trend values over the future period, while `trend==False` show mean values. In this notebook, which focuses on the future period, only trend values will be shown, and, consequently, `trend==True`. When the `trend` argument is set to True, regions with no significance are hatched. For individual models, a grid point is considered to have a statistically significant trend when the p-value is lower than 0.05 (in such cases, no hatching is shown). However, for determining trend significance for the ensemble median (understood as the median of the trend values of the chosen subset of models calculated for each grid cell), reliance is placed on agreement categories, following the advanced approach proposed in AR6 [IPCC](https://www.ipcc.ch/report/ar6/wg1/downloads/report/IPCC_AR6_WGI_Atlas.pdf) on pages 1945-1950. The `hatch_p_value_ensemble()` function is used to distinguish, for each grid point, between three possible cases:\n\n1. If more than 66% of the models are statistically significant (p-value < 0.05) and more than 80% of the models share the same sign, we consider the ensemble median trend to be statistically significant, and there is agreement on the sign. To represent this, no hatching is used.\n2. If less than 66% of the models are statistically significant, regardless of agreement on the sign of the trend, hatching is applied (indicating that the ensemble median trend is not statistically significant).\n3. If more than 66% of the models are statistically significant but less than 80% of the models share the same sign, we consider the ensemble median trend to be statistically significant, but there is no agreement on the sign of the trend. This is represented using crosses.\n\nDefine function to plot the caption of the figures (for the ensemble case)\nAdd caption to the figure\nAdd each line to the figure\nend captioning\nDefine function to plot the caption of the figures (for the individual models case)\nAdd caption to the figure\nAdd each line to the figure\n\n\n
\nCOLORBAR NOTE:
\nThe colorbar chosen to represent trends for the HDD15.5 index spans from red to blue. In this color scheme, negative trend values (shown in red) indicate a decrease in Heating Degree Days over time. Conversely, blueish colors indicate an increase in Heating Degree Days over time. This selection is based on the rationale that more Heating Degree Days are associated with colder conditions, typically represented by blueish colors, while fewer Heating Degree Days indicate warmer conditions, depicted by reddish colors.\n\n(section-3.2)=", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes of energy consumption in Europe\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q02 | Category: Climate_Projections\nSection: Uncertainty in projected changes of energy consumption in Europe > Analysis and results > 3. Plot and describe results > 3.1. Define plotting functions\n---\nThe functions presented here are used to plot the trends calculated over the future period (2016-2099) for each of the indices ('HDD15.5' and 'CDD22').\n\nFor a selected index, two layout types will be displayed, depending on the chosen function:\n\n1. Layout including the ensemble median and the ensemble spread for the trend: `plot_ensemble()` is used.\n2. Layout including every model trend: `plot_models()` is employed.\n\n`trend==True` allows displaying trend values over the future period, while `trend==False` show mean values. In this notebook, which focuses on the future period, only trend values will be shown, and, consequently, `trend==True`. When the `trend` argument is set to True, regions with no significance are hatched. For individual models, a grid point is considered to have a statistically significant trend when the p-value is lower than 0.05 (in such cases, no hatching is shown). However, for determining trend significance for the ensemble median (understood as the median of the trend values of the chosen subset of models calculated for each grid cell), reliance is placed on agreement categories, following the advanced approach proposed in AR6 [IPCC](https://www.ipcc.ch/report/ar6/wg1/downloads/report/IPCC_AR6_WGI_Atlas.pdf) on pages 1945-1950. The `hatch_p_value_ensemble()` function is used to distinguish, for each grid point, between three possible cases:\n\n1. If more than 66% of the models are statistically significant (p-value < 0.05) and more than 80% of the models share the same sign, we consider the ensemble median trend to be statistically significant, and there is agreement on the sign. To represent this, no hatching is used.\n2. If less than 66% of the models are statistically significant, regardless of agreement on the sign of the trend, hatching is applied (indicating that the ensemble median trend is not statistically significant).\n3. If more than 66% of the models are statistically significant but less than 80% of the models share the same sign, we consider the ensemble median trend to be statistically significant, but there is no agreement on the sign of the trend. This is represented using crosses.\n\nDefine function to plot the caption of the figures (for the ensemble case)\nAdd caption to the figure\nAdd each line to the figure\nend captioning\nDefine function to plot the caption of the figures (for the individual models case)\nAdd caption to the figure\nAdd each line to the figure\n\n\n
\nCOLORBAR NOTE:
\nThe colorbar chosen to represent trends for the HDD15.5 index spans from red to blue. In this color scheme, negative trend values (shown in red) indicate a decrease in Heating Degree Days over time. Conversely, blueish colors indicate an increase in Heating Degree Days over time. This selection is based on the rationale that more Heating Degree Days are associated with colder conditions, typically represented by blueish colors, while fewer Heating Degree Days indicate warmer conditions, depicted by reddish colors.\n\n(section-3.2)="} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q02__22a472539e45", "report_id": "climate_projections-cmip6_climate-impact-indicators_q02", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q02", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes of energy consumption in Europe > Analysis and results > 3. Plot and describe results > 3.2. Plot ensemble maps", "title": "Uncertainty in projected changes of energy consumption in Europe", "chunk_index": 14, "token_count": 260, "text_raw": "In this section, we invoke the `plot_ensemble()` function to visualise the trend calculated over the future period (2016-2099) for the model ensemble across Europe. Note that the model data used in this section has previously been interpolated to the \"regridding model\" grid (`\"gfdl_esm4\"` for this notebook).\n\nSpecifically, for each of the indices ('HDD15.5' and 'CDD22'), this section presents a single layout including trend values of the future period (2016-2099) for: (a) the ensemble median (understood as the median of the trend values of the chosen subset of models calculated for each grid cell) and (b) the ensemble spread (derived as the standard deviation of the distribution of the chosen subset of models).\n\nFig number counter\nCommon title\n\n(section-3.3)=", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes of energy consumption in Europe\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q02 | Category: Climate_Projections\nSection: Uncertainty in projected changes of energy consumption in Europe > Analysis and results > 3. Plot and describe results > 3.2. Plot ensemble maps\n---\nIn this section, we invoke the `plot_ensemble()` function to visualise the trend calculated over the future period (2016-2099) for the model ensemble across Europe. Note that the model data used in this section has previously been interpolated to the \"regridding model\" grid (`\"gfdl_esm4\"` for this notebook).\n\nSpecifically, for each of the indices ('HDD15.5' and 'CDD22'), this section presents a single layout including trend values of the future period (2016-2099) for: (a) the ensemble median (understood as the median of the trend values of the chosen subset of models calculated for each grid cell) and (b) the ensemble spread (derived as the standard deviation of the distribution of the chosen subset of models).\n\nFig number counter\nCommon title\n\n(section-3.3)="} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q02__e67f2ff6fb4f", "report_id": "climate_projections-cmip6_climate-impact-indicators_q02", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q02", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes of energy consumption in Europe > Analysis and results > 3. Plot and describe results > 3.3. Plot model maps", "title": "Uncertainty in projected changes of energy consumption in Europe", "chunk_index": 15, "token_count": 182, "text_raw": "In this section, we invoke the `plot_models()` function to visualise the trend calculated over the future period (2016-2099) for every model individually across Europe. Note that the model data used in this section maintains its original grid.\n\nSpecifically, for each of the indices ('HDD15.5' and 'CDD22'), this section presents a single layout including the trend for the future period (2016-2099) of every model.\n\n(section-3.4)=", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes of energy consumption in Europe\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q02 | Category: Climate_Projections\nSection: Uncertainty in projected changes of energy consumption in Europe > Analysis and results > 3. Plot and describe results > 3.3. Plot model maps\n---\nIn this section, we invoke the `plot_models()` function to visualise the trend calculated over the future period (2016-2099) for every model individually across Europe. Note that the model data used in this section maintains its original grid.\n\nSpecifically, for each of the indices ('HDD15.5' and 'CDD22'), this section presents a single layout including the trend for the future period (2016-2099) of every model.\n\n(section-3.4)="} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q02__d790459f8d54", "report_id": "climate_projections-cmip6_climate-impact-indicators_q02", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q02", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes of energy consumption in Europe > Analysis and results > 3. Plot and describe results > 3.4. Boxplots of the future trend", "title": "Uncertainty in projected changes of energy consumption in Europe", "chunk_index": 16, "token_count": 396, "text_raw": "Finally, we present boxplots representing the ensemble distribution of each climate model trend calculated over the future period (2016-2099) across Europe.\n\nDots represent the spatially-averaged future trend over the selected region (change of the number of days per decade) for each model (grey) and the ensemble mean (blue). The ensemble median is shown as a green line. Note that the spatially averaged values are calculated for each model from its original grid (i.e., no interpolated data has been used here).\n\nThe boxplot visually illustrates the distribution of trends among the climate models, with the box covering the first quartile (Q1 = 25th percentile) to the third quartile (Q3 = 75th percentile), and a green line indicating the ensemble median (Q2 = 50th percentile). Whiskers extend from the edges of the box to show the full data range.\n\nEnsemble mean\n\n
\n
\n

Fig 5. Boxplots illustrating the future trends of the distribution of the chosen subset of models for the Energy Degree Days indices daily averaged: (a) 'CDD22', and (b) 'HDD15.5'. The distribution is created by considering spatially averaged trends across Europe. The ensemble mean and the ensemble median trends are both included. Outliers in the distribution are denoted by a grey circle with a black contour.

\n

\n\n(section-3.5)=", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes of energy consumption in Europe\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q02 | Category: Climate_Projections\nSection: Uncertainty in projected changes of energy consumption in Europe > Analysis and results > 3. Plot and describe results > 3.4. Boxplots of the future trend\n---\nFinally, we present boxplots representing the ensemble distribution of each climate model trend calculated over the future period (2016-2099) across Europe.\n\nDots represent the spatially-averaged future trend over the selected region (change of the number of days per decade) for each model (grey) and the ensemble mean (blue). The ensemble median is shown as a green line. Note that the spatially averaged values are calculated for each model from its original grid (i.e., no interpolated data has been used here).\n\nThe boxplot visually illustrates the distribution of trends among the climate models, with the box covering the first quartile (Q1 = 25th percentile) to the third quartile (Q3 = 75th percentile), and a green line indicating the ensemble median (Q2 = 50th percentile). Whiskers extend from the edges of the box to show the full data range.\n\nEnsemble mean\n\n
\n
\n

Fig 5. Boxplots illustrating the future trends of the distribution of the chosen subset of models for the Energy Degree Days indices daily averaged: (a) 'CDD22', and (b) 'HDD15.5'. The distribution is created by considering spatially averaged trends across Europe. The ensemble mean and the ensemble median trends are both included. Outliers in the distribution are denoted by a grey circle with a black contour.

\n

\n\n(section-3.5)="} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q02__b1e88adc017d", "report_id": "climate_projections-cmip6_climate-impact-indicators_q02", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q02", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes of energy consumption in Europe > Analysis and results > 3. Plot and describe results > 3.5. Results summary and discussion", "title": "Uncertainty in projected changes of energy consumption in Europe", "chunk_index": 17, "token_count": 613, "text_raw": "- Trends calculated for the future period (2016-2099) exhibit a general decrease in Heating Degree Days (HDD15.5) across Europe for DJF, especially in the northern and central-eastern regions where winter Heating Degree Days are typically higher. For Cooling Degree Days (CDD22) trends calculated for summer in Europe, an increase can be observed, particularly in the southern half and central regions. However, no trend is observed for some northern regions and certain mountain areas, possibly due to the threshold temperature of 22°C being too high to be reached across these regions.\n\n- The boxplots illustrate a decrease in Heating Degree Days (HDD15.5) across Europe for DJF during the future period from 2016 to 2099. The ensemble median trend reaches a daily average value near -0.23 °C per decade, which is a larger decrease than that captured by ERA5 for the historical period (around -0.15 °C per decade). The interquantile range of the ensemble ranges from -0.275 to around -0.225 °C per decade.\n\n- The boxplot analysis reveals an increase in Cooling Degree Days (CDD22) across Europe for JJA, with an ensemble median trend value near 0.25 °C per decade and an interquantile range that spans approximately from 0.20 to 0.27 °C per decade. This increase is again larger than that reflected by the ERA5 calculations over the historical period (which is around 0.11°C/decade).\n\n- What do the results mean for users? Are the biases relevant?\n\n- The projected decrease in Heating Degree Days during winter and the projected increase in Cooling Degree Days during summer provide valuable information for decisions sensitive to future energy demand. However, it is crucial to consider the biases identified during the historical period (1971-2000) in the assessment \"CMIP6 Climate Projections: evaluating biases in energy-consumption-related indices in Europe\". Depending on the region, these biases can either enhance or diminish confidence in interpreting the assessment results. The inter-model spread, which is used to account for projected uncertainties, should also be taken into account, as these uncertainties can lead to misinterpretations regarding the magnitude of projected increases or decreases.\n\n\n
\nRESULTS NOTE:
\nIt is important to note that the results presented are specific to the 16 models chosen, and users should aim to assess as wide a range of models as possible before making a sub-selection.", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes of energy consumption in Europe\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q02 | Category: Climate_Projections\nSection: Uncertainty in projected changes of energy consumption in Europe > Analysis and results > 3. Plot and describe results > 3.5. Results summary and discussion\n---\n- Trends calculated for the future period (2016-2099) exhibit a general decrease in Heating Degree Days (HDD15.5) across Europe for DJF, especially in the northern and central-eastern regions where winter Heating Degree Days are typically higher. For Cooling Degree Days (CDD22) trends calculated for summer in Europe, an increase can be observed, particularly in the southern half and central regions. However, no trend is observed for some northern regions and certain mountain areas, possibly due to the threshold temperature of 22°C being too high to be reached across these regions.\n\n- The boxplots illustrate a decrease in Heating Degree Days (HDD15.5) across Europe for DJF during the future period from 2016 to 2099. The ensemble median trend reaches a daily average value near -0.23 °C per decade, which is a larger decrease than that captured by ERA5 for the historical period (around -0.15 °C per decade). The interquantile range of the ensemble ranges from -0.275 to around -0.225 °C per decade.\n\n- The boxplot analysis reveals an increase in Cooling Degree Days (CDD22) across Europe for JJA, with an ensemble median trend value near 0.25 °C per decade and an interquantile range that spans approximately from 0.20 to 0.27 °C per decade. This increase is again larger than that reflected by the ERA5 calculations over the historical period (which is around 0.11°C/decade).\n\n- What do the results mean for users? Are the biases relevant?\n\n- The projected decrease in Heating Degree Days during winter and the projected increase in Cooling Degree Days during summer provide valuable information for decisions sensitive to future energy demand. However, it is crucial to consider the biases identified during the historical period (1971-2000) in the assessment \"CMIP6 Climate Projections: evaluating biases in energy-consumption-related indices in Europe\". Depending on the region, these biases can either enhance or diminish confidence in interpreting the assessment results. The inter-model spread, which is used to account for projected uncertainties, should also be taken into account, as these uncertainties can lead to misinterpretations regarding the magnitude of projected increases or decreases.\n\n\n
\nRESULTS NOTE:
\nIt is important to note that the results presented are specific to the 16 models chosen, and users should aim to assess as wide a range of models as possible before making a sub-selection."} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q02__5374652e3485", "report_id": "climate_projections-cmip6_climate-impact-indicators_q02", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q02", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes of energy consumption in Europe > ℹ️ If you want to know more > Key resources", "title": "Uncertainty in projected changes of energy consumption in Europe", "chunk_index": 18, "token_count": 219, "text_raw": "Some key resources and further reading were linked throughout this assessment.\n\nThe CDS catalogue entries for the data used were:\n* CMIP6 climate projections (Daily - air temperature): https://cds.climate.copernicus.eu/datasets/projections-cmip6?tab=overview\n\nCode libraries used:\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)\n* [icclim](https://icclim.readthedocs.io/en/stable/) Python package", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes of energy consumption in Europe\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q02 | Category: Climate_Projections\nSection: Uncertainty in projected changes of energy consumption in Europe > ℹ️ If you want to know more > Key resources\n---\nSome key resources and further reading were linked throughout this assessment.\n\nThe CDS catalogue entries for the data used were:\n* CMIP6 climate projections (Daily - air temperature): https://cds.climate.copernicus.eu/datasets/projections-cmip6?tab=overview\n\nCode libraries used:\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)\n* [icclim](https://icclim.readthedocs.io/en/stable/) Python package"} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q02__3019e94aa926", "report_id": "climate_projections-cmip6_climate-impact-indicators_q02", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q02", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes of energy consumption in Europe > ℹ️ If you want to know more > References", "title": "Uncertainty in projected changes of energy consumption in Europe", "chunk_index": 19, "token_count": 791, "text_raw": "[[1]](https://doi.org/10.1007/s10113-010-0173-x) Bindi, M., Olesen, J.E. The responses of agriculture in Europe to climate change (2011). Reg Environ Change 11 (Suppl 1), 151–158. https://doi.org/10.1007/s10113-010-0173-x\n\n[[2]](https://doi.org/10.1016/j.foreco.2009.09.023) Lindner, M., Maroschek, M., Netherer, S., Kremer, A., Barbati, A., Garcia-Gonzalo, J., Seidi, R., Delzon, S., Corona, P., Kolstrom, M., Lexer, M.J., Marchetti, M. (2010). Climate change impacts, adaptive capacity, and vulnerability of European forest ecosystems. For. Ecol. Manage. 259(4): 698–709. https://doi.org/10.1016/j.foreco.2009.09.023\n\n[[3]](https://doi.org/10.1016/j.enbuild.2014.09.052) Santamouris, M., Cartalis, C., Synnefa, A., Kolokotsa, D. (2015). On the impact of urban heat island and global warming on the power demand and electricity consumption of buildings – a review. Energy Build. 98: 119–124. https://doi.org/10.1016/j.enbuild.2014.09.052\n\n[[4]](https://doi.org/10.1007/s10113-013-0499-2) Jacob, D., Petersen, J., Eggert, B. et al. (2014). EURO-CORDEX: new high-resolution climate change projections for European impact research. Reg Environ Change 14, 563–578. https://doi.org/10.1007/s10113-013-0499-2\n\n[[5]](http://hdl.handle.net/10013/epic.45156.d001) IPCC. 2014. In Climate Change 2014: Synthesis Report. Contribution of Working Groups I, II and III to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change, Core Writing Team, RK Pachauri, LA Meyer (eds). IPCC: Geneva, Switzerland 151 pp.\n\n[[6]](https://doi.org/10.1002/joc.5362) Spinoni, J., Vogt, J.V., Barbosa, P., Dosio, A., McCormick, N., Bigano, A. and Füssel, H.-M. (2018). Changes of heating and cooling degree-days in Europe from 1981 to 2100. Int. J. Climatol, 38: e191-e208. https://doi.org/10.1002/joc.5362\n\n[[7]](https://doi.org/10.1038/s43247-023-00878-3) Scoccimarro, E., Cattaneo, O., Gualdi, S. et al. (2023). Country-level energy demand for cooling has increased over the past two decades. Commun Earth Environ 4, 208. https://doi.org/10.1038/s43247-023-00878-3", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes of energy consumption in Europe\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q02 | Category: Climate_Projections\nSection: Uncertainty in projected changes of energy consumption in Europe > ℹ️ If you want to know more > References\n---\n[[1]](https://doi.org/10.1007/s10113-010-0173-x) Bindi, M., Olesen, J.E. The responses of agriculture in Europe to climate change (2011). Reg Environ Change 11 (Suppl 1), 151–158. https://doi.org/10.1007/s10113-010-0173-x\n\n[[2]](https://doi.org/10.1016/j.foreco.2009.09.023) Lindner, M., Maroschek, M., Netherer, S., Kremer, A., Barbati, A., Garcia-Gonzalo, J., Seidi, R., Delzon, S., Corona, P., Kolstrom, M., Lexer, M.J., Marchetti, M. (2010). Climate change impacts, adaptive capacity, and vulnerability of European forest ecosystems. For. Ecol. Manage. 259(4): 698–709. https://doi.org/10.1016/j.foreco.2009.09.023\n\n[[3]](https://doi.org/10.1016/j.enbuild.2014.09.052) Santamouris, M., Cartalis, C., Synnefa, A., Kolokotsa, D. (2015). On the impact of urban heat island and global warming on the power demand and electricity consumption of buildings – a review. Energy Build. 98: 119–124. https://doi.org/10.1016/j.enbuild.2014.09.052\n\n[[4]](https://doi.org/10.1007/s10113-013-0499-2) Jacob, D., Petersen, J., Eggert, B. et al. (2014). EURO-CORDEX: new high-resolution climate change projections for European impact research. Reg Environ Change 14, 563–578. https://doi.org/10.1007/s10113-013-0499-2\n\n[[5]](http://hdl.handle.net/10013/epic.45156.d001) IPCC. 2014. In Climate Change 2014: Synthesis Report. Contribution of Working Groups I, II and III to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change, Core Writing Team, RK Pachauri, LA Meyer (eds). IPCC: Geneva, Switzerland 151 pp.\n\n[[6]](https://doi.org/10.1002/joc.5362) Spinoni, J., Vogt, J.V., Barbosa, P., Dosio, A., McCormick, N., Bigano, A. and Füssel, H.-M. (2018). Changes of heating and cooling degree-days in Europe from 1981 to 2100. Int. J. Climatol, 38: e191-e208. https://doi.org/10.1002/joc.5362\n\n[[7]](https://doi.org/10.1038/s43247-023-00878-3) Scoccimarro, E., Cattaneo, O., Gualdi, S. et al. (2023). Country-level energy demand for cooling has increased over the past two decades. Commun Earth Environ 4, 208. https://doi.org/10.1038/s43247-023-00878-3"} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q03__d3a5fe8864e9", "report_id": "climate_projections-cmip6_climate-impact-indicators_q03", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q03", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level", "title": "Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level", "chunk_index": 0, "token_count": 112, "text_raw": "Production date: 23-05-2024\n\nProduced by: CMCC foundation - Euro-Mediterranean Center on Climate Change. Albert Martinez Boti.", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q03 | Category: Climate_Projections\nSection: Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level\n---\nProduction date: 23-05-2024\n\nProduced by: CMCC foundation - Euro-Mediterranean Center on Climate Change. Albert Martinez Boti."} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q03__666edf415dd5", "report_id": "climate_projections-cmip6_climate-impact-indicators_q03", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q03", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level > Quality assessment question", "title": "Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level", "chunk_index": 1, "token_count": 767, "text_raw": "* **What are the projected changes for a global warming level of 2°C and associated uncertainties of Energy Degree Days in Europe?**\n\nSectors affected by climate change are varied including agriculture [[1]](https://doi.org/10.1007/s10113-010-0173-x), forest ecosystems [[2]](https://doi.org/10.1016/j.foreco.2009.09.023), and energy consumption [[3]](https://doi.org/10.1016/j.enbuild.2014.09.052). Under projected future global warming over Europe [[4]](https://doi.org/10.1007/s10113-013-0499-2)[[5]](http://hdl.handle.net/10013/epic.45156.d001), the current increase in energy demand is expected to persist until the end of this century and beyond [[6]](https://doi.org/10.1002/joc.5362). Identifying which climate-change-related impacts are likely to increase, by how much, and inherent regional patterns, is important for any effective strategy for managing future climate risks. This notebook utilises data from a subset of models from CMIP6 Global Climate Models (GCMs) and explores the projected changes and uncertainties in future projections of energy-consumption-related indices by considering the ensemble inter-model spread of a subset of CMIP6 models at a global mean warming level of 2°C. Two energy-consumption-related indices are calculated from daily mean temperatures using the [icclim](https://icclim.readthedocs.io/en/stable/) Python package: Cooling Degree Days (CDDs) and Heating Degree Days (HDDs). Degree days measure how much warmer or colder it is compared to standard temperatures (usually 15.5°C for heating and 22°C for cooling). Higher degree day numbers indicate more extreme temperatures, which typically lead to increased energy use for heating or cooling buildings. In the presented code, CDD calculations use summer aggregation (CDD22), while HDD calculations focus on winter (HDD15.5), presenting results as daily averages rather than cumulative values. In this notebook, these calculations are performed over the historical period from 1971 to 2000 and compared to the global warming level of 2°C. The global warming level of 2°C can be defined as the first time the 30-year moving average (centre year) of global temperature is above 2°C compared to pre-industrial (Grigory Nikulin et al. (2018) [[7]](https://doi.org/10.1088/1748-9326/aab1b1)). The preindustrial period is defined here as the period spanning from 1861 to 1890 and the index calculations are performed for the Shared Socioeconomic Pathways SSP5-8.5. It is important to note that the results presented here pertain to a specific subset of the CMIP6 ensemble and may not be generalisable to the entire dataset. Also note that a separate assessment examines the representation of climatology and trends of these indices for the same models during the historical period (1971-2000), while another assessment looks at the projected trends of these indices for the same models during a fixed future period (2015-2099).", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q03 | Category: Climate_Projections\nSection: Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level > Quality assessment question\n---\n* **What are the projected changes for a global warming level of 2°C and associated uncertainties of Energy Degree Days in Europe?**\n\nSectors affected by climate change are varied including agriculture [[1]](https://doi.org/10.1007/s10113-010-0173-x), forest ecosystems [[2]](https://doi.org/10.1016/j.foreco.2009.09.023), and energy consumption [[3]](https://doi.org/10.1016/j.enbuild.2014.09.052). Under projected future global warming over Europe [[4]](https://doi.org/10.1007/s10113-013-0499-2)[[5]](http://hdl.handle.net/10013/epic.45156.d001), the current increase in energy demand is expected to persist until the end of this century and beyond [[6]](https://doi.org/10.1002/joc.5362). Identifying which climate-change-related impacts are likely to increase, by how much, and inherent regional patterns, is important for any effective strategy for managing future climate risks. This notebook utilises data from a subset of models from CMIP6 Global Climate Models (GCMs) and explores the projected changes and uncertainties in future projections of energy-consumption-related indices by considering the ensemble inter-model spread of a subset of CMIP6 models at a global mean warming level of 2°C. Two energy-consumption-related indices are calculated from daily mean temperatures using the [icclim](https://icclim.readthedocs.io/en/stable/) Python package: Cooling Degree Days (CDDs) and Heating Degree Days (HDDs). Degree days measure how much warmer or colder it is compared to standard temperatures (usually 15.5°C for heating and 22°C for cooling). Higher degree day numbers indicate more extreme temperatures, which typically lead to increased energy use for heating or cooling buildings. In the presented code, CDD calculations use summer aggregation (CDD22), while HDD calculations focus on winter (HDD15.5), presenting results as daily averages rather than cumulative values. In this notebook, these calculations are performed over the historical period from 1971 to 2000 and compared to the global warming level of 2°C. The global warming level of 2°C can be defined as the first time the 30-year moving average (centre year) of global temperature is above 2°C compared to pre-industrial (Grigory Nikulin et al. (2018) [[7]](https://doi.org/10.1088/1748-9326/aab1b1)). The preindustrial period is defined here as the period spanning from 1861 to 1890 and the index calculations are performed for the Shared Socioeconomic Pathways SSP5-8.5. It is important to note that the results presented here pertain to a specific subset of the CMIP6 ensemble and may not be generalisable to the entire dataset. Also note that a separate assessment examines the representation of climatology and trends of these indices for the same models during the historical period (1971-2000), while another assessment looks at the projected trends of these indices for the same models during a fixed future period (2015-2099)."} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q03__a01c4006d14c", "report_id": "climate_projections-cmip6_climate-impact-indicators_q03", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q03", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level > Quality assessment statement", "title": "Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level", "chunk_index": 2, "token_count": 510, "text_raw": "These are the key outcomes of this assessment\n\n* Under a 2°C global warming scenario, projections indicate further changes in Energy Degree Days for the selected subset of models.\n\n* HDD15.5 is projected to decrease across Europe for DJF, with the largest decreases expected in northern and continental Europe. Conversely, CDD22 is expected to increase, with the Mediterranean Basin experiencing the most significant rise.\n\n* Despite a clear agreement on the sign of the climate signal for the selected models, differences on the magnitude of the climate signal between models is high and larger ensemble should be considered when addressing specific cases to enhance the robustness of the analysis and account for uncertainties.\n\n* The projected decrease in Heating Degree Days during winter and the projected increase in Cooling Degree Days during summer in a world 2°C warmer than the preindustrial baseline provide valuable information for decisions sensitive to future energy demand. Using climate global warming levels instead of fixed future periods helps mitigate the systematic biases of the models [[8]](https://doi.org/10.5194/esd-13-321-2022). However, it is important to note that the centered thirty-year period when the global mean temperature reaches 2°C above the preindustrial baseline varies depending on the model considered, making it challenging to determine this timing. Users must identify whether their interest lies in projected values for fixed periods or in working with global warming levels. A combined approach may provide a more comprehensive assessment.\n```\n\nattachment:1b204289-a2d7-4cee-bad9-d5f270df7baf.png \n---\nalt: GWL_CDD22.png\nwidth: 550px\n---\nBoxplot illustrating the climate signal (i.e., the mean values for the warming level of 2°C compared to the historical period from 1971 to 2000) for the ensemble distribution of the 'CDD22' index daily averaged. The distribution is created by considering spatially averaged values across Europe. The ensemble mean and the ensemble median are both included. Outliers in the distribution are denoted by a grey circle with a black contour.\n```", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q03 | Category: Climate_Projections\nSection: Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level > Quality assessment statement\n---\nThese are the key outcomes of this assessment\n\n* Under a 2°C global warming scenario, projections indicate further changes in Energy Degree Days for the selected subset of models.\n\n* HDD15.5 is projected to decrease across Europe for DJF, with the largest decreases expected in northern and continental Europe. Conversely, CDD22 is expected to increase, with the Mediterranean Basin experiencing the most significant rise.\n\n* Despite a clear agreement on the sign of the climate signal for the selected models, differences on the magnitude of the climate signal between models is high and larger ensemble should be considered when addressing specific cases to enhance the robustness of the analysis and account for uncertainties.\n\n* The projected decrease in Heating Degree Days during winter and the projected increase in Cooling Degree Days during summer in a world 2°C warmer than the preindustrial baseline provide valuable information for decisions sensitive to future energy demand. Using climate global warming levels instead of fixed future periods helps mitigate the systematic biases of the models [[8]](https://doi.org/10.5194/esd-13-321-2022). However, it is important to note that the centered thirty-year period when the global mean temperature reaches 2°C above the preindustrial baseline varies depending on the model considered, making it challenging to determine this timing. Users must identify whether their interest lies in projected values for fixed periods or in working with global warming levels. A combined approach may provide a more comprehensive assessment.\n```\n\nattachment:1b204289-a2d7-4cee-bad9-d5f270df7baf.png \n---\nalt: GWL_CDD22.png\nwidth: 550px\n---\nBoxplot illustrating the climate signal (i.e., the mean values for the warming level of 2°C compared to the historical period from 1971 to 2000) for the ensemble distribution of the 'CDD22' index daily averaged. The distribution is created by considering spatially averaged values across Europe. The ensemble mean and the ensemble median are both included. Outliers in the distribution are denoted by a grey circle with a black contour.\n```"} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q03__783d1d33f66f", "report_id": "climate_projections-cmip6_climate-impact-indicators_q03", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q03", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level > Methodology", "title": "Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level", "chunk_index": 3, "token_count": 1026, "text_raw": "The reference methodology used here for the indices calculation is similar to the one followed by Scoccimarro et al., (2023) [[9]](https://doi.org/10.1038/s43247-023-00878-3). However the thermal comfort thresholds used in this notebook are slightly different. A winter comfort temperature of 15.5°C and a summer comfort temperature of 22.0°C are used here (as in the [CDS application](https://dast.copernicus-climate.eu/documents/app-heating-cooling-degree-days/C3S_EEA_HDD_CDD_application_user_guide_v0.9.pdf)). In the presented code, the CDD calculations are based on the JJA aggregation, with a comfort temperature of 22°C (CDD22), while HDD calculations focus on winter (DJF) with a comfort temperature of 15.5°C (HDD15.5). More specifically, to calculate CDD22, the sum of the differences between the daily mean temperature and the thermal comfort temperature of 22°C is computed. This calculation occurs only when the mean temperature is above the thermal comfort level; otherwise, the CDD22 for that day is set to 0. For example, a day with a mean temperature of 28°C would result in 6°C. Two consecutive hot days like this would total 12°C over the two-day period. Similarly, to calculate HDD15.5, the sum of the differences between the thermal comfort temperature of 15.5°C and the daily mean temperature is determined. This happens only when the mean temperature is below the thermal comfort level; otherwise, the HDD15.5 for that day is set to 0. Finally, to obtain more intuitive values, the sum is averaged over the number of days in the season to produce daily average values. This approach differs from the [CDS application](https://dast.copernicus-climate.eu/documents/app-heating-cooling-degree-days/C3S_EEA_HDD_CDD_application_user_guide_v0.9.pdf), where both the sum over a period and the daily average values can be displayed. In Spinoni et al., (2018) [[6]](https://doi.org/10.1002/joc.5362), as well as in the [CDS application](https://dast.copernicus-climate.eu/documents/app-heating-cooling-degree-days/C3S_EEA_HDD_CDD_application_user_guide_v0.9.pdf), more advanced methods for calculating CDD and HDD involve considering maximum, minimum, and mean temperatures. However, to prevent overloading the notebook and maintain simplicity while ensuring compatibility with the [icclim](https://icclim.readthedocs.io/en/stable/) Python package, we opted to utilise a single variable (2m mean temperature).\n\nThis notebook provides an assessment of the projected changes and their associated uncertainties, utilising a subset of 16 models from [CMIP6](https://cds.climate.copernicus.eu/datasets/projections-cmip6?tab=overview) under a global warming level of 2°C. The uncertainty is explored by analysing the ensemble inter-model spread of projected changes the energy-consumption-related indices 'HDD15.5' and 'CDD22', calculated for the specific global warming level of 2°C. In particular, spatial patterns of the climate signal (i.e., mean values for the warming level of 2°C compared to the historical period from 1971 to 2000) are examined and displayed for each model individually and for the ensemble median (calculated for each grid cell), alongside the ensemble inter-model spread to account for projected uncertainty. Additionally, spatially-averaged values are analysed and presented using box plots to provide an overview of climate signal behavior across the distribution of the chosen subset of models when averaged across Europe.\n\nThe analysis and results follow the next outline:\n\n**[](section-1)**\n * [](section-1.1)\n * [](section-1.2)\n * [](section-1.3)\n * [](section-1.4)\n * [](section-1.5)\n * [](section-1.6)\n\n**[](section-2)**\n * [](section-2.1)\n * [](section-2.2)\n * [](section-2.3)\n\n**[](section-3)**\n * [](section-3.1)\n * [](section-3.2)\n * [](section-3.3)\n * [](section-3.4)\n * [](section-3.5)", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q03 | Category: Climate_Projections\nSection: Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level > Methodology\n---\nThe reference methodology used here for the indices calculation is similar to the one followed by Scoccimarro et al., (2023) [[9]](https://doi.org/10.1038/s43247-023-00878-3). However the thermal comfort thresholds used in this notebook are slightly different. A winter comfort temperature of 15.5°C and a summer comfort temperature of 22.0°C are used here (as in the [CDS application](https://dast.copernicus-climate.eu/documents/app-heating-cooling-degree-days/C3S_EEA_HDD_CDD_application_user_guide_v0.9.pdf)). In the presented code, the CDD calculations are based on the JJA aggregation, with a comfort temperature of 22°C (CDD22), while HDD calculations focus on winter (DJF) with a comfort temperature of 15.5°C (HDD15.5). More specifically, to calculate CDD22, the sum of the differences between the daily mean temperature and the thermal comfort temperature of 22°C is computed. This calculation occurs only when the mean temperature is above the thermal comfort level; otherwise, the CDD22 for that day is set to 0. For example, a day with a mean temperature of 28°C would result in 6°C. Two consecutive hot days like this would total 12°C over the two-day period. Similarly, to calculate HDD15.5, the sum of the differences between the thermal comfort temperature of 15.5°C and the daily mean temperature is determined. This happens only when the mean temperature is below the thermal comfort level; otherwise, the HDD15.5 for that day is set to 0. Finally, to obtain more intuitive values, the sum is averaged over the number of days in the season to produce daily average values. This approach differs from the [CDS application](https://dast.copernicus-climate.eu/documents/app-heating-cooling-degree-days/C3S_EEA_HDD_CDD_application_user_guide_v0.9.pdf), where both the sum over a period and the daily average values can be displayed. In Spinoni et al., (2018) [[6]](https://doi.org/10.1002/joc.5362), as well as in the [CDS application](https://dast.copernicus-climate.eu/documents/app-heating-cooling-degree-days/C3S_EEA_HDD_CDD_application_user_guide_v0.9.pdf), more advanced methods for calculating CDD and HDD involve considering maximum, minimum, and mean temperatures. However, to prevent overloading the notebook and maintain simplicity while ensuring compatibility with the [icclim](https://icclim.readthedocs.io/en/stable/) Python package, we opted to utilise a single variable (2m mean temperature).\n\nThis notebook provides an assessment of the projected changes and their associated uncertainties, utilising a subset of 16 models from [CMIP6](https://cds.climate.copernicus.eu/datasets/projections-cmip6?tab=overview) under a global warming level of 2°C. The uncertainty is explored by analysing the ensemble inter-model spread of projected changes the energy-consumption-related indices 'HDD15.5' and 'CDD22', calculated for the specific global warming level of 2°C. In particular, spatial patterns of the climate signal (i.e., mean values for the warming level of 2°C compared to the historical period from 1971 to 2000) are examined and displayed for each model individually and for the ensemble median (calculated for each grid cell), alongside the ensemble inter-model spread to account for projected uncertainty. Additionally, spatially-averaged values are analysed and presented using box plots to provide an overview of climate signal behavior across the distribution of the chosen subset of models when averaged across Europe.\n\nThe analysis and results follow the next outline:\n\n**[](section-1)**\n * [](section-1.1)\n * [](section-1.2)\n * [](section-1.3)\n * [](section-1.4)\n * [](section-1.5)\n * [](section-1.6)\n\n**[](section-2)**\n * [](section-2.1)\n * [](section-2.2)\n * [](section-2.3)\n\n**[](section-3)**\n * [](section-3.1)\n * [](section-3.2)\n * [](section-3.3)\n * [](section-3.4)\n * [](section-3.5)"} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q03__cd06aec0c3ba", "report_id": "climate_projections-cmip6_climate-impact-indicators_q03", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q03", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level > Analysis and results > 1. Parameters, requests and functions definition > 1.2. Define Parameters", "title": "Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level", "chunk_index": 4, "token_count": 406, "text_raw": "In the \"Define Parameters\" section, various customisable options for the notebook are specified. Most of the parameters chosen are the same as those used in other assessments ([](./climate_projections-cmip6_climate-impact-indicators_q01)), being them:\n\n- `historical_slice` determines the historical (or control) period used (1971 to 2000 is choosen)\n- `index_timeseries` is a dictionary that set the temporal aggregation for every index considered within this notebook ('HDD15.5' and 'CDD22'). In the presented code, the CDD calculations are always based on the JJA aggregation, with a comfort temperature of 22°C (CDD22), while HDD calculations focus on winter (DJF) with a comfort temperature of 15.5°C (HDD15.5).\n- `collection_id` set the family of models. Only CMIP6 is implemented for this sub-notebook.\n- `area` allows specifying the geographical domain of interest.\n- The `interpolation_method` parameter allows selecting the interpolation method when regridding is performed over the indices.\n- The `chunk` selection allows the user to define if dividing into chunks when downloading the data on their local machine. Although it does not significantly affect the analysis, it is recommended to keep the default value for optimal performance.\n\nChoose annual or seasonal timeseries\nSelect the family of models\nInterpolation method\nArea to show\nChunks for download\n\n(section-1.3)=", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q03 | Category: Climate_Projections\nSection: Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level > Analysis and results > 1. Parameters, requests and functions definition > 1.2. Define Parameters\n---\nIn the \"Define Parameters\" section, various customisable options for the notebook are specified. Most of the parameters chosen are the same as those used in other assessments ([](./climate_projections-cmip6_climate-impact-indicators_q01)), being them:\n\n- `historical_slice` determines the historical (or control) period used (1971 to 2000 is choosen)\n- `index_timeseries` is a dictionary that set the temporal aggregation for every index considered within this notebook ('HDD15.5' and 'CDD22'). In the presented code, the CDD calculations are always based on the JJA aggregation, with a comfort temperature of 22°C (CDD22), while HDD calculations focus on winter (DJF) with a comfort temperature of 15.5°C (HDD15.5).\n- `collection_id` set the family of models. Only CMIP6 is implemented for this sub-notebook.\n- `area` allows specifying the geographical domain of interest.\n- The `interpolation_method` parameter allows selecting the interpolation method when regridding is performed over the indices.\n- The `chunk` selection allows the user to define if dividing into chunks when downloading the data on their local machine. Although it does not significantly affect the analysis, it is recommended to keep the default value for optimal performance.\n\nChoose annual or seasonal timeseries\nSelect the family of models\nInterpolation method\nArea to show\nChunks for download\n\n(section-1.3)="} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q03__0ad181063cdb", "report_id": "climate_projections-cmip6_climate-impact-indicators_q03", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q03", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level > Analysis and results > 1. Parameters, requests and functions definition > 1.3. Define models", "title": "Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level", "chunk_index": 5, "token_count": 635, "text_raw": "The following climate analyses are performed considering a subset of GCMs from CMIP6. Models names are listed in the parameters below. The selected CMIP6 models have available both the historical and SSP8.5 experiments, and are the same as those used in other assessments ([](./climate_projections-cmip6_climate-impact-indicators_q01)). Additionally, for each model, the thirty-year period corresponding to the Global Warming Level of 2°C above the preindustrial period is specified.\n\nThe Global Warming Level of 2°C can be defined as the earliest point at which a 30-year moving average of global temperature exceeds by two degrees compared to the 1861–1890 baseline [[7]](https://doi.org/10.1088/1748-9326/aab1b1). This thirty-year period varies depending on the Global Climate Model being analysed. To determine this timeframe, the global mean surface temperature for the preindustrial baseline is first calculated for each model. Then, a rolling mean over 30 years is applied to the future period from 2015 to 2099, following the SSP5-8.5 scenario. Finally, the earliest 30-year period at which each model reaches the global warming level of 2°C above the preindustrial period is identified.\n\n\n
\nNOTE on the 30-year slice used for Global Warming Level calculations:
\nGiven the 30-year duration of the historical/control period, the period of mean global surface temperature reaching 2 degrees above preindustrial levels was defined as a 30-year interval (while some studies use 20-year slices for this purpose). The pre-industrial period was also calculated as a 30-year slice (as done in [6]), despite the standard practice of considering the period from 1850 to 1900. This approach not only saves computational time but also ensures consistency in working with 30-year slices.\n\n\n
\nNOTE on the Global Warming Level calculation:
\nTo streamline the current notebook and avoid excessive length and complexity, the calculation of the Global Warming Level of 2°C above the preindustrial period for each model has been conducted externally.\n\nDefine models\nDefine dictionaries to use in titles and caption\n\n(section-1.4)=", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q03 | Category: Climate_Projections\nSection: Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level > Analysis and results > 1. Parameters, requests and functions definition > 1.3. Define models\n---\nThe following climate analyses are performed considering a subset of GCMs from CMIP6. Models names are listed in the parameters below. The selected CMIP6 models have available both the historical and SSP8.5 experiments, and are the same as those used in other assessments ([](./climate_projections-cmip6_climate-impact-indicators_q01)). Additionally, for each model, the thirty-year period corresponding to the Global Warming Level of 2°C above the preindustrial period is specified.\n\nThe Global Warming Level of 2°C can be defined as the earliest point at which a 30-year moving average of global temperature exceeds by two degrees compared to the 1861–1890 baseline [[7]](https://doi.org/10.1088/1748-9326/aab1b1). This thirty-year period varies depending on the Global Climate Model being analysed. To determine this timeframe, the global mean surface temperature for the preindustrial baseline is first calculated for each model. Then, a rolling mean over 30 years is applied to the future period from 2015 to 2099, following the SSP5-8.5 scenario. Finally, the earliest 30-year period at which each model reaches the global warming level of 2°C above the preindustrial period is identified.\n\n\n
\nNOTE on the 30-year slice used for Global Warming Level calculations:
\nGiven the 30-year duration of the historical/control period, the period of mean global surface temperature reaching 2 degrees above preindustrial levels was defined as a 30-year interval (while some studies use 20-year slices for this purpose). The pre-industrial period was also calculated as a 30-year slice (as done in [6]), despite the standard practice of considering the period from 1850 to 1900. This approach not only saves computational time but also ensures consistency in working with 30-year slices.\n\n\n
\nNOTE on the Global Warming Level calculation:
\nTo streamline the current notebook and avoid excessive length and complexity, the calculation of the Global Warming Level of 2°C above the preindustrial period for each model has been conducted externally.\n\nDefine models\nDefine dictionaries to use in titles and caption\n\n(section-1.4)="} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q03__ee3c58eb54a1", "report_id": "climate_projections-cmip6_climate-impact-indicators_q03", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q03", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level > Analysis and results > 1. Parameters, requests and functions definition > 1.4. Define land-sea mask request", "title": "Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level", "chunk_index": 6, "token_count": 154, "text_raw": "Within this sub-notebook, ERA5 will be used to download the land-sea mask when plotting. In this section, we set the required parameters for the cds-api data-request of ERA5 land-sea mask.\n\n(section-1.5)=", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q03 | Category: Climate_Projections\nSection: Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level > Analysis and results > 1. Parameters, requests and functions definition > 1.4. Define land-sea mask request\n---\nWithin this sub-notebook, ERA5 will be used to download the land-sea mask when plotting. In this section, we set the required parameters for the cds-api data-request of ERA5 land-sea mask.\n\n(section-1.5)="} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q03__aa09b0c08e9a", "report_id": "climate_projections-cmip6_climate-impact-indicators_q03", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q03", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level > Analysis and results > 1. Parameters, requests and functions definition > 1.5. Define model requests", "title": "Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level", "chunk_index": 7, "token_count": 206, "text_raw": "In this section we set the required parameters for the cds-api data-request.\n\nWhen `weights = True`, spatial weighting is applied for calculations requiring spatial data aggregation. This is particularly relevant for CMIP6 GCMs with regular lon-lat grids that do not consider varying surface extensions at different latitudes. In contrast, CORDEX RCMs, using rotated grids, inherently account for different cell surfaces based on latitude, eliminating the need for a latitude cosine multiplicative factor (`weights = False`).\n\n(section-1.6)=", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q03 | Category: Climate_Projections\nSection: Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level > Analysis and results > 1. Parameters, requests and functions definition > 1.5. Define model requests\n---\nIn this section we set the required parameters for the cds-api data-request.\n\nWhen `weights = True`, spatial weighting is applied for calculations requiring spatial data aggregation. This is particularly relevant for CMIP6 GCMs with regular lon-lat grids that do not consider varying surface extensions at different latitudes. In contrast, CORDEX RCMs, using rotated grids, inherently account for different cell surfaces based on latitude, eliminating the need for a latitude cosine multiplicative factor (`weights = False`).\n\n(section-1.6)="} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q03__4e1f4aad417f", "report_id": "climate_projections-cmip6_climate-impact-indicators_q03", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q03", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level > Analysis and results > 1. Parameters, requests and functions definition > 1.6. Functions to cache", "title": "Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level", "chunk_index": 8, "token_count": 277, "text_raw": "In this section, functions that will be executed in the caching phase are defined. Caching is the process of storing copies of files in a temporary storage location, so that they can be accessed more quickly. This process also checks if the user has already downloaded a file, avoiding redundant downloads.\n\nFunctions description:\n\n- The `select_timeseries` function subsets the dataset based on the chosen `timeseries` parameter, which could be a specific season (e.g., \"JJA\") or \"annual.\n\n- The `compute_indices` function utilises the icclim package to calculate the energy-consumption-related indices.\n\n- Finally, the `compute_indices_and_trends` function calculates the energy consumption-related indices for the corresponding temporal aggregation using the `compute_indices` function and determines the indices mean over the period of interest.\n\nOriginal bounds for conservative interpolation\n\n(section-2)=", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q03 | Category: Climate_Projections\nSection: Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level > Analysis and results > 1. Parameters, requests and functions definition > 1.6. Functions to cache\n---\nIn this section, functions that will be executed in the caching phase are defined. Caching is the process of storing copies of files in a temporary storage location, so that they can be accessed more quickly. This process also checks if the user has already downloaded a file, avoiding redundant downloads.\n\nFunctions description:\n\n- The `select_timeseries` function subsets the dataset based on the chosen `timeseries` parameter, which could be a specific season (e.g., \"JJA\") or \"annual.\n\n- The `compute_indices` function utilises the icclim package to calculate the energy-consumption-related indices.\n\n- Finally, the `compute_indices_and_trends` function calculates the energy consumption-related indices for the corresponding temporal aggregation using the `compute_indices` function and determines the indices mean over the period of interest.\n\nOriginal bounds for conservative interpolation\n\n(section-2)="} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q03__ec60e0a3b25a", "report_id": "climate_projections-cmip6_climate-impact-indicators_q03", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q03", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level > Analysis and results > 2. Downloading and processing > 2.1. Download and transform the regridding model", "title": "Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level", "chunk_index": 9, "token_count": 384, "text_raw": "In this section, the `download.download_and_transform` function from the 'c3s_eqc_automatic_quality_control' package is employed to download daily data from the selected CMIP6 regridding model, compute the energy-consumption-related indices for the selected temporal aggregation (\"DJF\" for HDD and \"JJA\" for CDD), calculate the mean over the historical period (1971-2000) and for the thirty-year period corresponding to the 2°C warming level, and cache the result (to avoid redundant downloads and processing).\n\nThe regridding model is intended here as the model whose grid will be used to interpolate the others. This ensures all models share a common grid, facilitating the calculation of median values for each cell point. The regridding model within this notebook is \"gfdl_esm4\" but a different one can be selected by just modifying the `model_regrid` parameter at [](section-1.3). It is key to highlight the importance of the chosen target grid depending on the specific application.\n\n\n
\nNOTE on the Global Warming Level calculation:
\nThe thirty year-period corresponding to the 2°C Global Warming Level is different for each model. Its calculation has been conducted externally.\n\n(section-2.2)=", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q03 | Category: Climate_Projections\nSection: Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level > Analysis and results > 2. Downloading and processing > 2.1. Download and transform the regridding model\n---\nIn this section, the `download.download_and_transform` function from the 'c3s_eqc_automatic_quality_control' package is employed to download daily data from the selected CMIP6 regridding model, compute the energy-consumption-related indices for the selected temporal aggregation (\"DJF\" for HDD and \"JJA\" for CDD), calculate the mean over the historical period (1971-2000) and for the thirty-year period corresponding to the 2°C warming level, and cache the result (to avoid redundant downloads and processing).\n\nThe regridding model is intended here as the model whose grid will be used to interpolate the others. This ensures all models share a common grid, facilitating the calculation of median values for each cell point. The regridding model within this notebook is \"gfdl_esm4\" but a different one can be selected by just modifying the `model_regrid` parameter at [](section-1.3). It is key to highlight the importance of the chosen target grid depending on the specific application.\n\n\n
\nNOTE on the Global Warming Level calculation:
\nThe thirty year-period corresponding to the 2°C Global Warming Level is different for each model. Its calculation has been conducted externally.\n\n(section-2.2)="} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q03__404e91b99ea0", "report_id": "climate_projections-cmip6_climate-impact-indicators_q03", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q03", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level > Analysis and results > 2. Downloading and processing > 2.2. Download and transform models", "title": "Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level", "chunk_index": 10, "token_count": 483, "text_raw": "In this section, we utilise the `download_and_transform` function from the 'c3s_eqc_automatic_quality_control' package to download daily data from the CMIP6 models for the historical period (1971-2000) and for the thirty-year period corresponding to the 2°C warming level of each model, compute the energy-consumption-related indices for the selected temporal aggregation (\"DJF\" for HDD and \"JJA\" for CDD), calculate the mean over the historical period (1971-2000) and for the thirty-year period corresponding to the 2°C warming level, interpolate to the regridding model's grid (only for the cases in which it is specified, in the other cases, the original model's grid is mantained), and cache the result (to avoid redundant downloads and processing).\n\n\n
\nNOTE on the Global Warming Level calculation:
\nThe thirty year-period corresponding to the 2°C Global Warming Level is different for each model. Its calculation has been conducted externally.\n\nOriginal model\nInterpolated model\n\n```text\nmodel='access_cm2'\nmodel='awi_cm_1_1_mr'\nmodel='cmcc_esm2'\nmodel='cnrm_cm6_1_hr'\nmodel='cnrm_esm2_1'\nmodel='ec_earth3_cc'\nmodel='gfdl_esm4'\nmodel='inm_cm4_8'\nmodel='inm_cm5_0'\nmodel='kiost_esm'\nmodel='miroc6'\nmodel='miroc_es2l'\nmodel='mpi_esm1_2_lr'\nmodel='mri_esm2_0'\nmodel='noresm2_mm'\nmodel='nesm3'\n```\n\n(section-2.3)=", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q03 | Category: Climate_Projections\nSection: Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level > Analysis and results > 2. Downloading and processing > 2.2. Download and transform models\n---\nIn this section, we utilise the `download_and_transform` function from the 'c3s_eqc_automatic_quality_control' package to download daily data from the CMIP6 models for the historical period (1971-2000) and for the thirty-year period corresponding to the 2°C warming level of each model, compute the energy-consumption-related indices for the selected temporal aggregation (\"DJF\" for HDD and \"JJA\" for CDD), calculate the mean over the historical period (1971-2000) and for the thirty-year period corresponding to the 2°C warming level, interpolate to the regridding model's grid (only for the cases in which it is specified, in the other cases, the original model's grid is mantained), and cache the result (to avoid redundant downloads and processing).\n\n\n
\nNOTE on the Global Warming Level calculation:
\nThe thirty year-period corresponding to the 2°C Global Warming Level is different for each model. Its calculation has been conducted externally.\n\nOriginal model\nInterpolated model\n\n```text\nmodel='access_cm2'\nmodel='awi_cm_1_1_mr'\nmodel='cmcc_esm2'\nmodel='cnrm_cm6_1_hr'\nmodel='cnrm_esm2_1'\nmodel='ec_earth3_cc'\nmodel='gfdl_esm4'\nmodel='inm_cm4_8'\nmodel='inm_cm5_0'\nmodel='kiost_esm'\nmodel='miroc6'\nmodel='miroc_es2l'\nmodel='mpi_esm1_2_lr'\nmodel='mri_esm2_0'\nmodel='noresm2_mm'\nmodel='nesm3'\n```\n\n(section-2.3)="} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q03__010e9ce99bfa", "report_id": "climate_projections-cmip6_climate-impact-indicators_q03", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q03", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level > Analysis and results > 2. Downloading and processing > 2.3. Apply land-sea mask, change attributes and cut the region to show", "title": "Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level", "chunk_index": 11, "token_count": 248, "text_raw": "This section performs the following tasks:\n\n1. Cut the region of interest.\n2. Downloads the sea mask for ERA5.\n3. Regrids ERA5's mask to the `model_regrid` grid and applies it to the regridded data\n4. Regrids the ERA5 land-sea mask to the model's original grid and applies it to them.\n5. Change some variable attributes for plotting purposes.\n\n**Note:** `ds_interpolated` contains data from the models regridded to the regridding model's grid. `model_datasets` contain the same data but in the original grid of each model.\n\nEdit attributes\n\n(section-3)=", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q03 | Category: Climate_Projections\nSection: Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level > Analysis and results > 2. Downloading and processing > 2.3. Apply land-sea mask, change attributes and cut the region to show\n---\nThis section performs the following tasks:\n\n1. Cut the region of interest.\n2. Downloads the sea mask for ERA5.\n3. Regrids ERA5's mask to the `model_regrid` grid and applies it to the regridded data\n4. Regrids the ERA5 land-sea mask to the model's original grid and applies it to them.\n5. Change some variable attributes for plotting purposes.\n\n**Note:** `ds_interpolated` contains data from the models regridded to the regridding model's grid. `model_datasets` contain the same data but in the original grid of each model.\n\nEdit attributes\n\n(section-3)="} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q03__62b01b8a72e3", "report_id": "climate_projections-cmip6_climate-impact-indicators_q03", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q03", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level > Analysis and results > 3. Plot and describe results", "title": "Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level", "chunk_index": 12, "token_count": 233, "text_raw": "This section will display the following results:\n\n- Maps representing the spatial distribution of the climate change signal (i.e., mean values for the warming level of 2°C compared to the historical period 1971-2000) of the indices 'HDD15.5' and 'CDD22' for each model individually, the ensemble median (understood as the median of the climate signal values of the chosen subset of models calculated for each grid cell), and the ensemble spread (derived as the standard deviation of the distribution of the chosen subset of models).\n\n- Boxplots representing statistical distributions (PDFs) constructed from the spatially-averaged climate signal of each considered model.\n\n(section-3.1)=", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q03 | Category: Climate_Projections\nSection: Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level > Analysis and results > 3. Plot and describe results\n---\nThis section will display the following results:\n\n- Maps representing the spatial distribution of the climate change signal (i.e., mean values for the warming level of 2°C compared to the historical period 1971-2000) of the indices 'HDD15.5' and 'CDD22' for each model individually, the ensemble median (understood as the median of the climate signal values of the chosen subset of models calculated for each grid cell), and the ensemble spread (derived as the standard deviation of the distribution of the chosen subset of models).\n\n- Boxplots representing statistical distributions (PDFs) constructed from the spatially-averaged climate signal of each considered model.\n\n(section-3.1)="} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q03__e459460b84df", "report_id": "climate_projections-cmip6_climate-impact-indicators_q03", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q03", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level > Analysis and results > 3. Plot and describe results > 3.1. Define plotting functions", "title": "Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level", "chunk_index": 13, "token_count": 418, "text_raw": "The functions presented here are used to plot the climate signal (i.e., mean values for the warming level of 2°C compared to the historical period) for each of the indices ('HDD15.5' and 'CDD22').\n\nFor a selected index, two layout types will be displayed, depending on the chosen function:\n\n1. Layout including the ensemble median and the ensemble spread: `plot_ensemble()` is used.\n2. Layout including every model: `plot_models()` is employed.\n\nDefine function to plot the caption of the figures (for the ensemble case)\nAdd caption to the figure\nAdd each line to the figure\nend captioning\nDefine function to plot the cation of the figures (for the individual models case)\nAdd caption to the figure\nAdd each line to the figure\n\n\n
\nCOLORBAR NOTE:
\nThe colorbar chosen to represent the climate signal for the HDD15.5 index spans from intense to light red. In this color scheme, large negative climate signal values (shown in more intense reds) indicate a significant decrease in Heating Degree Days compared to the historical period. Conversely, lighter red colors indicate a less pronounced decrease in Heating Degree Days compared to the historical period values. This selection is based on the rationale that more Heating Degree Days are associated with colder conditions, typically represented by blueish colors, while fewer Heating Degree Days are associated with warmer conditions, depicted by reddish colors.\n\n(section-3.2)=", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q03 | Category: Climate_Projections\nSection: Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level > Analysis and results > 3. Plot and describe results > 3.1. Define plotting functions\n---\nThe functions presented here are used to plot the climate signal (i.e., mean values for the warming level of 2°C compared to the historical period) for each of the indices ('HDD15.5' and 'CDD22').\n\nFor a selected index, two layout types will be displayed, depending on the chosen function:\n\n1. Layout including the ensemble median and the ensemble spread: `plot_ensemble()` is used.\n2. Layout including every model: `plot_models()` is employed.\n\nDefine function to plot the caption of the figures (for the ensemble case)\nAdd caption to the figure\nAdd each line to the figure\nend captioning\nDefine function to plot the cation of the figures (for the individual models case)\nAdd caption to the figure\nAdd each line to the figure\n\n\n
\nCOLORBAR NOTE:
\nThe colorbar chosen to represent the climate signal for the HDD15.5 index spans from intense to light red. In this color scheme, large negative climate signal values (shown in more intense reds) indicate a significant decrease in Heating Degree Days compared to the historical period. Conversely, lighter red colors indicate a less pronounced decrease in Heating Degree Days compared to the historical period values. This selection is based on the rationale that more Heating Degree Days are associated with colder conditions, typically represented by blueish colors, while fewer Heating Degree Days are associated with warmer conditions, depicted by reddish colors.\n\n(section-3.2)="} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q03__67be5683977b", "report_id": "climate_projections-cmip6_climate-impact-indicators_q03", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q03", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level > Analysis and results > 3. Plot and describe results > 3.2. Plot ensemble maps", "title": "Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level", "chunk_index": 14, "token_count": 264, "text_raw": "In this section, we invoke the `plot_ensemble()` function to visualise the climate signal (i.e., mean values for the warming level of 2°C compared to the historical period) for each of the indices ('HDD15.5' and 'CDD22') across Europe. The layout includes: (a) the ensemble median (understood as the median of the climate signal values of the chosen subset of models calculated for each grid cell) and (b) the ensemble spread (derived as the standard deviation of the distribution of the chosen subset of models).\n\nNote that the model data used in this section has previously been interpolated to the \"regridding model\" grid (`\"gfdl_esm4\"` for this notebook).\n\nFig number counter\nCommon title\n\n(section-3.3)=", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q03 | Category: Climate_Projections\nSection: Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level > Analysis and results > 3. Plot and describe results > 3.2. Plot ensemble maps\n---\nIn this section, we invoke the `plot_ensemble()` function to visualise the climate signal (i.e., mean values for the warming level of 2°C compared to the historical period) for each of the indices ('HDD15.5' and 'CDD22') across Europe. The layout includes: (a) the ensemble median (understood as the median of the climate signal values of the chosen subset of models calculated for each grid cell) and (b) the ensemble spread (derived as the standard deviation of the distribution of the chosen subset of models).\n\nNote that the model data used in this section has previously been interpolated to the \"regridding model\" grid (`\"gfdl_esm4\"` for this notebook).\n\nFig number counter\nCommon title\n\n(section-3.3)="} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q03__45795be49295", "report_id": "climate_projections-cmip6_climate-impact-indicators_q03", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q03", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level > Analysis and results > 3. Plot and describe results > 3.3. Plot model maps", "title": "Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level", "chunk_index": 15, "token_count": 178, "text_raw": "In this section, we invoke the `plot_models()` function to visualise the climate signal for every model individually across Europe. Note that the model data used in this section maintains its original grid.\n\nSpecifically, for each of the indices ('HDD15.5' and CDD22'), this section presents a single layout including the climate signal of every model.\n\n(section-3.4)=", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q03 | Category: Climate_Projections\nSection: Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level > Analysis and results > 3. Plot and describe results > 3.3. Plot model maps\n---\nIn this section, we invoke the `plot_models()` function to visualise the climate signal for every model individually across Europe. Note that the model data used in this section maintains its original grid.\n\nSpecifically, for each of the indices ('HDD15.5' and CDD22'), this section presents a single layout including the climate signal of every model.\n\n(section-3.4)="} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q03__8c0e6f3f98a1", "report_id": "climate_projections-cmip6_climate-impact-indicators_q03", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q03", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level > Analysis and results > 3. Plot and describe results > 3.4. Boxplots of the climate change signal", "title": "Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level", "chunk_index": 16, "token_count": 427, "text_raw": "Finally, we present boxplots representing the ensemble distribution of each climate model signal for the 2°C global warming level.\n\nDots represent the spatially-averaged climate signal over the selected region for each model (grey) and the ensemble mean (blue). The ensemble median is shown as a green line. Note that the spatially averaged values are calculated for each model from its original grid (i.e., no interpolated data has been used here).\n\nThe boxplot visually illustrates the distribution of the climate signal among the climate models, with the box covering the first quartile (Q1 = 25th percentile) to the third quartile (Q3 = 75th percentile), and a green line indicating the ensemble median (Q2 = 50th percentile). Whiskers extend from the edges of the box to show the full data range.\n\nEnsemble mean\n\n
\n
\n

Fig 5. Boxplots illustrating the climate signal (i.e., mean values for the warming level of 2°C compared to the historical period from 1971 to 2000) of the distribution of the chosen subset of models for the Energy Degree Days indices daily averaged: (a) 'CDD22', and (b) 'HDD15.5'. The distribution is created by considering spatially averaged values across Europe. The ensemble mean and the ensemble median are both included. Outliers in the distribution are denoted by a grey circle with a black contour.

\n

\n\n(section-3.5)=", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q03 | Category: Climate_Projections\nSection: Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level > Analysis and results > 3. Plot and describe results > 3.4. Boxplots of the climate change signal\n---\nFinally, we present boxplots representing the ensemble distribution of each climate model signal for the 2°C global warming level.\n\nDots represent the spatially-averaged climate signal over the selected region for each model (grey) and the ensemble mean (blue). The ensemble median is shown as a green line. Note that the spatially averaged values are calculated for each model from its original grid (i.e., no interpolated data has been used here).\n\nThe boxplot visually illustrates the distribution of the climate signal among the climate models, with the box covering the first quartile (Q1 = 25th percentile) to the third quartile (Q3 = 75th percentile), and a green line indicating the ensemble median (Q2 = 50th percentile). Whiskers extend from the edges of the box to show the full data range.\n\nEnsemble mean\n\n
\n
\n

Fig 5. Boxplots illustrating the climate signal (i.e., mean values for the warming level of 2°C compared to the historical period from 1971 to 2000) of the distribution of the chosen subset of models for the Energy Degree Days indices daily averaged: (a) 'CDD22', and (b) 'HDD15.5'. The distribution is created by considering spatially averaged values across Europe. The ensemble mean and the ensemble median are both included. Outliers in the distribution are denoted by a grey circle with a black contour.

\n

\n\n(section-3.5)="} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q03__76c7f7c99bb2", "report_id": "climate_projections-cmip6_climate-impact-indicators_q03", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q03", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level > Analysis and results > 3. Plot and describe results > 3.5. Results summary and discussion", "title": "Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level", "chunk_index": 17, "token_count": 712, "text_raw": "- The Heating Degree Days (HDD15.5) during the DJF season (i.e., the sum of the differences between the thermal comfort temperature of 15.5°C and the daily mean temperature, only when the mean temperature is below the thermal comfort temperature) are projected to generally decrease in a world 2°C warmer than the preindustrial baseline (1861-1890), compared to the average daily values typical of the control period (1971-2000) across Europe. The largest decrease is expected in the northern parts of Europe and continental Europe.\n\n- The Cooling Degree Days (CDD22) during the JJA season (i.e., the sum of the differences between the daily mean temperature and the thermal comfort temperature of 22°C, only when the mean temperature is above the thermal comfort temperature) are projected to generally increase in a world 2°C warmer than the preindustrial baseline (1861-1890), compared to the average daily values typical of the control period (1971-2000) across Europe. The largest increase is expected in the Mediterranean Basin, while northern Europe will experience a lower increase, possibly because exceeding a mean temperature of 22°C in those regions may not be too frequent even in a world 2°C warmer than the preindustrial era.\n\n- Boxplots indicate that, across the considered region, the Heating Degree Days for winter (HDD15.5) will have daily average values (if the ensemble median is considered) 2.2°C lower compared to the historical period's DJF daily average values. If the focus is on the Cooling Degree Days in JJA (CDD22), an increase can be observed, reaching a daily average value for the ensemble median which is almost 1.5°C higher than that calculated for the historical period.\n\n- What do the results mean for users? Are the biases relevant?\n - The projected decrease in Heating Degree Days during winter and the projected increase in Cooling Degree Days during summer in a world 2°C warmer than the preindustrial baseline provide useful information for decisions related to future energy demand. Utilising climate global warming levels rather than fixed future periods helps reduce the systematic biases present in models [[8]](https://doi.org/10.5194/esd-13-321-2022). However, it is important to recognise that the centered thirty-year period when the global mean temperature reaches 2°C above the preindustrial baseline varies depending on the model used, making this timing difficult to determine. Users need to determine whether they are more interested in projected values for specific periods or in working with global warming levels. A combined approach could yield a more comprehensive assessment.\n\n\n
\nRESULTS NOTE:
\nIt is important to note that the results presented are specific to the 16 models chosen, and users should aim to assess as wide a range of models as possible before making a sub-selection.", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q03 | Category: Climate_Projections\nSection: Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level > Analysis and results > 3. Plot and describe results > 3.5. Results summary and discussion\n---\n- The Heating Degree Days (HDD15.5) during the DJF season (i.e., the sum of the differences between the thermal comfort temperature of 15.5°C and the daily mean temperature, only when the mean temperature is below the thermal comfort temperature) are projected to generally decrease in a world 2°C warmer than the preindustrial baseline (1861-1890), compared to the average daily values typical of the control period (1971-2000) across Europe. The largest decrease is expected in the northern parts of Europe and continental Europe.\n\n- The Cooling Degree Days (CDD22) during the JJA season (i.e., the sum of the differences between the daily mean temperature and the thermal comfort temperature of 22°C, only when the mean temperature is above the thermal comfort temperature) are projected to generally increase in a world 2°C warmer than the preindustrial baseline (1861-1890), compared to the average daily values typical of the control period (1971-2000) across Europe. The largest increase is expected in the Mediterranean Basin, while northern Europe will experience a lower increase, possibly because exceeding a mean temperature of 22°C in those regions may not be too frequent even in a world 2°C warmer than the preindustrial era.\n\n- Boxplots indicate that, across the considered region, the Heating Degree Days for winter (HDD15.5) will have daily average values (if the ensemble median is considered) 2.2°C lower compared to the historical period's DJF daily average values. If the focus is on the Cooling Degree Days in JJA (CDD22), an increase can be observed, reaching a daily average value for the ensemble median which is almost 1.5°C higher than that calculated for the historical period.\n\n- What do the results mean for users? Are the biases relevant?\n - The projected decrease in Heating Degree Days during winter and the projected increase in Cooling Degree Days during summer in a world 2°C warmer than the preindustrial baseline provide useful information for decisions related to future energy demand. Utilising climate global warming levels rather than fixed future periods helps reduce the systematic biases present in models [[8]](https://doi.org/10.5194/esd-13-321-2022). However, it is important to recognise that the centered thirty-year period when the global mean temperature reaches 2°C above the preindustrial baseline varies depending on the model used, making this timing difficult to determine. Users need to determine whether they are more interested in projected values for specific periods or in working with global warming levels. A combined approach could yield a more comprehensive assessment.\n\n\n
\nRESULTS NOTE:
\nIt is important to note that the results presented are specific to the 16 models chosen, and users should aim to assess as wide a range of models as possible before making a sub-selection."} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q03__5374652e3485", "report_id": "climate_projections-cmip6_climate-impact-indicators_q03", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q03", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level > ℹ️ If you want to know more > Key resources", "title": "Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level", "chunk_index": 18, "token_count": 237, "text_raw": "Some key resources and further reading were linked throughout this assessment.\n\nThe CDS catalogue entries for the data used were:\n* CMIP6 climate projections (Daily - air temperature): https://cds.climate.copernicus.eu/datasets/projections-cmip6?tab=overview\n\nCode libraries used:\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)\n* [icclim](https://icclim.readthedocs.io/en/stable/) Python package", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q03 | Category: Climate_Projections\nSection: Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level > ℹ️ If you want to know more > Key resources\n---\nSome key resources and further reading were linked throughout this assessment.\n\nThe CDS catalogue entries for the data used were:\n* CMIP6 climate projections (Daily - air temperature): https://cds.climate.copernicus.eu/datasets/projections-cmip6?tab=overview\n\nCode libraries used:\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)\n* [icclim](https://icclim.readthedocs.io/en/stable/) Python package"} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q03__f6f6b3baf731", "report_id": "climate_projections-cmip6_climate-impact-indicators_q03", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q03", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level > ℹ️ If you want to know more > References", "title": "Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level", "chunk_index": 19, "token_count": 1066, "text_raw": "[[1]](https://doi.org/10.1007/s10113-010-0173-x) Bindi, M., Olesen, J.E. The responses of agriculture in Europe to climate change (2011). Reg Environ Change 11 (Suppl 1), 151–158. https://doi.org/10.1007/s10113-010-0173-x\n\n[[2]](https://doi.org/10.1016/j.foreco.2009.09.023) Lindner, M., Maroschek, M., Netherer, S., Kremer, A., Barbati, A., Garcia-Gonzalo, J., Seidi, R., Delzon, S., Corona, P., Kolstrom, M., Lexer, M.J., Marchetti, M. (2010). Climate change impacts, adaptive capacity, and vulnerability of European forest ecosystems. For. Ecol. Manage. 259(4): 698–709. https://doi.org/10.1016/j.foreco.2009.09.023\n\n[[3]](https://doi.org/10.1016/j.enbuild.2014.09.052) Santamouris, M., Cartalis, C., Synnefa, A., Kolokotsa, D. (2015). On the impact of urban heat island and global warming on the power demand and electricity consumption of buildings – a review. Energy Build. 98: 119–124. https://doi.org/10.1016/j.enbuild.2014.09.052\n\n[[4]](https://doi.org/10.1007/s10113-013-0499-2) Jacob, D., Petersen, J., Eggert, B. et al. (2014). EURO-CORDEX: new high-resolution climate change projections for European impact research. Reg Environ Change 14, 563–578. https://doi.org/10.1007/s10113-013-0499-2\n\n[[5]](http://hdl.handle.net/10013/epic.45156.d001) IPCC. 2014. In Climate Change 2014: Synthesis Report. Contribution of Working Groups I, II and III to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change, Core Writing Team, RK Pachauri, LA Meyer (eds). IPCC: Geneva, Switzerland 151 pp.\n\n[[6]](https://doi.org/10.1002/joc.5362) Spinoni, J., Vogt, J.V., Barbosa, P., Dosio, A., McCormick, N., Bigano, A. and Füssel, H.-M. (2018). Changes of heating and cooling degree-days in Europe from 1981 to 2100. Int. J. Climatol, 38: e191-e208. https://doi.org/10.1002/joc.5362\n\n[[7]](https://doi.org/10.1088/1748-9326/aab1b1) Nikulin, G., Lennard, C., Dosio, A., Kjellström, E., Chen, Y., Hänsler, A., Kupiainen, M., Laprise, R., Laura Mariotti, L., Maule C.F., et al. (2018). The effects of 1.5 and 2 degrees of global warming on Africa in the CORDEX ensemble. Environ. Res. Lett. 13 065003. https://doi.org/10.1088/1748-9326/aab1b1\n\n[[8]](https://doi.org/10.5194/esd-13-321-2022) Cos, J., Doblas-Reyes, F., Jury, M., Marcos, R., Bretonnière, P.-A., and Samsó, M. (2022). The Mediterranean climate change hotspot in the CMIP5 and CMIP6 projections, Earth Syst. Dynam., 13, 321–340. https://doi.org/10.5194/esd-13-321-2022\n\n[[9]](https://doi.org/10.1038/s43247-023-00878-3) Scoccimarro, E., Cattaneo, O., Gualdi, S. et al. (2023). Country-level energy demand for cooling has increased over the past two decades. Commun Earth Environ 4, 208. https://doi.org/10.1038/s43247-023-00878-3", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q03 | Category: Climate_Projections\nSection: Uncertainty in projected changes of energy consumption in Europe at a 2°C Global Warming Level > ℹ️ If you want to know more > References\n---\n[[1]](https://doi.org/10.1007/s10113-010-0173-x) Bindi, M., Olesen, J.E. The responses of agriculture in Europe to climate change (2011). Reg Environ Change 11 (Suppl 1), 151–158. https://doi.org/10.1007/s10113-010-0173-x\n\n[[2]](https://doi.org/10.1016/j.foreco.2009.09.023) Lindner, M., Maroschek, M., Netherer, S., Kremer, A., Barbati, A., Garcia-Gonzalo, J., Seidi, R., Delzon, S., Corona, P., Kolstrom, M., Lexer, M.J., Marchetti, M. (2010). Climate change impacts, adaptive capacity, and vulnerability of European forest ecosystems. For. Ecol. Manage. 259(4): 698–709. https://doi.org/10.1016/j.foreco.2009.09.023\n\n[[3]](https://doi.org/10.1016/j.enbuild.2014.09.052) Santamouris, M., Cartalis, C., Synnefa, A., Kolokotsa, D. (2015). On the impact of urban heat island and global warming on the power demand and electricity consumption of buildings – a review. Energy Build. 98: 119–124. https://doi.org/10.1016/j.enbuild.2014.09.052\n\n[[4]](https://doi.org/10.1007/s10113-013-0499-2) Jacob, D., Petersen, J., Eggert, B. et al. (2014). EURO-CORDEX: new high-resolution climate change projections for European impact research. Reg Environ Change 14, 563–578. https://doi.org/10.1007/s10113-013-0499-2\n\n[[5]](http://hdl.handle.net/10013/epic.45156.d001) IPCC. 2014. In Climate Change 2014: Synthesis Report. Contribution of Working Groups I, II and III to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change, Core Writing Team, RK Pachauri, LA Meyer (eds). IPCC: Geneva, Switzerland 151 pp.\n\n[[6]](https://doi.org/10.1002/joc.5362) Spinoni, J., Vogt, J.V., Barbosa, P., Dosio, A., McCormick, N., Bigano, A. and Füssel, H.-M. (2018). Changes of heating and cooling degree-days in Europe from 1981 to 2100. Int. J. Climatol, 38: e191-e208. https://doi.org/10.1002/joc.5362\n\n[[7]](https://doi.org/10.1088/1748-9326/aab1b1) Nikulin, G., Lennard, C., Dosio, A., Kjellström, E., Chen, Y., Hänsler, A., Kupiainen, M., Laprise, R., Laura Mariotti, L., Maule C.F., et al. (2018). The effects of 1.5 and 2 degrees of global warming on Africa in the CORDEX ensemble. Environ. Res. Lett. 13 065003. https://doi.org/10.1088/1748-9326/aab1b1\n\n[[8]](https://doi.org/10.5194/esd-13-321-2022) Cos, J., Doblas-Reyes, F., Jury, M., Marcos, R., Bretonnière, P.-A., and Samsó, M. (2022). The Mediterranean climate change hotspot in the CMIP5 and CMIP6 projections, Earth Syst. Dynam., 13, 321–340. https://doi.org/10.5194/esd-13-321-2022\n\n[[9]](https://doi.org/10.1038/s43247-023-00878-3) Scoccimarro, E., Cattaneo, O., Gualdi, S. et al. (2023). Country-level energy demand for cooling has increased over the past two decades. Commun Earth Environ 4, 208. https://doi.org/10.1038/s43247-023-00878-3"} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q04__3d3e6401958f", "report_id": "climate_projections-cmip6_climate-impact-indicators_q04", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q04", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Projections of future ice-free periods for the Arctic and Antarctic", "title": "Projections of future ice-free periods for the Arctic and Antarctic", "chunk_index": 0, "token_count": 94, "text_raw": "Production date: 2025-04-23\n\nProduced by: Timothy Williams, Nansen Environmental and Remote Sensing Center (NERSC)", "text_with_prefix": "EQC Quality Assessment: \"Projections of future ice-free periods for the Arctic and Antarctic\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q04 | Category: Climate_Projections\nSection: Projections of future ice-free periods for the Arctic and Antarctic\n---\nProduction date: 2025-04-23\n\nProduced by: Timothy Williams, Nansen Environmental and Remote Sensing Center (NERSC)"} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q04__f1d45fb7a8ff", "report_id": "climate_projections-cmip6_climate-impact-indicators_q04", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q04", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Projections of future ice-free periods for the Arctic and Antarctic > Quality assessment question", "title": "Projections of future ice-free periods for the Arctic and Antarctic", "chunk_index": 1, "token_count": 103, "text_raw": "**How likely are sea-ice-free conditions in the Arctic and Antarctic, and along the Arctic shipping routes, and when under which warming scenarios would they become more likely?**", "text_with_prefix": "EQC Quality Assessment: \"Projections of future ice-free periods for the Arctic and Antarctic\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q04 | Category: Climate_Projections\nSection: Projections of future ice-free periods for the Arctic and Antarctic > Quality assessment question\n---\n**How likely are sea-ice-free conditions in the Arctic and Antarctic, and along the Arctic shipping routes, and when under which warming scenarios would they become more likely?**"} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q04__1c6c04bbaed9", "report_id": "climate_projections-cmip6_climate-impact-indicators_q04", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q04", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Projections of future ice-free periods for the Arctic and Antarctic > Quality assessment statement", "title": "Projections of future ice-free periods for the Arctic and Antarctic", "chunk_index": 2, "token_count": 689, "text_raw": "These are the key outcomes of this assessment\n\n- Depending on the warming scenario, the Arctic itself has varying chances of being ice-free in the summer. While some authors project the first ice-free year as being before 2050 (regardless of warming scenario) (eg. [SIMIP community, 2020](https://doi.org/10.1029/2019GL086749); [X. Zhao et al., 2022](https://doi.org/10.1029/2022EF002708)), we find (considering the full set of models proving sea ice concentration) more dependence on the warming scenario, with only 40% of models projecting an ice-free Arctic in 2050 under `ssp1_2_6` while 60% of them project it under `ssp5_8_5`. In 2070 this range has increased to between 50% and 85%. The period with high probabilities also lengthens according to the warming scenario, from just September in `ssp1_2_6` to the 6-month period from June to December in `ssp5_8_5`.\n\n- The maximum Arctic area is projected to drop by between $4\\times 10^6$ km$^2$ and $8\\times 10^6$ km$^2$ from the pre-1960 levels of about $17\\times 10^6$ km$^2$.\n\n- Around Antarctica, the climate models predict too little ice by quite a large margin, as seen by [Roach et al. (2023)](https://doi.org/10.1029/2019GL086729) and the [CDS assessment of historical sea ice extent](https://ecmwf-projects.github.io/c3s2-eqc-quality-assessment/Climate_Projections/CMIP6/climate_projections-cmip6_model-performance_q02.html), and this limits the conclusions that can be made about accessibility in this region. However with a definition of ice-free conditions consistent with the Arctic, we also see a similar pattern of increased overall probability and length of the period of high probabilities with increased warming.\n\n- We also see this behaviour for the two Arctic shipping routes, the Northern Sea Route (NSR) and the Transpolar Sea Route (TSR). There is higher probability of the TSR being ice-free in December than for the NSR. This is consistent with the results of other authors ([Chen et al, 2023](https://doi.org/10.1016/j.accre.2023.11.011); [Min et al., 2022](https://doi.org/10.1029/2022GL099157)).\n```", "text_with_prefix": "EQC Quality Assessment: \"Projections of future ice-free periods for the Arctic and Antarctic\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q04 | Category: Climate_Projections\nSection: Projections of future ice-free periods for the Arctic and Antarctic > Quality assessment statement\n---\nThese are the key outcomes of this assessment\n\n- Depending on the warming scenario, the Arctic itself has varying chances of being ice-free in the summer. While some authors project the first ice-free year as being before 2050 (regardless of warming scenario) (eg. [SIMIP community, 2020](https://doi.org/10.1029/2019GL086749); [X. Zhao et al., 2022](https://doi.org/10.1029/2022EF002708)), we find (considering the full set of models proving sea ice concentration) more dependence on the warming scenario, with only 40% of models projecting an ice-free Arctic in 2050 under `ssp1_2_6` while 60% of them project it under `ssp5_8_5`. In 2070 this range has increased to between 50% and 85%. The period with high probabilities also lengthens according to the warming scenario, from just September in `ssp1_2_6` to the 6-month period from June to December in `ssp5_8_5`.\n\n- The maximum Arctic area is projected to drop by between $4\\times 10^6$ km$^2$ and $8\\times 10^6$ km$^2$ from the pre-1960 levels of about $17\\times 10^6$ km$^2$.\n\n- Around Antarctica, the climate models predict too little ice by quite a large margin, as seen by [Roach et al. (2023)](https://doi.org/10.1029/2019GL086729) and the [CDS assessment of historical sea ice extent](https://ecmwf-projects.github.io/c3s2-eqc-quality-assessment/Climate_Projections/CMIP6/climate_projections-cmip6_model-performance_q02.html), and this limits the conclusions that can be made about accessibility in this region. However with a definition of ice-free conditions consistent with the Arctic, we also see a similar pattern of increased overall probability and length of the period of high probabilities with increased warming.\n\n- We also see this behaviour for the two Arctic shipping routes, the Northern Sea Route (NSR) and the Transpolar Sea Route (TSR). There is higher probability of the TSR being ice-free in December than for the NSR. This is consistent with the results of other authors ([Chen et al, 2023](https://doi.org/10.1016/j.accre.2023.11.011); [Min et al., 2022](https://doi.org/10.1029/2022GL099157)).\n```"} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q04__81dea7f44515", "report_id": "climate_projections-cmip6_climate-impact-indicators_q04", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q04", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Projections of future ice-free periods for the Arctic and Antarctic > Methodology", "title": "Projections of future ice-free periods for the Arctic and Antarctic", "chunk_index": 3, "token_count": 956, "text_raw": "Related to this assessment are the assessments of [sea ice extent](https://ecmwf-projects.github.io/c3s2-eqc-quality-assessment/Climate_Projections/CMIP6/climate_projections-cmip6_model-performance_q02.html) and [Arctic sea ice thickness](https://ecmwf-projects.github.io/c3s2-eqc-quality-assessment/Climate_Projections/CMIP6/climate_projections-cmip6_model-performance_q03.html) in the CMIP6 historical experiment. The sea ice thickness bias is not directly relevant to this one, but it could have been used in subsetting the models.\n\nFor each experiment, we take all the models that provide sea ice concentration (SIC) outputs for our ensemble - i.e. we don't do any subsetting of models in the results shown. We did do some sensitivity testing for the accessibility results however, and found no change to these results when using different model subsets. Specifically, noting that [Pan et al. (2023)](https://doi.org/10.1029/2022GL102077) found that climate models whose ocean model component was NEMO (Nucleus for European Modelling of the Ocean) had noticeably reduced Arctic sea ice in projections compared to other climate models, we tried either excluding or only using such models. Users should still bear in mind that using a different subset could still effect results. For example, the [SIMIP community (2020)](https://doi.org/10.1029/2019GL086749) selected models whose ensemble spread of sea ice area in the historical experiment included the satellite estimate. Alternatively, subsetting with \"emergent constraints\" (physically explainable empirical relationships) has been used to reduce the uncertainty in predictions. In examples of this approach, [X. Zhao et al. (2022)](https://doi.org/10.1029/2022EF002708) used the accuracy of the mean April SIT and the response of the minimum sea ice area to the mean April SIT (compared to the value from Pan-Arctic Ice-Ocean Assimlation System, or PIOMAS) to select models, while [Wang et al. (2021)](https://doi.org/10.1088/1748-9326/ac0b17) used the accuracy of the surface air temperature and the sensitivity of the ice area to the surface air temperature. Another approach is to use weighted statistics (as done by [J. Zhou et al., 2022](https://doi.org/10.1029/2022EF002708)), where weights are determined by model skill and interdependance.\n\nWe also do not consider the individual ensemble members available for each model, as the [SIMIP community (2020)](https://doi.org/10.1029/2019GL086749) did, but only the ensemble mean for each model. This is done to save computational cost, but also avoids biasing the mean towards models with larger ensembles. Also note that we only use the monthly mean - considering daily means as done by [Heuze & Jahn (2024)](https://doi.org/10.1038/s41467-024-54508-3) can lead to projected estimates of a much earlier first ice-free year (before 2030) than estimated using monthly means.\n\nAs well as the historical experiment, which has 33 models that output SIC, we use the following projection experiments, chosen because they provide a sufficiently large ensemble:\n- `ssp1_2_6` (22 models outputting SIC)\n- `ssp2_4_5` (22 models outputting SIC)\n- `ssp3_7_0` (21 models outputting SIC)\n- `ssp5_8_5` (21 models outputting SIC)", "text_with_prefix": "EQC Quality Assessment: \"Projections of future ice-free periods for the Arctic and Antarctic\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q04 | Category: Climate_Projections\nSection: Projections of future ice-free periods for the Arctic and Antarctic > Methodology\n---\nRelated to this assessment are the assessments of [sea ice extent](https://ecmwf-projects.github.io/c3s2-eqc-quality-assessment/Climate_Projections/CMIP6/climate_projections-cmip6_model-performance_q02.html) and [Arctic sea ice thickness](https://ecmwf-projects.github.io/c3s2-eqc-quality-assessment/Climate_Projections/CMIP6/climate_projections-cmip6_model-performance_q03.html) in the CMIP6 historical experiment. The sea ice thickness bias is not directly relevant to this one, but it could have been used in subsetting the models.\n\nFor each experiment, we take all the models that provide sea ice concentration (SIC) outputs for our ensemble - i.e. we don't do any subsetting of models in the results shown. We did do some sensitivity testing for the accessibility results however, and found no change to these results when using different model subsets. Specifically, noting that [Pan et al. (2023)](https://doi.org/10.1029/2022GL102077) found that climate models whose ocean model component was NEMO (Nucleus for European Modelling of the Ocean) had noticeably reduced Arctic sea ice in projections compared to other climate models, we tried either excluding or only using such models. Users should still bear in mind that using a different subset could still effect results. For example, the [SIMIP community (2020)](https://doi.org/10.1029/2019GL086749) selected models whose ensemble spread of sea ice area in the historical experiment included the satellite estimate. Alternatively, subsetting with \"emergent constraints\" (physically explainable empirical relationships) has been used to reduce the uncertainty in predictions. In examples of this approach, [X. Zhao et al. (2022)](https://doi.org/10.1029/2022EF002708) used the accuracy of the mean April SIT and the response of the minimum sea ice area to the mean April SIT (compared to the value from Pan-Arctic Ice-Ocean Assimlation System, or PIOMAS) to select models, while [Wang et al. (2021)](https://doi.org/10.1088/1748-9326/ac0b17) used the accuracy of the surface air temperature and the sensitivity of the ice area to the surface air temperature. Another approach is to use weighted statistics (as done by [J. Zhou et al., 2022](https://doi.org/10.1029/2022EF002708)), where weights are determined by model skill and interdependance.\n\nWe also do not consider the individual ensemble members available for each model, as the [SIMIP community (2020)](https://doi.org/10.1029/2019GL086749) did, but only the ensemble mean for each model. This is done to save computational cost, but also avoids biasing the mean towards models with larger ensembles. Also note that we only use the monthly mean - considering daily means as done by [Heuze & Jahn (2024)](https://doi.org/10.1038/s41467-024-54508-3) can lead to projected estimates of a much earlier first ice-free year (before 2030) than estimated using monthly means.\n\nAs well as the historical experiment, which has 33 models that output SIC, we use the following projection experiments, chosen because they provide a sufficiently large ensemble:\n- `ssp1_2_6` (22 models outputting SIC)\n- `ssp2_4_5` (22 models outputting SIC)\n- `ssp3_7_0` (21 models outputting SIC)\n- `ssp5_8_5` (21 models outputting SIC)"} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q04__fdaf2f49bd2f", "report_id": "climate_projections-cmip6_climate-impact-indicators_q04", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q04", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Projections of future ice-free periods for the Arctic and Antarctic > Methodology", "title": "Projections of future ice-free periods for the Arctic and Antarctic", "chunk_index": 4, "token_count": 936, "text_raw": "5` (22 models outputting SIC)\n- `ssp3_7_0` (21 models outputting SIC)\n- `ssp5_8_5` (21 models outputting SIC)\n\nThe experiments are ordered by expected warming with `ssp_1_2_6` giving the least warming (reaching 2.6 W/m$^2$ in 2100, which is a kind of \"best case\" warming scenario) and `ssp5_8_5` giving high warming (8.5 W/m$^2$ in 2100).\nThe first number refers to which of the five [Shared Socioeconomic Pathways](https://www.dkrz.de/en/communication/climate-simulations/cmip6-en/the-ssp-scenarios) (SSP) that the experiment corresponds to.\n\nWe calculate the minimum and maximum sea ice areas for the Arctic and Antarctica, show the historical and projected results as well as results derived from satellite observations, and show some climatological maps to show projected changes in the spatial distribution of SIC.\nTo calculate the area we remap the model SIC onto the equal-area grids used by the observations. The ice-covered area in each grid cell can be determined from the SIC and the grid cell area, and then summed up to get the total ice-covered area.\nFor each experiment, we then have an ensemble of areal time series (one from each model) and we plot the median and IQL (inter-quartile limits) of this ensemble.\nFor the climatological maps we follow a similar procedure but average over time instead of space before plotting on the same grids that the observations use.\n\nWe also calculate the probability of them being \"sea-ice-free\" in the future as the fraction of models predicting the sea ice area dropping below a given region-dependent threshold (discussed below). We also do the same calculations for two Arctic shipping routes - the Transpolar Sea Route (TSR) and the Northern |Sea Route (NSR).\n\nOne point to note here is that we need to choose an area threshold for when a region is ice-free. A consequence of this is that the probabilities we calculate are a little arbitrary and so we should pay more attention to how they are changing with time than the absolute values.\nThe thresholds we use to decide when the regions are ice-free are as follows. The ice-free threshold for the Arctic is usually taken as $10^6$ km$^2$ (e.g. [Jahn et al., 2016](https://doi.org/10.1002/2016GL070067)). For the Antarctic this is too high since nearly all of the CMIP6 models significantly underestimate the Antarctic area and accordingly the minimum area for these models is less than $10^6$ km$^2$ even in the historical period ([Roach et al., 2023](https://doi.org/10.1029/2019GL086729); [CDS assessment of historical sea ice extent](https://ecmwf-projects.github.io/c3s2-eqc-quality-assessment/Climate_Projections/CMIP6/climate_projections-cmip6_model-performance_q02.html)). Therefore we choose the threshold to be $0.25 \\times 10^6$ km$^2$, which is about one seventh of the pre-1970 Antarctic minima (according to the models) of about $1.8 \\times 10^6$ km$^2$. This relationship is the approximate relationship in the Arctic, since the pre-1970 Arctic minima are about $7 \\times 10^6$ km$^2$. For the two shipping routes, we also divide the pre-1970 Antarctic minima (again according to the models) by seven, giving thresholds of $0.14 \\times 10^6$ km$^2$ for the NSR and $0.21 \\times 10^6$ km$^2$ for the TSR.", "text_with_prefix": "EQC Quality Assessment: \"Projections of future ice-free periods for the Arctic and Antarctic\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q04 | Category: Climate_Projections\nSection: Projections of future ice-free periods for the Arctic and Antarctic > Methodology\n---\n5` (22 models outputting SIC)\n- `ssp3_7_0` (21 models outputting SIC)\n- `ssp5_8_5` (21 models outputting SIC)\n\nThe experiments are ordered by expected warming with `ssp_1_2_6` giving the least warming (reaching 2.6 W/m$^2$ in 2100, which is a kind of \"best case\" warming scenario) and `ssp5_8_5` giving high warming (8.5 W/m$^2$ in 2100).\nThe first number refers to which of the five [Shared Socioeconomic Pathways](https://www.dkrz.de/en/communication/climate-simulations/cmip6-en/the-ssp-scenarios) (SSP) that the experiment corresponds to.\n\nWe calculate the minimum and maximum sea ice areas for the Arctic and Antarctica, show the historical and projected results as well as results derived from satellite observations, and show some climatological maps to show projected changes in the spatial distribution of SIC.\nTo calculate the area we remap the model SIC onto the equal-area grids used by the observations. The ice-covered area in each grid cell can be determined from the SIC and the grid cell area, and then summed up to get the total ice-covered area.\nFor each experiment, we then have an ensemble of areal time series (one from each model) and we plot the median and IQL (inter-quartile limits) of this ensemble.\nFor the climatological maps we follow a similar procedure but average over time instead of space before plotting on the same grids that the observations use.\n\nWe also calculate the probability of them being \"sea-ice-free\" in the future as the fraction of models predicting the sea ice area dropping below a given region-dependent threshold (discussed below). We also do the same calculations for two Arctic shipping routes - the Transpolar Sea Route (TSR) and the Northern |Sea Route (NSR).\n\nOne point to note here is that we need to choose an area threshold for when a region is ice-free. A consequence of this is that the probabilities we calculate are a little arbitrary and so we should pay more attention to how they are changing with time than the absolute values.\nThe thresholds we use to decide when the regions are ice-free are as follows. The ice-free threshold for the Arctic is usually taken as $10^6$ km$^2$ (e.g. [Jahn et al., 2016](https://doi.org/10.1002/2016GL070067)). For the Antarctic this is too high since nearly all of the CMIP6 models significantly underestimate the Antarctic area and accordingly the minimum area for these models is less than $10^6$ km$^2$ even in the historical period ([Roach et al., 2023](https://doi.org/10.1029/2019GL086729); [CDS assessment of historical sea ice extent](https://ecmwf-projects.github.io/c3s2-eqc-quality-assessment/Climate_Projections/CMIP6/climate_projections-cmip6_model-performance_q02.html)). Therefore we choose the threshold to be $0.25 \\times 10^6$ km$^2$, which is about one seventh of the pre-1970 Antarctic minima (according to the models) of about $1.8 \\times 10^6$ km$^2$. This relationship is the approximate relationship in the Arctic, since the pre-1970 Arctic minima are about $7 \\times 10^6$ km$^2$. For the two shipping routes, we also divide the pre-1970 Antarctic minima (again according to the models) by seven, giving thresholds of $0.14 \\times 10^6$ km$^2$ for the NSR and $0.21 \\times 10^6$ km$^2$ for the TSR."} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q04__2c5eadf2ec13", "report_id": "climate_projections-cmip6_climate-impact-indicators_q04", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q04", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Projections of future ice-free periods for the Arctic and Antarctic > Methodology", "title": "Projections of future ice-free periods for the Arctic and Antarctic", "chunk_index": 5, "token_count": 671, "text_raw": "$0.14 \\times 10^6$ km$^2$ for the NSR and $0.21 \\times 10^6$ km$^2$ for the TSR.\n\nOther authors have also studied accessibility of these and other shipping routes (eg. [Chen et al, 2023](https://doi.org/10.1016/j.accre.2023.11.011);[X. Zhao et al., 2022](https://doi.org/10.1029/2022EF002708); [Wang et al, 2021](https://doi.org/10.1088/1748-9326/ac0b17); [Min et al., 2022](https://doi.org/10.1029/2022GL099157); [Li et al., 2021](https://doi.org/10.1007/s10584-021-03172-3)). For these authors, accessibility is determined by the polar class of the ship (how much ice strengthening the ship has) (if any) and the sea ice concentration and thickness. These variables are generally used as inputs into numerical ship-routing software like the Arctic Transport Accessibility Model ([Smith & Stephenson, 2013](https://doi.org/10.1017/jog.2020.73)) to create an ensemble of routes (created by using different start and end points for the journey) which are then analysed. Our approach, which would be applicable to open water vessels without ice strengthening, is comparatively simpler but is enough to give an overview of how likely the routes are to be clear and the effect of increased warming.\n\nThe \"Analysis and results\" section is organised as follows:\n\n**[](section-1)**\n\n**[](section-2)**\n\n**[](section-3)**\n\n        **[](section-3.1)**\n\n        **[](section-3.2)**\n\n        **[](section-3.3)**\n\n        **[](section-3.4)**\n\n        **[](section-3.5)**\n\n                **[](section-3.5.1)**\n\n                **[](section-3.5.2)**", "text_with_prefix": "EQC Quality Assessment: \"Projections of future ice-free periods for the Arctic and Antarctic\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q04 | Category: Climate_Projections\nSection: Projections of future ice-free periods for the Arctic and Antarctic > Methodology\n---\n$0.14 \\times 10^6$ km$^2$ for the NSR and $0.21 \\times 10^6$ km$^2$ for the TSR.\n\nOther authors have also studied accessibility of these and other shipping routes (eg. [Chen et al, 2023](https://doi.org/10.1016/j.accre.2023.11.011);[X. Zhao et al., 2022](https://doi.org/10.1029/2022EF002708); [Wang et al, 2021](https://doi.org/10.1088/1748-9326/ac0b17); [Min et al., 2022](https://doi.org/10.1029/2022GL099157); [Li et al., 2021](https://doi.org/10.1007/s10584-021-03172-3)). For these authors, accessibility is determined by the polar class of the ship (how much ice strengthening the ship has) (if any) and the sea ice concentration and thickness. These variables are generally used as inputs into numerical ship-routing software like the Arctic Transport Accessibility Model ([Smith & Stephenson, 2013](https://doi.org/10.1017/jog.2020.73)) to create an ensemble of routes (created by using different start and end points for the journey) which are then analysed. Our approach, which would be applicable to open water vessels without ice strengthening, is comparatively simpler but is enough to give an overview of how likely the routes are to be clear and the effect of increased warming.\n\nThe \"Analysis and results\" section is organised as follows:\n\n**[](section-1)**\n\n**[](section-2)**\n\n**[](section-3)**\n\n        **[](section-3.1)**\n\n        **[](section-3.2)**\n\n        **[](section-3.3)**\n\n        **[](section-3.4)**\n\n        **[](section-3.5)**\n\n                **[](section-3.5.1)**\n\n                **[](section-3.5.2)**"} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q04__43f32989630f", "report_id": "climate_projections-cmip6_climate-impact-indicators_q04", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q04", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Projections of future ice-free periods for the Arctic and Antarctic > Analysis and results > 1. Import libraries, set parameters and definitions of functions > 1.2 Set parameters", "title": "Projections of future ice-free periods for the Arctic and Antarctic", "chunk_index": 6, "token_count": 389, "text_raw": "- Set the time period to be analysed with `year_start` and `year_stop`.\n- Set the regions to be analysed with the list `sea_masks`.\n- Set the area thresholds for determining when each region is ice-free with `area_thresholds`.\n- Set the concentration threshold `sic_threshold` for determining sea ice extent (we use 30% to be consistent with the ice edge product).\n- Set the months we want to determine climatologies for with `clim_months`.\n- Set the averaging window (in years) for the climatologies for with `clim_length`.\n- Set the starting years for the climatologies for with `clim_start_years`.\n- Set the map projections for plotting climatologies with `projections`.\n- Set the map extents for plotting climatologies with `map_slices`.\n- Set the experiments to be considered with the list `experiments`.\n- Set the models to be evaluated with the dictionary `models_dict`.\n\nSelect masks\nExtent thresholds for being considered ice-free (units 10^6 km^2)\n- we take the Antarctic threshold to be about one seventh of the pre-1970 minima,\nsince this is the relationship in the Arctic\nSea Ice Concentration Threshold\nMonths to get climatologies for\nLengths of climatology windows\nStart years of the climatologies\nMap projections for plotting climatologies\nSlices in x and y directions (on observation grid) for zooming in on climatologies", "text_with_prefix": "EQC Quality Assessment: \"Projections of future ice-free periods for the Arctic and Antarctic\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q04 | Category: Climate_Projections\nSection: Projections of future ice-free periods for the Arctic and Antarctic > Analysis and results > 1. Import libraries, set parameters and definitions of functions > 1.2 Set parameters\n---\n- Set the time period to be analysed with `year_start` and `year_stop`.\n- Set the regions to be analysed with the list `sea_masks`.\n- Set the area thresholds for determining when each region is ice-free with `area_thresholds`.\n- Set the concentration threshold `sic_threshold` for determining sea ice extent (we use 30% to be consistent with the ice edge product).\n- Set the months we want to determine climatologies for with `clim_months`.\n- Set the averaging window (in years) for the climatologies for with `clim_length`.\n- Set the starting years for the climatologies for with `clim_start_years`.\n- Set the map projections for plotting climatologies with `projections`.\n- Set the map extents for plotting climatologies with `map_slices`.\n- Set the experiments to be considered with the list `experiments`.\n- Set the models to be evaluated with the dictionary `models_dict`.\n\nSelect masks\nExtent thresholds for being considered ice-free (units 10^6 km^2)\n- we take the Antarctic threshold to be about one seventh of the pre-1970 minima,\nsince this is the relationship in the Arctic\nSea Ice Concentration Threshold\nMonths to get climatologies for\nLengths of climatology windows\nStart years of the climatologies\nMap projections for plotting climatologies\nSlices in x and y directions (on observation grid) for zooming in on climatologies"} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q04__de9e1cd59c26", "report_id": "climate_projections-cmip6_climate-impact-indicators_q04", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q04", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Projections of future ice-free periods for the Arctic and Antarctic > Analysis and results > 1. Import libraries, set parameters and definitions of functions > 1.4 Define functions to compute time series of extent and area", "title": "Projections of future ice-free periods for the Arctic and Antarctic", "chunk_index": 7, "token_count": 236, "text_raw": "- `apply_sea_mask` chooses the region to analyse.\n- `compute_extent_and_area_from_sic` computes the extent and area from a sea ice concentration field.\n- `interpolate_to_satellite_grid` interpolates the model sea ice concentration to the reference grid (chosen to be the grid of the satellite sea ice concentration dataset).\n- `compute_interpolated_sea_ice_extent_and_area` calls `interpolate_to_satellite_grid` to interpolate the model sea ice concentration to the reference grid and then calls `compute_extent_and_area_from_sic` to calculate the extent and area.\n\nConvert longitude\nCompute diagnostics\nMerge and add attributes\nRemove nan columns\nMonthly resample", "text_with_prefix": "EQC Quality Assessment: \"Projections of future ice-free periods for the Arctic and Antarctic\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q04 | Category: Climate_Projections\nSection: Projections of future ice-free periods for the Arctic and Antarctic > Analysis and results > 1. Import libraries, set parameters and definitions of functions > 1.4 Define functions to compute time series of extent and area\n---\n- `apply_sea_mask` chooses the region to analyse.\n- `compute_extent_and_area_from_sic` computes the extent and area from a sea ice concentration field.\n- `interpolate_to_satellite_grid` interpolates the model sea ice concentration to the reference grid (chosen to be the grid of the satellite sea ice concentration dataset).\n- `compute_interpolated_sea_ice_extent_and_area` calls `interpolate_to_satellite_grid` to interpolate the model sea ice concentration to the reference grid and then calls `compute_extent_and_area_from_sic` to calculate the extent and area.\n\nConvert longitude\nCompute diagnostics\nMerge and add attributes\nRemove nan columns\nMonthly resample"} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q04__24a96635e1fa", "report_id": "climate_projections-cmip6_climate-impact-indicators_q04", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q04", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Projections of future ice-free periods for the Arctic and Antarctic > Analysis and results > 1. Import libraries, set parameters and definitions of functions > 1.5 Functions to post-process and plot time series", "title": "Projections of future ice-free periods for the Arctic and Antarctic", "chunk_index": 8, "token_count": 314, "text_raw": "- `postprocess_dataset` makes the combined dataset easier to work with by making sure all each individual dataset uses the same calendar, and by renaming some variables and attributes.\n- `full_year_only_resample` resamples a time series to yearly, by either taking the minimum or maximum for the year.\n- `plot_timeseries`loops over each experiment and plots the median and the interquartile range (IQR) of the sea ice area of the ensemble of models. It also plots the area given by the satellite sea ice concentration products. The time series are reduced to yearly frequency according to the argument `reduction` which is a string eg `\"min\"` or `\"max\"`.\n- `plot_time_series_regional_reduction` loops over two different time periods, the full CMIP6 period and a \"zoomed-in\" period (1980-2080 or 2020-2100), and calls `plot_time_series` for each.\n\nDefine colors\nGet dataarrays\nPlot satellites\nplot the ice-free level if it is passed in\nset xtick label format", "text_with_prefix": "EQC Quality Assessment: \"Projections of future ice-free periods for the Arctic and Antarctic\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q04 | Category: Climate_Projections\nSection: Projections of future ice-free periods for the Arctic and Antarctic > Analysis and results > 1. Import libraries, set parameters and definitions of functions > 1.5 Functions to post-process and plot time series\n---\n- `postprocess_dataset` makes the combined dataset easier to work with by making sure all each individual dataset uses the same calendar, and by renaming some variables and attributes.\n- `full_year_only_resample` resamples a time series to yearly, by either taking the minimum or maximum for the year.\n- `plot_timeseries`loops over each experiment and plots the median and the interquartile range (IQR) of the sea ice area of the ensemble of models. It also plots the area given by the satellite sea ice concentration products. The time series are reduced to yearly frequency according to the argument `reduction` which is a string eg `\"min\"` or `\"max\"`.\n- `plot_time_series_regional_reduction` loops over two different time periods, the full CMIP6 period and a \"zoomed-in\" period (1980-2080 or 2020-2100), and calls `plot_time_series` for each.\n\nDefine colors\nGet dataarrays\nPlot satellites\nplot the ice-free level if it is passed in\nset xtick label format"} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q04__2c6feebc0947", "report_id": "climate_projections-cmip6_climate-impact-indicators_q04", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q04", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Projections of future ice-free periods for the Arctic and Antarctic > Analysis and results > 1. Import libraries, set parameters and definitions of functions > 1.6 Functions to create the monthly climatologies", "title": "Projections of future ice-free periods for the Arctic and Antarctic", "chunk_index": 9, "token_count": 315, "text_raw": "- `compute_monthly_climatology` is the function that is used by `download.download_and_transform`. It takes in an `xarray.Dataset` and takes the temporal mean for each month, before calling `interpolate_to_satellite_grid` to interpolate the result to the satellite grid.\n- `postprocess_climatology` makes the dataset easier to use by renaming some variables and ensuring the sea ice concentration has units `%`.\n- `get_monthly_climatology_model` finalises the request to be passed to `download.download_and_transform`, calls `download.download_and_transform`, and then calls `postprocess_climatology` and returns the post-processed dataset.\n- `get_monthly_climatologies_cmip6` creates the ensemble mean of monthly climatologies of the sea ice concentration, for a given experiment and time period. It does this by looping over the models in a given experiment and and calling `get_monthly_climatology_model` for each.\n\nrename month\nrename SIC and convert to %\nsome models produce extra variables so drop any that are not needed", "text_with_prefix": "EQC Quality Assessment: \"Projections of future ice-free periods for the Arctic and Antarctic\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q04 | Category: Climate_Projections\nSection: Projections of future ice-free periods for the Arctic and Antarctic > Analysis and results > 1. Import libraries, set parameters and definitions of functions > 1.6 Functions to create the monthly climatologies\n---\n- `compute_monthly_climatology` is the function that is used by `download.download_and_transform`. It takes in an `xarray.Dataset` and takes the temporal mean for each month, before calling `interpolate_to_satellite_grid` to interpolate the result to the satellite grid.\n- `postprocess_climatology` makes the dataset easier to use by renaming some variables and ensuring the sea ice concentration has units `%`.\n- `get_monthly_climatology_model` finalises the request to be passed to `download.download_and_transform`, calls `download.download_and_transform`, and then calls `postprocess_climatology` and returns the post-processed dataset.\n- `get_monthly_climatologies_cmip6` creates the ensemble mean of monthly climatologies of the sea ice concentration, for a given experiment and time period. It does this by looping over the models in a given experiment and and calling `get_monthly_climatology_model` for each.\n\nrename month\nrename SIC and convert to %\nsome models produce extra variables so drop any that are not needed"} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q04__234da62d8dc1", "report_id": "climate_projections-cmip6_climate-impact-indicators_q04", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q04", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Projections of future ice-free periods for the Arctic and Antarctic > Analysis and results > 1. Import libraries, set parameters and definitions of functions > 1.7 Functions to plot climatology maps", "title": "Projections of future ice-free periods for the Arctic and Antarctic", "chunk_index": 10, "token_count": 186, "text_raw": "- `get_datasets_eumetsat_clim` downloads satellite data for each hemisphere for one day. These datasets will be used to add a mask to CMIP6 projection data to improve their visual appearance.\n- `make_sic_maps` plots the climatology maps for each experiment (in columns) and time period (in rows).\n- `compare_sic_maps` is a wrapper that organises the inputs to `make_sic_maps` before calling it.", "text_with_prefix": "EQC Quality Assessment: \"Projections of future ice-free periods for the Arctic and Antarctic\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q04 | Category: Climate_Projections\nSection: Projections of future ice-free periods for the Arctic and Antarctic > Analysis and results > 1. Import libraries, set parameters and definitions of functions > 1.7 Functions to plot climatology maps\n---\n- `get_datasets_eumetsat_clim` downloads satellite data for each hemisphere for one day. These datasets will be used to add a mask to CMIP6 projection data to improve their visual appearance.\n- `make_sic_maps` plots the climatology maps for each experiment (in columns) and time period (in rows).\n- `compare_sic_maps` is a wrapper that organises the inputs to `make_sic_maps` before calling it."} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q04__3bd56d6934f8", "report_id": "climate_projections-cmip6_climate-impact-indicators_q04", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q04", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Projections of future ice-free periods for the Arctic and Antarctic > Analysis and results > 1. Import libraries, set parameters and definitions of functions > 1.8 Functions to plot the access to a given area", "title": "Projections of future ice-free periods for the Arctic and Antarctic", "chunk_index": 11, "token_count": 401, "text_raw": "- `plot_sea_masks` makes maps illustrating where we define the transpolar shipping route and the northern sea shipping route.\n- `get_access_table` sorts the time series of sea ice area to make a dataset with dimensions `year` and `month` containing the probability that a region will be clear for that year and month. The probability is calculated by considering each model in an experiment.\n- `get_access_tables` loops over each CMIP6 projection experiment and calls `get_access_table`.\n- `plot_access_tables` then plots the probability of a region being clear for each experiment by year and month.\n- `plot_access_vs_month_one_year` plots `Probability clear` against `month` for all the experiments and for a given year. That is, it combines vertical slices of the access tables for all the different experiments.\n- `plot_access_vs_month` loops over a number of years and calls `plot_access_vs_month_one_year`.\n\ninitialise mask from SIC land mask\napply sea mask and zoom in on it\nadd area to title\nCreate a boolean mask where area > area_threshold\nConvert boolean values to integers (True -> 1, False -> 0)\nCalculate the fraction of models where area > area_threshold along the 'model' dimension\nGet the time values and extract year and month\nFind unique years and months\nCreate an empty 2D array with dimensions ('month', 'year')\nFill the 2D array with the corresponding probability_clear values\n\n(section-2)=", "text_with_prefix": "EQC Quality Assessment: \"Projections of future ice-free periods for the Arctic and Antarctic\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q04 | Category: Climate_Projections\nSection: Projections of future ice-free periods for the Arctic and Antarctic > Analysis and results > 1. Import libraries, set parameters and definitions of functions > 1.8 Functions to plot the access to a given area\n---\n- `plot_sea_masks` makes maps illustrating where we define the transpolar shipping route and the northern sea shipping route.\n- `get_access_table` sorts the time series of sea ice area to make a dataset with dimensions `year` and `month` containing the probability that a region will be clear for that year and month. The probability is calculated by considering each model in an experiment.\n- `get_access_tables` loops over each CMIP6 projection experiment and calls `get_access_table`.\n- `plot_access_tables` then plots the probability of a region being clear for each experiment by year and month.\n- `plot_access_vs_month_one_year` plots `Probability clear` against `month` for all the experiments and for a given year. That is, it combines vertical slices of the access tables for all the different experiments.\n- `plot_access_vs_month` loops over a number of years and calls `plot_access_vs_month_one_year`.\n\ninitialise mask from SIC land mask\napply sea mask and zoom in on it\nadd area to title\nCreate a boolean mask where area > area_threshold\nConvert boolean values to integers (True -> 1, False -> 0)\nCalculate the fraction of models where area > area_threshold along the 'model' dimension\nGet the time values and extract year and month\nFind unique years and months\nCreate an empty 2D array with dimensions ('month', 'year')\nFill the 2D array with the corresponding probability_clear values\n\n(section-2)="} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q04__59e25881820c", "report_id": "climate_projections-cmip6_climate-impact-indicators_q04", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q04", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Projections of future ice-free periods for the Arctic and Antarctic > Analysis and results > 2. Download and transform data > 2.1 Set some downloading parameters", "title": "Projections of future ice-free periods for the Arctic and Antarctic", "chunk_index": 12, "token_count": 191, "text_raw": "- `io_kwargs` are some options to speed-up the downloading.\n- `common_kwargs` adds the `transform_func` option to make `download.download_and_transform` transform the data using `compute_interpolated_sea_ice_extent_and_area`.\n- `transform_func_kwargs` are options to be passed to `compute_interpolated_sea_ice_extent_and_area`.\n- `interpolation_kwargs` are options for interpolation.\n- `region_name_mapper` makes the region names more user friendly.\n\nParameters to speed up IO", "text_with_prefix": "EQC Quality Assessment: \"Projections of future ice-free periods for the Arctic and Antarctic\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q04 | Category: Climate_Projections\nSection: Projections of future ice-free periods for the Arctic and Antarctic > Analysis and results > 2. Download and transform data > 2.1 Set some downloading parameters\n---\n- `io_kwargs` are some options to speed-up the downloading.\n- `common_kwargs` adds the `transform_func` option to make `download.download_and_transform` transform the data using `compute_interpolated_sea_ice_extent_and_area`.\n- `transform_func_kwargs` are options to be passed to `compute_interpolated_sea_ice_extent_and_area`.\n- `interpolation_kwargs` are options for interpolation.\n- `region_name_mapper` makes the region names more user friendly.\n\nParameters to speed up IO"} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q04__b957f8b33fff", "report_id": "climate_projections-cmip6_climate-impact-indicators_q04", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q04", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Projections of future ice-free periods for the Arctic and Antarctic > Analysis and results > 2. Download and transform data > 2.2 Download and transform satellite data for time series", "title": "Projections of future ice-free periods for the Arctic and Antarctic", "chunk_index": 13, "token_count": 194, "text_raw": "Loops over the two satellites and the four masks to get time series of sea ice extent and area for the satellite sea ice concentration products. `download.download_and_transform` downloads the data and transforms it with `compute_interpolated_sea_ice_extent_and_area` before saving it to disk. The dataset is then post-processed with `postprocess_dataset`. The time series are stored in the dictionary `datasets_satellite`.\n\ndownload SIC for one date to get land mask for grid to illustrate the different sea masks", "text_with_prefix": "EQC Quality Assessment: \"Projections of future ice-free periods for the Arctic and Antarctic\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q04 | Category: Climate_Projections\nSection: Projections of future ice-free periods for the Arctic and Antarctic > Analysis and results > 2. Download and transform data > 2.2 Download and transform satellite data for time series\n---\nLoops over the two satellites and the four masks to get time series of sea ice extent and area for the satellite sea ice concentration products. `download.download_and_transform` downloads the data and transforms it with `compute_interpolated_sea_ice_extent_and_area` before saving it to disk. The dataset is then post-processed with `postprocess_dataset`. The time series are stored in the dictionary `datasets_satellite`.\n\ndownload SIC for one date to get land mask for grid to illustrate the different sea masks"} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q04__3a9b6670fa1b", "report_id": "climate_projections-cmip6_climate-impact-indicators_q04", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q04", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Projections of future ice-free periods for the Arctic and Antarctic > Analysis and results > 2. Download and transform data > 2.3 Download and transform CMIP6 data for time series", "title": "Projections of future ice-free periods for the Arctic and Antarctic", "chunk_index": 14, "token_count": 185, "text_raw": "Loops over each model in each experiment, and the four masks, to get time series of sea ice extent and area for the CMIP6 models. For each combination, `download.download_and_transform` downloads the data and transforms it with `compute_interpolated_sea_ice_extent_and_area` before saving it to disk. The dataset is then post-processed with `postprocess_dataset`. The time series are stored in the dictionary `datasets_cmip6`.", "text_with_prefix": "EQC Quality Assessment: \"Projections of future ice-free periods for the Arctic and Antarctic\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q04 | Category: Climate_Projections\nSection: Projections of future ice-free periods for the Arctic and Antarctic > Analysis and results > 2. Download and transform data > 2.3 Download and transform CMIP6 data for time series\n---\nLoops over each model in each experiment, and the four masks, to get time series of sea ice extent and area for the CMIP6 models. For each combination, `download.download_and_transform` downloads the data and transforms it with `compute_interpolated_sea_ice_extent_and_area` before saving it to disk. The dataset is then post-processed with `postprocess_dataset`. The time series are stored in the dictionary `datasets_cmip6`."} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q04__4ec6200b1494", "report_id": "climate_projections-cmip6_climate-impact-indicators_q04", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q04", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Projections of future ice-free periods for the Arctic and Antarctic > Analysis and results > 2. Download and transform data > 2.4 Download and process climatologies for CMIP6 projections", "title": "Projections of future ice-free periods for the Arctic and Antarctic", "chunk_index": 15, "token_count": 174, "text_raw": "This step loops over each hemisphere, each model in each experiment, and each time period, and calls `get_monthly_climatologies_cmip6` to get the ensemble mean (average over all models in an experiment) for each hemisphere, experiment and time period. The climatologies are stored in the dictionary `datasets_cmip6_clim`.\n\noptions for plotting climatologies\n\n(section-3)=", "text_with_prefix": "EQC Quality Assessment: \"Projections of future ice-free periods for the Arctic and Antarctic\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q04 | Category: Climate_Projections\nSection: Projections of future ice-free periods for the Arctic and Antarctic > Analysis and results > 2. Download and transform data > 2.4 Download and process climatologies for CMIP6 projections\n---\nThis step loops over each hemisphere, each model in each experiment, and each time period, and calls `get_monthly_climatologies_cmip6` to get the ensemble mean (average over all models in an experiment) for each hemisphere, experiment and time period. The climatologies are stored in the dictionary `datasets_cmip6_clim`.\n\noptions for plotting climatologies\n\n(section-3)="} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q04__dd9bf6d71795", "report_id": "climate_projections-cmip6_climate-impact-indicators_q04", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q04", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Projections of future ice-free periods for the Arctic and Antarctic > Analysis and results > 3. Results", "title": "Projections of future ice-free periods for the Arctic and Antarctic", "chunk_index": 16, "token_count": 163, "text_raw": "In this section we plot yearly minima and maxima for sea ice area, for the Arctic and Antarctic. We also estimate the probability that a region is ice-free,\nby taking the ensemble of models and using the percentage of models predicting it as having low enough area as the probability. We chart this by month-by-month to also see how much of the year periods with a high chance of being ice-free are happening.\n\n(section-3.1)=", "text_with_prefix": "EQC Quality Assessment: \"Projections of future ice-free periods for the Arctic and Antarctic\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q04 | Category: Climate_Projections\nSection: Projections of future ice-free periods for the Arctic and Antarctic > Analysis and results > 3. Results\n---\nIn this section we plot yearly minima and maxima for sea ice area, for the Arctic and Antarctic. We also estimate the probability that a region is ice-free,\nby taking the ensemble of models and using the percentage of models predicting it as having low enough area as the probability. We chart this by month-by-month to also see how much of the year periods with a high chance of being ice-free are happening.\n\n(section-3.1)="} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q04__ea50b079de7e", "report_id": "climate_projections-cmip6_climate-impact-indicators_q04", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q04", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Projections of future ice-free periods for the Arctic and Antarctic > Analysis and results > 3. Results > 3.1 Arctic sea ice minima and the projected probabilities of an ice-free Arctic", "title": "Projections of future ice-free periods for the Arctic and Antarctic", "chunk_index": 17, "token_count": 996, "text_raw": "In the Arctic region, the yearly minimum in sea ice area shows a clear effect of increased warming. Before about 1990 the historical CMIP6 is relatively constant. After this time however it drops rapidly. This agrees quite well with the satellite data when they are available. As we come to the start of the projection period (2015-2100), the yearly minima continues to drop. The minimum area is highest for the `ssp1_2_6` experiment and its median plateaus around the ice-free level, so about half the models predict the Arctic will be ice-free in the projection period under this scenario. The medians for the `ssp2_4_5` and `ssp3_7_0` minimum extents are similar, both reaching the ice-free level about 2050, and even dropping close to zero by the end of the century. The `ssp5_8_5` median is consistently the lowest and reaches the ice-free level around 2040. The median drops to near zero around 2070.\n\nThe yearly minimum in sea ice area follows mostly the same pattern, although the time when the median and lower quartile reach zero is slightly later since there is no concentration threshold used in the calculation (a grid cell is classed as ice-free if the concentration is below 30%).\n\nThe maps below show the spatial pattern of the decrease in sea ice cover by plotting the ensemble mean of the sea ice concentration for a series of 20-year climatologies for the month of September. The magenta lines plot the ice edge (15% concentration contour). Going from left to right (increasing the amount of warming in the projection), there is quite a big difference, especially from 2035. From 2055 only the `ssp_1_2_6` experiment has a significant amount of ice.\n\nThe heat maps below show the proportion of models having a minimum Arctic sea ice area less than $10^6$ km$^2$. For the `ssp1_2_6` experiment, the chances of being ice-free in September reach 20% around 2040, 40% around 2050 and 50% around 2070. In August there are consistently 40-50% probabilities after 2080, while in October the chances are about 20-30% from this time. From 2090, about 15% of the models predict ice-free conditions in July under this warming scenario.\n\nWe can see however that the chances are being overestimated (perhaps indicating a bias of about 10%) since some years in the past (at time of writing i.e. 2015-2024) have 10% chance of the Arctic being ice-free (1 in 10 models predict this) while it has never been observed yet. Choosing a different subset of models could also avoid this bias.\n\nIn the `ssp2_4_5` experiment, the pattern is similar but the probabilities have increased somewhat. In September they are above 80% from about 2085, with it being 95-100% for three of the later years. In August and October the chances reach about 70% (in 2085 and 2093 respectively), and 20% of the models are predicting ice-free conditions in July and November.\n\nIn the `ssp3_7_0` experiment, there are now 85%-90% chances of being ice-free in August-October after about 2085. The chances in July and November have also increased (above 40% after about 2083, and sometimes reaching 60%), and there is also a 20%-30% chance of it being ice-free in June and December from about 2095.\n\nIn the `ssp5_8_5` experiment, there are now chances of the Arctic being ice-free from August to October of around 70-80% from about 2070. In September, the chances of being ice-free are consistently over 80% from around 2065. July and November have high chances (70%-85%) of an ice-free Arctic in the last decade or two under this scenario, and the chances of June and December start to become significant earlier than under `ssp3_7_0`. In this experiment (`ssp5_8_5`), even May and January have 20% probability in the last 5 years or so.", "text_with_prefix": "EQC Quality Assessment: \"Projections of future ice-free periods for the Arctic and Antarctic\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q04 | Category: Climate_Projections\nSection: Projections of future ice-free periods for the Arctic and Antarctic > Analysis and results > 3. Results > 3.1 Arctic sea ice minima and the projected probabilities of an ice-free Arctic\n---\nIn the Arctic region, the yearly minimum in sea ice area shows a clear effect of increased warming. Before about 1990 the historical CMIP6 is relatively constant. After this time however it drops rapidly. This agrees quite well with the satellite data when they are available. As we come to the start of the projection period (2015-2100), the yearly minima continues to drop. The minimum area is highest for the `ssp1_2_6` experiment and its median plateaus around the ice-free level, so about half the models predict the Arctic will be ice-free in the projection period under this scenario. The medians for the `ssp2_4_5` and `ssp3_7_0` minimum extents are similar, both reaching the ice-free level about 2050, and even dropping close to zero by the end of the century. The `ssp5_8_5` median is consistently the lowest and reaches the ice-free level around 2040. The median drops to near zero around 2070.\n\nThe yearly minimum in sea ice area follows mostly the same pattern, although the time when the median and lower quartile reach zero is slightly later since there is no concentration threshold used in the calculation (a grid cell is classed as ice-free if the concentration is below 30%).\n\nThe maps below show the spatial pattern of the decrease in sea ice cover by plotting the ensemble mean of the sea ice concentration for a series of 20-year climatologies for the month of September. The magenta lines plot the ice edge (15% concentration contour). Going from left to right (increasing the amount of warming in the projection), there is quite a big difference, especially from 2035. From 2055 only the `ssp_1_2_6` experiment has a significant amount of ice.\n\nThe heat maps below show the proportion of models having a minimum Arctic sea ice area less than $10^6$ km$^2$. For the `ssp1_2_6` experiment, the chances of being ice-free in September reach 20% around 2040, 40% around 2050 and 50% around 2070. In August there are consistently 40-50% probabilities after 2080, while in October the chances are about 20-30% from this time. From 2090, about 15% of the models predict ice-free conditions in July under this warming scenario.\n\nWe can see however that the chances are being overestimated (perhaps indicating a bias of about 10%) since some years in the past (at time of writing i.e. 2015-2024) have 10% chance of the Arctic being ice-free (1 in 10 models predict this) while it has never been observed yet. Choosing a different subset of models could also avoid this bias.\n\nIn the `ssp2_4_5` experiment, the pattern is similar but the probabilities have increased somewhat. In September they are above 80% from about 2085, with it being 95-100% for three of the later years. In August and October the chances reach about 70% (in 2085 and 2093 respectively), and 20% of the models are predicting ice-free conditions in July and November.\n\nIn the `ssp3_7_0` experiment, there are now 85%-90% chances of being ice-free in August-October after about 2085. The chances in July and November have also increased (above 40% after about 2083, and sometimes reaching 60%), and there is also a 20%-30% chance of it being ice-free in June and December from about 2095.\n\nIn the `ssp5_8_5` experiment, there are now chances of the Arctic being ice-free from August to October of around 70-80% from about 2070. In September, the chances of being ice-free are consistently over 80% from around 2065. July and November have high chances (70%-85%) of an ice-free Arctic in the last decade or two under this scenario, and the chances of June and December start to become significant earlier than under `ssp3_7_0`. In this experiment (`ssp5_8_5`), even May and January have 20% probability in the last 5 years or so."} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q04__00da348378aa", "report_id": "climate_projections-cmip6_climate-impact-indicators_q04", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q04", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Projections of future ice-free periods for the Arctic and Antarctic > Analysis and results > 3. Results > 3.1 Arctic sea ice minima and the projected probabilities of an ice-free Arctic", "title": "Projections of future ice-free periods for the Arctic and Antarctic", "chunk_index": 18, "token_count": 691, "text_raw": "start to become significant earlier than under `ssp3_7_0`. In this experiment (`ssp5_8_5`), even May and January have 20% probability in the last 5 years or so.\n\nIn summary, the heat maps above show that as the warming increases, we can clearly see that the overall probability of being ice-free is increasing, and the length of the season where there is high probability is also increasing from just September (`ssp1_2_6`) to the 6-month period from June to December (`ssp5_8_5`).\nThis can also be seen in the plots below, where we compare the experiments more directly by plotting the monthly probabilities for a selection of years.\n\nIn the years before 2060, the probabilities are not ordered by warming strength, but around 2060 `ssp3_7_0` reaches the `ssp2_4_5` level, and after that year the probabilities' order does correspond to warming strength.\nIn 2040, the maximum probability ranges from about 20% to 40%; in 2050 and 2060 it ranges from about 40% to 60%; in 2070, 2080 and 2090 it ranges from about 45% to about 85%. In the latter three years the main effect is the widening of the season with the highest probabilities.\n\nComparing to other authors estimates, the [SIMIP community (2020)](https://doi.org/10.1029/2019GL086749) found that nearly all of their selected models projected ice-free conditions by 2050, regardless of warming scenario. [X. Zhao et al. (2022)](https://doi.org/10.1088/1748-9326/ac9d4d) agree with this - they projected the first ice-free year to be $2049\\pm 12$ under `ssp2_4_5` and $2043\\pm 9$ under `ssp5_8_5`, giving a similar range to the [SIMIP community (2020)](https://doi.org/10.1029/2019GL086749). While our 2050 probability of just under 60% for `ssp5_8_5` is quite high, 40% for `ssp1_2_6` is quite a lot lower. This difference possibly reflects the effect both of using a different subset of models and of using the full ensemble set provided for each model, as [SIMIP community (2020)](https://doi.org/10.1029/2019GL086749) did (instead of just using the ensemble mean for each model).\n\n(section-3.2)=", "text_with_prefix": "EQC Quality Assessment: \"Projections of future ice-free periods for the Arctic and Antarctic\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q04 | Category: Climate_Projections\nSection: Projections of future ice-free periods for the Arctic and Antarctic > Analysis and results > 3. Results > 3.1 Arctic sea ice minima and the projected probabilities of an ice-free Arctic\n---\nstart to become significant earlier than under `ssp3_7_0`. In this experiment (`ssp5_8_5`), even May and January have 20% probability in the last 5 years or so.\n\nIn summary, the heat maps above show that as the warming increases, we can clearly see that the overall probability of being ice-free is increasing, and the length of the season where there is high probability is also increasing from just September (`ssp1_2_6`) to the 6-month period from June to December (`ssp5_8_5`).\nThis can also be seen in the plots below, where we compare the experiments more directly by plotting the monthly probabilities for a selection of years.\n\nIn the years before 2060, the probabilities are not ordered by warming strength, but around 2060 `ssp3_7_0` reaches the `ssp2_4_5` level, and after that year the probabilities' order does correspond to warming strength.\nIn 2040, the maximum probability ranges from about 20% to 40%; in 2050 and 2060 it ranges from about 40% to 60%; in 2070, 2080 and 2090 it ranges from about 45% to about 85%. In the latter three years the main effect is the widening of the season with the highest probabilities.\n\nComparing to other authors estimates, the [SIMIP community (2020)](https://doi.org/10.1029/2019GL086749) found that nearly all of their selected models projected ice-free conditions by 2050, regardless of warming scenario. [X. Zhao et al. (2022)](https://doi.org/10.1088/1748-9326/ac9d4d) agree with this - they projected the first ice-free year to be $2049\\pm 12$ under `ssp2_4_5` and $2043\\pm 9$ under `ssp5_8_5`, giving a similar range to the [SIMIP community (2020)](https://doi.org/10.1029/2019GL086749). While our 2050 probability of just under 60% for `ssp5_8_5` is quite high, 40% for `ssp1_2_6` is quite a lot lower. This difference possibly reflects the effect both of using a different subset of models and of using the full ensemble set provided for each model, as [SIMIP community (2020)](https://doi.org/10.1029/2019GL086749) did (instead of just using the ensemble mean for each model).\n\n(section-3.2)="} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q04__a72571f96e85", "report_id": "climate_projections-cmip6_climate-impact-indicators_q04", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q04", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Projections of future ice-free periods for the Arctic and Antarctic > Analysis and results > 3. Results > 3.2 Arctic sea ice maxima", "title": "Projections of future ice-free periods for the Arctic and Antarctic", "chunk_index": 19, "token_count": 362, "text_raw": "Like the Arctic minimum in sea ice area, the maximum sea ice area is constant until about 1990 the historical CMIP6 is relatively constant, before starting to drop rapidly. However, unlike with the minimum area, there is a clear bias in this variable, with both the ERA5 reanalysis and the CMIP6 ensembles overestimating it by about $1.5\\times10^6$km$^2$. As we come to the start of the projection period (2015-2100), the yearly maxima continues to drop. Unlike with the minima, the maximum sea ice area shows a clear response to increased warming - as the warming increases the median Arctic maximum area drops. The experiment with the most warming (`ssp5_8_5`) also shows a very large spread in this variable by 2100.\n\nLooking at the maps below, we can see that there is not too much difference between the maps in March before about 2055. After this however, the `ssp5_8_5` experiment especially starts to have reduced concentration in the central Arctic, and also in the Hudson Bay. The `ssp3_7_0` experiment also shows this pattern to some extent. The drop in maximum sea ice area with time can also be seen, e.g. in the Bering and Labrador Seas.\n\n(section-3.3)=", "text_with_prefix": "EQC Quality Assessment: \"Projections of future ice-free periods for the Arctic and Antarctic\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q04 | Category: Climate_Projections\nSection: Projections of future ice-free periods for the Arctic and Antarctic > Analysis and results > 3. Results > 3.2 Arctic sea ice maxima\n---\nLike the Arctic minimum in sea ice area, the maximum sea ice area is constant until about 1990 the historical CMIP6 is relatively constant, before starting to drop rapidly. However, unlike with the minimum area, there is a clear bias in this variable, with both the ERA5 reanalysis and the CMIP6 ensembles overestimating it by about $1.5\\times10^6$km$^2$. As we come to the start of the projection period (2015-2100), the yearly maxima continues to drop. Unlike with the minima, the maximum sea ice area shows a clear response to increased warming - as the warming increases the median Arctic maximum area drops. The experiment with the most warming (`ssp5_8_5`) also shows a very large spread in this variable by 2100.\n\nLooking at the maps below, we can see that there is not too much difference between the maps in March before about 2055. After this however, the `ssp5_8_5` experiment especially starts to have reduced concentration in the central Arctic, and also in the Hudson Bay. The `ssp3_7_0` experiment also shows this pattern to some extent. The drop in maximum sea ice area with time can also be seen, e.g. in the Bering and Labrador Seas.\n\n(section-3.3)="} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q04__138a1f3acfb2", "report_id": "climate_projections-cmip6_climate-impact-indicators_q04", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q04", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Projections of future ice-free periods for the Arctic and Antarctic > Analysis and results > 3. Results > 3.3 Antarctic sea ice minima and the projected probabilities of being ice-free around Antarctica", "title": "Projections of future ice-free periods for the Arctic and Antarctic", "chunk_index": 20, "token_count": 800, "text_raw": "The CMIP6 models show a large bias in the Antarctic minimum sea ice area compared to the satellite data, underestimating it by about 1-$2\\times10^6$km$^2$. The satellites seem to show a relatively constant minimum, while there is a steady drop predicted by the CMIP6 models in this time period when observations are available (1979-2017).\n\nKeeping in mind the large bias in the CMIP6 models in this area ([Roach et al., 2023](https://doi.org/10.1029/2019GL086729); [CDS assessment of historical sea ice extent](https://ecmwf-projects.github.io/c3s2-eqc-quality-assessment/Climate_Projections/CMIP6/climate_projections-cmip6_model-performance_q02.html)), the modelled Antarctic minimum in sea ice area shows a clear effect of increased warming. After about 1960 the CMIP6 ensemble median starts to slowly drop, and then after 2015 the rate depends on the warming scenario. The median only reaches our ice-free limit of $0.25\\times10^6$ km$^2$ under the `ssp3_7_0` and `ssp5_8_5` scenarios, but the `ssp2_4_5` median is close to it. Under the `ssp1_2_6` scenario the area stays about $0.4\\times 10^6$ km$^2$ from about 2040. The spread of the models is quite high.\n\nA clear effect of climate change can be see in the maps below of the Antarctic sea ice concentration for March. Apart from the `ssp1_2_6` experiment, there is very little ice outside the Weddell and Ross Sea, and the `ssp5_8_5` has mostly lost its ice in the Ross Sea by 2075.\n\nThe heat maps below show the proportion of models dropping below $0.25\\times10^6$ km$^2$ for a given month and year, and below them we compare the monthly probabilities more directly for all the experiments for a selection of years.\n\nAs with the Arctic, increased warming increases the overall probability and also the length of the period when there is high probability. However, unlike the Arctic, the Antarctic region has lower chance of being ice-free, only getting consistently over 70% only in the warmest scenario `ssp5_8_5`. The chances in February reach 50% around 2040 and 80% around 2088 in that experiment.\nIn March and January the probability also reaches 60% (in 2070 and 2095 respectively), and even 70% (from about 2080 in March).\n\nThe `ssp_2_4_5` and `ssp_3_7_0` experiments show similar probabilities, consistently reaching 60% in February in the last two decades or so. In March they get to 50% from about 2080 (a few years earlier under `ssp_2_4_5`).\n\nThe `ssp1_2_6` experiment has probabilities of being ice-free oscillating between 20%-40% in February, and between 20%-35% in March.\n\n(section-3.4)=", "text_with_prefix": "EQC Quality Assessment: \"Projections of future ice-free periods for the Arctic and Antarctic\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q04 | Category: Climate_Projections\nSection: Projections of future ice-free periods for the Arctic and Antarctic > Analysis and results > 3. Results > 3.3 Antarctic sea ice minima and the projected probabilities of being ice-free around Antarctica\n---\nThe CMIP6 models show a large bias in the Antarctic minimum sea ice area compared to the satellite data, underestimating it by about 1-$2\\times10^6$km$^2$. The satellites seem to show a relatively constant minimum, while there is a steady drop predicted by the CMIP6 models in this time period when observations are available (1979-2017).\n\nKeeping in mind the large bias in the CMIP6 models in this area ([Roach et al., 2023](https://doi.org/10.1029/2019GL086729); [CDS assessment of historical sea ice extent](https://ecmwf-projects.github.io/c3s2-eqc-quality-assessment/Climate_Projections/CMIP6/climate_projections-cmip6_model-performance_q02.html)), the modelled Antarctic minimum in sea ice area shows a clear effect of increased warming. After about 1960 the CMIP6 ensemble median starts to slowly drop, and then after 2015 the rate depends on the warming scenario. The median only reaches our ice-free limit of $0.25\\times10^6$ km$^2$ under the `ssp3_7_0` and `ssp5_8_5` scenarios, but the `ssp2_4_5` median is close to it. Under the `ssp1_2_6` scenario the area stays about $0.4\\times 10^6$ km$^2$ from about 2040. The spread of the models is quite high.\n\nA clear effect of climate change can be see in the maps below of the Antarctic sea ice concentration for March. Apart from the `ssp1_2_6` experiment, there is very little ice outside the Weddell and Ross Sea, and the `ssp5_8_5` has mostly lost its ice in the Ross Sea by 2075.\n\nThe heat maps below show the proportion of models dropping below $0.25\\times10^6$ km$^2$ for a given month and year, and below them we compare the monthly probabilities more directly for all the experiments for a selection of years.\n\nAs with the Arctic, increased warming increases the overall probability and also the length of the period when there is high probability. However, unlike the Arctic, the Antarctic region has lower chance of being ice-free, only getting consistently over 70% only in the warmest scenario `ssp5_8_5`. The chances in February reach 50% around 2040 and 80% around 2088 in that experiment.\nIn March and January the probability also reaches 60% (in 2070 and 2095 respectively), and even 70% (from about 2080 in March).\n\nThe `ssp_2_4_5` and `ssp_3_7_0` experiments show similar probabilities, consistently reaching 60% in February in the last two decades or so. In March they get to 50% from about 2080 (a few years earlier under `ssp_2_4_5`).\n\nThe `ssp1_2_6` experiment has probabilities of being ice-free oscillating between 20%-40% in February, and between 20%-35% in March.\n\n(section-3.4)="} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q04__f3e0abb5e786", "report_id": "climate_projections-cmip6_climate-impact-indicators_q04", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q04", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Projections of future ice-free periods for the Arctic and Antarctic > Analysis and results > 3. Results > 3.4 Antarctic sea ice maxima", "title": "Projections of future ice-free periods for the Arctic and Antarctic", "chunk_index": 21, "token_count": 304, "text_raw": "The projected Antarctic sea ice maxima behave in a similar way to the Arctic maxima with time and warming, with a similar relative drop (30%-50%). The CMIP6 models again show a large bias in the Antarctic sea ice area compared to the satellite data, although the difference is less at the start of the start of the satellite period (1979). However, since the satellites show a slight increase in minimum area, while there is a steady drop predicted by the CMIP6 models, the difference increases with time, reaching about $2\\times10^6$km$^2$ by 2015.\n\nLess effect can be seen in the maps below of the ensemble mean Antarctic sea ice concentration in September than in March. For the `ssp1_2_6` experiment, there is little difference between the last time period and the first, but there small decreases in area, which are relatively even around the continent, visible for the other experiments. In the last two time periods, there is also a slight decrease in area as the warming increases.\n\n(section-3.5)=", "text_with_prefix": "EQC Quality Assessment: \"Projections of future ice-free periods for the Arctic and Antarctic\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q04 | Category: Climate_Projections\nSection: Projections of future ice-free periods for the Arctic and Antarctic > Analysis and results > 3. Results > 3.4 Antarctic sea ice maxima\n---\nThe projected Antarctic sea ice maxima behave in a similar way to the Arctic maxima with time and warming, with a similar relative drop (30%-50%). The CMIP6 models again show a large bias in the Antarctic sea ice area compared to the satellite data, although the difference is less at the start of the start of the satellite period (1979). However, since the satellites show a slight increase in minimum area, while there is a steady drop predicted by the CMIP6 models, the difference increases with time, reaching about $2\\times10^6$km$^2$ by 2015.\n\nLess effect can be seen in the maps below of the ensemble mean Antarctic sea ice concentration in September than in March. For the `ssp1_2_6` experiment, there is little difference between the last time period and the first, but there small decreases in area, which are relatively even around the continent, visible for the other experiments. In the last two time periods, there is also a slight decrease in area as the warming increases.\n\n(section-3.5)="} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q04__e09c05750051", "report_id": "climate_projections-cmip6_climate-impact-indicators_q04", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q04", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Projections of future ice-free periods for the Arctic and Antarctic > Analysis and results > 3. Results > 3.5 Projected access to Arctic shipping routes", "title": "Projections of future ice-free periods for the Arctic and Antarctic", "chunk_index": 22, "token_count": 1014, "text_raw": "We consider two Arctic shipping routes, the Transpolar Shipping Route (TSR) and the Northern Sea Route (NSR). We do not consider the Northwest Passage (NWP) since the climate models generally don't have enough resolution to represent the Canadian archipelago. The figure below shows our definition of these two routes.\n\n(section-3.5.1)=\n##### 3.5.1 Northern sea route\nIn the plots below we can see the minimum sea ice area in the NSR as stable at about $10^6 $km$^2$ until about 1970 when it starts dropping. The historical area is fairly close to the observations in this area, athough they are initially a bit low and a bit high after about 2005. Most experiments' median have dropped below the ice free level by 2040.\n\nThe heat maps below show the proportion of models dropping below $0.14\\times10^6$ km$^2$ for a given month and year, and below them we compare the monthly probabilities more directly for all the experiments for a selection of years.\n\nWe can see even in the experiment with the least warming, there is a high chance of the Northern Sea Route (NSR) being clear in September, especially after about 2040, when the probability reaches about 70%. After this the probabilty fluctuates between about 70% and 80%. August has quite high probabilities too, getting to around 70% around 2060. The probability in October reaches about 50%, while it reaches about 20% in July. The chances increase as the experiments change to produce more warming, and the chances of having the NSR clear in November and July start to increase as well. In November and July `ssp3_7_0` has about 60-70% (sometimes higher) chance after 2082 or 2083. `ssp5_8_5` has 80% chance in November and 70-80% chance in July, both from about 2075. In July `ssp3_7_0` has about 60-70% chance after 2080, while `ssp5_8_5` has 80% chance from about 2075. The experiment `ssp5_8_5` even has 40-50% chance of it being clear in December in later years, and 80-100% of the models are projecting the NSR to be free in August, September, and October from about 2055 under this scenario.\n\nThe probabilities shown below for September in 2040 and 2050 are only slightly lower than the results of [Chen et al. (2023)](https://doi.org/10.1016/j.accre.2023.11.011) (see Figures 6a and 6b), who estimated the probability (averaged over 2045-2055) of the Sannikov Strait being passable by open water vessels (without ice strengthening) of 60% under `ssp1_2_6` and 100% under `ssp5_8_5`. The size of their navigability window (about July-November) is also similar to ours. This strait (along with the Dmitry Laptev Strait which had slightly lower passability than the Sannikov Strait) is key for the overall passability of the NSR overall. This consistency is reassuring given the quite different methods of calculating the probabilities and the different subsets of the climate models used.\n\n(section-3.5.2)=\n##### 3.5.2 Transpolar shipping route\nIn the plots below we can see the minimum sea ice area in the TSR as stable at about $1.5 \\times 10^6 $km$^2$ until about 1970 when it starts dropping. The area from historical experiment is fairly close to the observations in this area. Most experiments' median have dropped below the ice free level by 2050, with the exception of the `ssp1_2_6` experiment, which is not consistently beneath it until about 2070. This experiment has quite a large spread compared to the others.\n\nThe heat maps below show the proportion of models dropping below $0.21\\times10^6$ km$^2$ for a given month and year, and below them we compare the monthly probabilities more directly for all the experiments for a selection of years.", "text_with_prefix": "EQC Quality Assessment: \"Projections of future ice-free periods for the Arctic and Antarctic\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q04 | Category: Climate_Projections\nSection: Projections of future ice-free periods for the Arctic and Antarctic > Analysis and results > 3. Results > 3.5 Projected access to Arctic shipping routes\n---\nWe consider two Arctic shipping routes, the Transpolar Shipping Route (TSR) and the Northern Sea Route (NSR). We do not consider the Northwest Passage (NWP) since the climate models generally don't have enough resolution to represent the Canadian archipelago. The figure below shows our definition of these two routes.\n\n(section-3.5.1)=\n##### 3.5.1 Northern sea route\nIn the plots below we can see the minimum sea ice area in the NSR as stable at about $10^6 $km$^2$ until about 1970 when it starts dropping. The historical area is fairly close to the observations in this area, athough they are initially a bit low and a bit high after about 2005. Most experiments' median have dropped below the ice free level by 2040.\n\nThe heat maps below show the proportion of models dropping below $0.14\\times10^6$ km$^2$ for a given month and year, and below them we compare the monthly probabilities more directly for all the experiments for a selection of years.\n\nWe can see even in the experiment with the least warming, there is a high chance of the Northern Sea Route (NSR) being clear in September, especially after about 2040, when the probability reaches about 70%. After this the probabilty fluctuates between about 70% and 80%. August has quite high probabilities too, getting to around 70% around 2060. The probability in October reaches about 50%, while it reaches about 20% in July. The chances increase as the experiments change to produce more warming, and the chances of having the NSR clear in November and July start to increase as well. In November and July `ssp3_7_0` has about 60-70% (sometimes higher) chance after 2082 or 2083. `ssp5_8_5` has 80% chance in November and 70-80% chance in July, both from about 2075. In July `ssp3_7_0` has about 60-70% chance after 2080, while `ssp5_8_5` has 80% chance from about 2075. The experiment `ssp5_8_5` even has 40-50% chance of it being clear in December in later years, and 80-100% of the models are projecting the NSR to be free in August, September, and October from about 2055 under this scenario.\n\nThe probabilities shown below for September in 2040 and 2050 are only slightly lower than the results of [Chen et al. (2023)](https://doi.org/10.1016/j.accre.2023.11.011) (see Figures 6a and 6b), who estimated the probability (averaged over 2045-2055) of the Sannikov Strait being passable by open water vessels (without ice strengthening) of 60% under `ssp1_2_6` and 100% under `ssp5_8_5`. The size of their navigability window (about July-November) is also similar to ours. This strait (along with the Dmitry Laptev Strait which had slightly lower passability than the Sannikov Strait) is key for the overall passability of the NSR overall. This consistency is reassuring given the quite different methods of calculating the probabilities and the different subsets of the climate models used.\n\n(section-3.5.2)=\n##### 3.5.2 Transpolar shipping route\nIn the plots below we can see the minimum sea ice area in the TSR as stable at about $1.5 \\times 10^6 $km$^2$ until about 1970 when it starts dropping. The area from historical experiment is fairly close to the observations in this area. Most experiments' median have dropped below the ice free level by 2050, with the exception of the `ssp1_2_6` experiment, which is not consistently beneath it until about 2070. This experiment has quite a large spread compared to the others.\n\nThe heat maps below show the proportion of models dropping below $0.21\\times10^6$ km$^2$ for a given month and year, and below them we compare the monthly probabilities more directly for all the experiments for a selection of years."} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q04__0645b19cea2b", "report_id": "climate_projections-cmip6_climate-impact-indicators_q04", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q04", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Projections of future ice-free periods for the Arctic and Antarctic > Analysis and results > 3. Results > 3.5 Projected access to Arctic shipping routes", "title": "Projections of future ice-free periods for the Arctic and Antarctic", "chunk_index": 23, "token_count": 499, "text_raw": "of models dropping below $0.21\\times10^6$ km$^2$ for a given month and year, and below them we compare the monthly probabilities more directly for all the experiments for a selection of years.\n\nThe Transpolar shipping route (TSR) has slightly lower chances of being ice-free in the months August-October than the NSR but somewhat surprisingly it is a bit higher in other months, with it having some chance of it being clear in winter months in the experiments with higher warming. \nFor example, under `ssp5_8_5` there is a 35% chance of the TSR being clear from February to May after 2089 (after 2085 for February), and there is more than 60% chance of it being clear in December from about 2088, and the same chance in January from 2095.\nUnder `ssp3_7_0` there is more than 20% chance of the TSR being ice-free in December, and also in January from 2090.\n\nThe maps below plot the ensemble mean sea ice concentration for December for a series of time intervals between 2015 and 2095. In the 2075-2095 period, the `ssp5_8_5` has a distinct gap where we have defined our TSR, while there is still ice off the Russian coast blocking the NSR. The effect is also present to a lesser extent for the `ssp2_4_5` and `ssp3_7_0` xperiments, so perhaps ice-strengthened vessels could still traverse the TSR under these scenarios also. This is consistent with the results of [Min et al (2022)](https://doi.org/10.1029/2022GL099157) (see Figure 3), who found that from 2080 to 2100, the TSR would be a more robust option than the NSR or the North-West Passage for open water vessels.", "text_with_prefix": "EQC Quality Assessment: \"Projections of future ice-free periods for the Arctic and Antarctic\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q04 | Category: Climate_Projections\nSection: Projections of future ice-free periods for the Arctic and Antarctic > Analysis and results > 3. Results > 3.5 Projected access to Arctic shipping routes\n---\nof models dropping below $0.21\\times10^6$ km$^2$ for a given month and year, and below them we compare the monthly probabilities more directly for all the experiments for a selection of years.\n\nThe Transpolar shipping route (TSR) has slightly lower chances of being ice-free in the months August-October than the NSR but somewhat surprisingly it is a bit higher in other months, with it having some chance of it being clear in winter months in the experiments with higher warming. \nFor example, under `ssp5_8_5` there is a 35% chance of the TSR being clear from February to May after 2089 (after 2085 for February), and there is more than 60% chance of it being clear in December from about 2088, and the same chance in January from 2095.\nUnder `ssp3_7_0` there is more than 20% chance of the TSR being ice-free in December, and also in January from 2090.\n\nThe maps below plot the ensemble mean sea ice concentration for December for a series of time intervals between 2015 and 2095. In the 2075-2095 period, the `ssp5_8_5` has a distinct gap where we have defined our TSR, while there is still ice off the Russian coast blocking the NSR. The effect is also present to a lesser extent for the `ssp2_4_5` and `ssp3_7_0` xperiments, so perhaps ice-strengthened vessels could still traverse the TSR under these scenarios also. This is consistent with the results of [Min et al (2022)](https://doi.org/10.1029/2022GL099157) (see Figure 3), who found that from 2080 to 2100, the TSR would be a more robust option than the NSR or the North-West Passage for open water vessels."} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q04__72bede2ee6ec", "report_id": "climate_projections-cmip6_climate-impact-indicators_q04", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q04", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Projections of future ice-free periods for the Arctic and Antarctic > ℹ️ If you want to know more > Key resources", "title": "Projections of future ice-free periods for the Arctic and Antarctic", "chunk_index": 24, "token_count": 375, "text_raw": "Introductory sea ice materials:\n- [Role of sea ice in the climate](https://marine.copernicus.eu/explainers/why-ocean-important/sea-ice)\n- [Sea ice as an indicator of climate change](https://climate.copernicus.eu/climate-indicators/sea-ice#)\n- [Observing sea ice with satellites](https://www.metoffice.gov.uk/research/climate/cryosphere-oceans/sea-ice/measure)\n\nIntroductory CMIP6 materials:\n- [A short introduction to CMIP and CMIP6](https://www.wcrp-climate.org/wgcm-cmip/cmip-video)\n- [CMIP6: the next generation of climate models explained](https://www.carbonbrief.org/cmip6-the-next-generation-of-climate-models-explained/)\n\nCode libraries used:\n- [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control) developed by [B-Open](https://www.bopen.eu/)", "text_with_prefix": "EQC Quality Assessment: \"Projections of future ice-free periods for the Arctic and Antarctic\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q04 | Category: Climate_Projections\nSection: Projections of future ice-free periods for the Arctic and Antarctic > ℹ️ If you want to know more > Key resources\n---\nIntroductory sea ice materials:\n- [Role of sea ice in the climate](https://marine.copernicus.eu/explainers/why-ocean-important/sea-ice)\n- [Sea ice as an indicator of climate change](https://climate.copernicus.eu/climate-indicators/sea-ice#)\n- [Observing sea ice with satellites](https://www.metoffice.gov.uk/research/climate/cryosphere-oceans/sea-ice/measure)\n\nIntroductory CMIP6 materials:\n- [A short introduction to CMIP and CMIP6](https://www.wcrp-climate.org/wgcm-cmip/cmip-video)\n- [CMIP6: the next generation of climate models explained](https://www.carbonbrief.org/cmip6-the-next-generation-of-climate-models-explained/)\n\nCode libraries used:\n- [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control) developed by [B-Open](https://www.bopen.eu/)"} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q04__9e9c74cdaba0", "report_id": "climate_projections-cmip6_climate-impact-indicators_q04", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q04", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Projections of future ice-free periods for the Arctic and Antarctic > ℹ️ If you want to know more > References", "title": "Projections of future ice-free periods for the Arctic and Antarctic", "chunk_index": 25, "token_count": 1046, "text_raw": "1. Chen, J. L., Kang, S. C., Wu, A. D., Chen, L. H., & Li, Y. W. (2023). Accessibility in key areas of the Arctic in the 21st mid-century. Advances in Climate Change Research, 14(6), 896-903, [ https://doi.org/10.1016/j.accre.2023.11.011 ](https://doi.org/10.1016/j.accre.2023.11.011).\n\n1. Chen, J., Kang, S., You, Q., Zhang, Y., & Du, W. (2022). Projected changes in sea ice and the navigability of the Arctic Passages under global warming of 2℃ and 3℃. Anthropocene, 40, 100349, [ https://doi.org/10.1016/j.ancene.2022.100349 ](https://doi.org/10.1016/j.ancene.2022.100349).\n\n1. Chen, J. L., Kang, S. C., Guo, J. M., Xu, M., & Zhang, Z. M. (2021). Variation of sea ice and perspectives of the Northwest Passage in the Arctic Ocean. Advances in Climate Change Research, 12(4), 447-455, [ https://doi.org/10.1016/j.accre.2021.02.002 ](https://doi.org/10.1016/j.accre.2021.02.002).\n\n1. Davy, R., and S. Outten (2020). The Arctic Surface Climate in CMIP6: Status and Developments since CMIP5. J. Climate, 33, 8047–8068, [ https://doi.org/10.1175/JCLI-D-19-0990.1 ](https://doi.org/10.1175/JCLI-D-19-0990.1).\n\n1. Heuzé, C. & Jahn, A. (2024). The first ice-free day in the Arctic Ocean could occur before 2030. Nat. Commun. 15, 10101, [ https://doi.org/10.1038/s41467-024-54508-3 ](https://doi.org/10.1038/s41467-024-54508-3).\n\n1. Jahn, A., Kay, J. E., Holland, M. M., & Hall, D. M. (2016). How predictable is the timing of a summer ice‐free Arctic? Geophysical Research Letters, 43(17), 9113-9120 , [ https://doi.org/10.1002/2016GL070067 ](https://doi.org/10.1002/2016GL070067).\n\n1. Jahn, A., Holland, M.M. & Kay, J.E. (2024). Projections of an ice-free Arctic Ocean. Nat. Rev. Earth. Environ. 5, 164–176, [ https://doi.org/10.1038/s43017-023-00515-9 ](https://doi.org/10.1038/s43017-023-00515-9).\n\n1. Li, X., Stephenson, S. R., Lynch, A. H., Goldstein, M. A., Bailey, D. A., & Veland, S. (2021). Arctic shipping guidance from the CMIP6 ensemble on operational and infrastructural timescales. Climatic Change, 167, 1-19, [ https://doi.org/10.1007/s10584-021-03172-3 ](https://doi.org/10.1007/s10584-021-03172-3).\n\n1. Melia, N., Haines, K., and Hawkins, E. (2016). Sea ice decline and 21st century trans-Arctic shipping routes, Geophys. Res. Lett., 43, 9720–9728, [ https://doi.org/10.1002/2016GL069315 ](https://doi.org/10.1002/2016GL069315).", "text_with_prefix": "EQC Quality Assessment: \"Projections of future ice-free periods for the Arctic and Antarctic\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q04 | Category: Climate_Projections\nSection: Projections of future ice-free periods for the Arctic and Antarctic > ℹ️ If you want to know more > References\n---\n1. Chen, J. L., Kang, S. C., Wu, A. D., Chen, L. H., & Li, Y. W. (2023). Accessibility in key areas of the Arctic in the 21st mid-century. Advances in Climate Change Research, 14(6), 896-903, [ https://doi.org/10.1016/j.accre.2023.11.011 ](https://doi.org/10.1016/j.accre.2023.11.011).\n\n1. Chen, J., Kang, S., You, Q., Zhang, Y., & Du, W. (2022). Projected changes in sea ice and the navigability of the Arctic Passages under global warming of 2℃ and 3℃. Anthropocene, 40, 100349, [ https://doi.org/10.1016/j.ancene.2022.100349 ](https://doi.org/10.1016/j.ancene.2022.100349).\n\n1. Chen, J. L., Kang, S. C., Guo, J. M., Xu, M., & Zhang, Z. M. (2021). Variation of sea ice and perspectives of the Northwest Passage in the Arctic Ocean. Advances in Climate Change Research, 12(4), 447-455, [ https://doi.org/10.1016/j.accre.2021.02.002 ](https://doi.org/10.1016/j.accre.2021.02.002).\n\n1. Davy, R., and S. Outten (2020). The Arctic Surface Climate in CMIP6: Status and Developments since CMIP5. J. Climate, 33, 8047–8068, [ https://doi.org/10.1175/JCLI-D-19-0990.1 ](https://doi.org/10.1175/JCLI-D-19-0990.1).\n\n1. Heuzé, C. & Jahn, A. (2024). The first ice-free day in the Arctic Ocean could occur before 2030. Nat. Commun. 15, 10101, [ https://doi.org/10.1038/s41467-024-54508-3 ](https://doi.org/10.1038/s41467-024-54508-3).\n\n1. Jahn, A., Kay, J. E., Holland, M. M., & Hall, D. M. (2016). How predictable is the timing of a summer ice‐free Arctic? Geophysical Research Letters, 43(17), 9113-9120 , [ https://doi.org/10.1002/2016GL070067 ](https://doi.org/10.1002/2016GL070067).\n\n1. Jahn, A., Holland, M.M. & Kay, J.E. (2024). Projections of an ice-free Arctic Ocean. Nat. Rev. Earth. Environ. 5, 164–176, [ https://doi.org/10.1038/s43017-023-00515-9 ](https://doi.org/10.1038/s43017-023-00515-9).\n\n1. Li, X., Stephenson, S. R., Lynch, A. H., Goldstein, M. A., Bailey, D. A., & Veland, S. (2021). Arctic shipping guidance from the CMIP6 ensemble on operational and infrastructural timescales. Climatic Change, 167, 1-19, [ https://doi.org/10.1007/s10584-021-03172-3 ](https://doi.org/10.1007/s10584-021-03172-3).\n\n1. Melia, N., Haines, K., and Hawkins, E. (2016). Sea ice decline and 21st century trans-Arctic shipping routes, Geophys. Res. Lett., 43, 9720–9728, [ https://doi.org/10.1002/2016GL069315 ](https://doi.org/10.1002/2016GL069315)."} {"chunk_id": "climate_projections-cmip6_climate-impact-indicators_q04__10dbcb98f96f", "report_id": "climate_projections-cmip6_climate-impact-indicators_q04", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-impact-indicators_q04", "aspect_base": "climate-impact-indicators", "category": "Climate_Projections", "match_confidence": "exact", "section": "Projections of future ice-free periods for the Arctic and Antarctic > ℹ️ If you want to know more > References", "title": "Projections of future ice-free periods for the Arctic and Antarctic", "chunk_index": 26, "token_count": 1076, "text_raw": "0–9728, [ https://doi.org/10.1002/2016GL069315 ](https://doi.org/10.1002/2016GL069315).\n\n1. Min, C., Yang, Q., Chen, D., Yang, Y., Zhou, X., Shu, Q., & Liu, J. (2022). The emerging Arctic shipping corridors. Geophysical Research Letters, 49(10), e2022GL099157. [ https://doi.org/10.1029/2022GL099157 ](https://doi.org/10.1029/2022GL099157).\n\n1. Pan, R., Shu, Q., Wang, Q., Wang, S., Song, Z., He, Y., & Qiao, F. (2023). Future Arctic climate change in CMIP6 strikingly intensified by NEMO‐family climate models. Geophysical Research Letters, 50(4), e2022GL102077, [ https://doi.org/10.1029/2022GL102077 ](https://doi.org/10.1029/2022GL102077).\n\n1. Roach, L. A., J. Dörr, C. R. Holmes, F. Massonnet, E. W. Blockley, D. Notz, T. Rackow, M. N. Raphael, S. P. O'Farrell, D. A. Bailey, and C. M. Bitz (2020). Antarctic sea ice area in CMIP6. Geophysical Research Letters, 47, e2019GL086729, [ https://doi.org/10.1029/2019GL086729 ](https://doi.org/10.1029/2019GL086729).\n\n1. SIMIP Community (2020). Arctic sea ice in CMIP6. Geophysical Research Letters, 47, e2019GL086749, [ https://doi.org/10.1029/2019GL086749 ](https://doi.org/10.1029/2019GL086749).\n\n1. Smith, L. C., & Stephenson, S. R. (2013). New Trans-Arctic shipping routes navigable by midcentury. Proceedings of the National Academy of Sciences, 110(13), E1191-E1195, [ https://doi.org/10.1073/pnas.1214212110 ](https://doi.org/10.1073/pnas.1214212110).\n\n1. Wei, T., Yan, Q., Qi, W., Ding, M., & Wang, C. (2020). Projections of Arctic sea ice conditions and shipping routes in the twenty-first century using CMIP6 forcing scenarios. Environmental Research Letters, 15(10), 104079, [ https://doi.org/10.1175/JCLI-D-19-0990.1 ](https://doi.org/10.1175/JCLI-D-19-0990.1).\n\n1. Wang, B., Zhou, X., Ding, Q., & Liu, J. (2021). Increasing confidence in projecting the Arctic ice-free year with emergent constraints. Environmental Research Letters, 16(9), 094016, [ https://doi.org/10.1088/1748-9326/ac0b17 ](https://doi.org/10.1088/1748-9326/ac0b17).\n\n1. Zhao, J., He, S., Wang, H., & Li, F. (2022). Constraining CMIP6 Projections of an ice‐free Arctic Using a weighting scheme. Earth's Future, 10(10), e2022EF002708, [ https://doi.org/10.1029/2022EF002708 ](https://doi.org/10.1029/2022EF002708).\n\n1. Zhou, X., Wang, B., & Huang, F. (2022). Evaluating sea ice thickness simulation is critical for projecting a summer ice-free Arctic Ocean. Environmental Research Letters, 17(11), 114033, [ https://doi.org/10.1088/1748-9326/ac9d4d ](https://doi.org/10.1088/1748-9326/ac9d4d).", "text_with_prefix": "EQC Quality Assessment: \"Projections of future ice-free periods for the Arctic and Antarctic\"\nDataset: projections-cmip6 [CDS]\nAspect: climate-impact-indicators_q04 | Category: Climate_Projections\nSection: Projections of future ice-free periods for the Arctic and Antarctic > ℹ️ If you want to know more > References\n---\n0–9728, [ https://doi.org/10.1002/2016GL069315 ](https://doi.org/10.1002/2016GL069315).\n\n1. Min, C., Yang, Q., Chen, D., Yang, Y., Zhou, X., Shu, Q., & Liu, J. (2022). The emerging Arctic shipping corridors. Geophysical Research Letters, 49(10), e2022GL099157. [ https://doi.org/10.1029/2022GL099157 ](https://doi.org/10.1029/2022GL099157).\n\n1. Pan, R., Shu, Q., Wang, Q., Wang, S., Song, Z., He, Y., & Qiao, F. (2023). Future Arctic climate change in CMIP6 strikingly intensified by NEMO‐family climate models. Geophysical Research Letters, 50(4), e2022GL102077, [ https://doi.org/10.1029/2022GL102077 ](https://doi.org/10.1029/2022GL102077).\n\n1. Roach, L. A., J. Dörr, C. R. Holmes, F. Massonnet, E. W. Blockley, D. Notz, T. Rackow, M. N. Raphael, S. P. O'Farrell, D. A. Bailey, and C. M. Bitz (2020). Antarctic sea ice area in CMIP6. Geophysical Research Letters, 47, e2019GL086729, [ https://doi.org/10.1029/2019GL086729 ](https://doi.org/10.1029/2019GL086729).\n\n1. SIMIP Community (2020). Arctic sea ice in CMIP6. Geophysical Research Letters, 47, e2019GL086749, [ https://doi.org/10.1029/2019GL086749 ](https://doi.org/10.1029/2019GL086749).\n\n1. Smith, L. C., & Stephenson, S. R. (2013). New Trans-Arctic shipping routes navigable by midcentury. Proceedings of the National Academy of Sciences, 110(13), E1191-E1195, [ https://doi.org/10.1073/pnas.1214212110 ](https://doi.org/10.1073/pnas.1214212110).\n\n1. Wei, T., Yan, Q., Qi, W., Ding, M., & Wang, C. (2020). Projections of Arctic sea ice conditions and shipping routes in the twenty-first century using CMIP6 forcing scenarios. Environmental Research Letters, 15(10), 104079, [ https://doi.org/10.1175/JCLI-D-19-0990.1 ](https://doi.org/10.1175/JCLI-D-19-0990.1).\n\n1. Wang, B., Zhou, X., Ding, Q., & Liu, J. (2021). Increasing confidence in projecting the Arctic ice-free year with emergent constraints. Environmental Research Letters, 16(9), 094016, [ https://doi.org/10.1088/1748-9326/ac0b17 ](https://doi.org/10.1088/1748-9326/ac0b17).\n\n1. Zhao, J., He, S., Wang, H., & Li, F. (2022). Constraining CMIP6 Projections of an ice‐free Arctic Using a weighting scheme. Earth's Future, 10(10), e2022EF002708, [ https://doi.org/10.1029/2022EF002708 ](https://doi.org/10.1029/2022EF002708).\n\n1. Zhou, X., Wang, B., & Huang, F. (2022). Evaluating sea ice thickness simulation is critical for projecting a summer ice-free Arctic Ocean. Environmental Research Letters, 17(11), 114033, [ https://doi.org/10.1088/1748-9326/ac9d4d ](https://doi.org/10.1088/1748-9326/ac9d4d)."} {"chunk_id": "climate_projections-cmip6_model-performance_q02__40ec278b07ae", "report_id": "climate_projections-cmip6_model-performance_q02", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "model-performance_q02", "aspect_base": "model-performance", "category": "Climate_Projections", "match_confidence": "exact", "section": "Historical accuracy of sea ice extent in the CMIP6 experiments", "title": "Historical accuracy of sea ice extent in the CMIP6 experiments", "chunk_index": 0, "token_count": 92, "text_raw": "Production date: 2024-05-31\n\nProduced by: Timothy Williams, Nansen Environmental and Remote Sensing Center (NERSC)", "text_with_prefix": "EQC Quality Assessment: \"Historical accuracy of sea ice extent in the CMIP6 experiments\"\nDataset: projections-cmip6 [CDS]\nAspect: model-performance_q02 | Category: Climate_Projections\nSection: Historical accuracy of sea ice extent in the CMIP6 experiments\n---\nProduction date: 2024-05-31\n\nProduced by: Timothy Williams, Nansen Environmental and Remote Sensing Center (NERSC)"} {"chunk_id": "climate_projections-cmip6_model-performance_q02__27fcac862052", "report_id": "climate_projections-cmip6_model-performance_q02", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "model-performance_q02", "aspect_base": "model-performance", "category": "Climate_Projections", "match_confidence": "exact", "section": "Historical accuracy of sea ice extent in the CMIP6 experiments > Quality assessment statement", "title": "Historical accuracy of sea ice extent in the CMIP6 experiments", "chunk_index": 1, "token_count": 975, "text_raw": "These are the key outcomes of this assessment\n\n- Bearing in mind that sea ice concentration estimates from passive microwave observations themselves are quite uncertain, we find that the errors between them and the CMIP6 models in the historical experiment can have significant biases and also have quite a large spread, so should be treated with some caution.\n- We find that the CMIP6 models generally underestimate the sea ice extent and area. Errors for the Antarctic are approximately double those in the Arctic. While the Arctic sea ice minima are generally quite accurate, the Arctic maxima are consistently underestimated, as are the Antarctic extrema.\n- At the time of the Arctic minimum, there is a general underestimation in the pack ice, but an overestimation in the marginal ice zone (MIZ) and near the coasts.\n- At the time of the Arctic maximum, the concentration in the pack is relatively unbiased, but there are some areas where there is strong underestimation (Bering Sea, Sea of Okhotsk)and overestimation (Greenland Sea, Labrador Sea).\n- Arctic December: one of the other CMIP6 quality assessments looks at projections of accessibility of Arctic shipping routes, where sea ice in December is particularly interesting. Therefore we also check the reliability of the December concentrations in the CMIP6 models. We find that the concentration in the pack is quite similar to the observations, although there is significant underestimation in Hudson Bay, and less pronounced underestimation in the Bering Sea. In later years, there is also slight underestimation off the Russian coast. The ice extent in the Greenland and Barents Sea is consistently overestimated.\n- At the time of the Antarctic minimum, there is strong underestimation in the Weddell, Bellingshausen and Amundsen Seas, while south of the Pacific and Indian Oceans, there is less bias. However, there there is too little ice at the coast and too much away from it.\n- At the time of the Antarctic maximum, there is in general too little ice everywhere, with the region away from the coast at longitude about 140W south of the Atlantic Ocean, and in the region south of the Indian Ocean having the most pronounced underestimation. The underestimation on those areas is also increasing with time, while the concentration from satellite is staying relatively constant in this month as well.\n- These results are generally consistent with analyses from other authors - for the Arctic by the [SIMIP community (2020)](https://doi.org/10.1029/2019GL086749), [Davy and Outten (2020)](https://doi.org/10.1175/JCLI-D-19-0990.1), [Shu et al (2020)](https://doi.org/10.1029/2020GL087965), [Watts et al (2021)](https://doi.org/10.1175/JCLI-D-20-0491.1), [Henke et al (2023)](https://doi.org/10.1080/15230430.2023.2271592) and [Frankignoul et al (2024)](https://doi.org/10.1175/JCLI-D-23-0452.1); for the Antarctic by [Roach et al (2020)](https://doi.org/10.1029/2019GL086729), [Shu et al (2020)](https://doi.org/10.1029/2020GL087965), [Nie et al (2023)](https://doi.org/10.1029/2023GL105265) and [Li et al (2023)](https://doi.org/10.3390/rs15082048).\n- We ourselves don't consider time series of the sea ice minima themselves, but only plot climatologies of the multi-model mean to see how the differences between the models and the observations are distributed spatially. However, this has been done by a few others, e.g. [Shu et al (2020)](https://doi.org/10.1029/2020GL087965), who found the observed Arctic September sea ice extent (SIE) declining trend between 1979 and 2014 is slightly underestimated in CMIP6 models, while the observed weak but significant upward trend of the Antarctic SIE is not captured.\n```", "text_with_prefix": "EQC Quality Assessment: \"Historical accuracy of sea ice extent in the CMIP6 experiments\"\nDataset: projections-cmip6 [CDS]\nAspect: model-performance_q02 | Category: Climate_Projections\nSection: Historical accuracy of sea ice extent in the CMIP6 experiments > Quality assessment statement\n---\nThese are the key outcomes of this assessment\n\n- Bearing in mind that sea ice concentration estimates from passive microwave observations themselves are quite uncertain, we find that the errors between them and the CMIP6 models in the historical experiment can have significant biases and also have quite a large spread, so should be treated with some caution.\n- We find that the CMIP6 models generally underestimate the sea ice extent and area. Errors for the Antarctic are approximately double those in the Arctic. While the Arctic sea ice minima are generally quite accurate, the Arctic maxima are consistently underestimated, as are the Antarctic extrema.\n- At the time of the Arctic minimum, there is a general underestimation in the pack ice, but an overestimation in the marginal ice zone (MIZ) and near the coasts.\n- At the time of the Arctic maximum, the concentration in the pack is relatively unbiased, but there are some areas where there is strong underestimation (Bering Sea, Sea of Okhotsk)and overestimation (Greenland Sea, Labrador Sea).\n- Arctic December: one of the other CMIP6 quality assessments looks at projections of accessibility of Arctic shipping routes, where sea ice in December is particularly interesting. Therefore we also check the reliability of the December concentrations in the CMIP6 models. We find that the concentration in the pack is quite similar to the observations, although there is significant underestimation in Hudson Bay, and less pronounced underestimation in the Bering Sea. In later years, there is also slight underestimation off the Russian coast. The ice extent in the Greenland and Barents Sea is consistently overestimated.\n- At the time of the Antarctic minimum, there is strong underestimation in the Weddell, Bellingshausen and Amundsen Seas, while south of the Pacific and Indian Oceans, there is less bias. However, there there is too little ice at the coast and too much away from it.\n- At the time of the Antarctic maximum, there is in general too little ice everywhere, with the region away from the coast at longitude about 140W south of the Atlantic Ocean, and in the region south of the Indian Ocean having the most pronounced underestimation. The underestimation on those areas is also increasing with time, while the concentration from satellite is staying relatively constant in this month as well.\n- These results are generally consistent with analyses from other authors - for the Arctic by the [SIMIP community (2020)](https://doi.org/10.1029/2019GL086749), [Davy and Outten (2020)](https://doi.org/10.1175/JCLI-D-19-0990.1), [Shu et al (2020)](https://doi.org/10.1029/2020GL087965), [Watts et al (2021)](https://doi.org/10.1175/JCLI-D-20-0491.1), [Henke et al (2023)](https://doi.org/10.1080/15230430.2023.2271592) and [Frankignoul et al (2024)](https://doi.org/10.1175/JCLI-D-23-0452.1); for the Antarctic by [Roach et al (2020)](https://doi.org/10.1029/2019GL086729), [Shu et al (2020)](https://doi.org/10.1029/2020GL087965), [Nie et al (2023)](https://doi.org/10.1029/2023GL105265) and [Li et al (2023)](https://doi.org/10.3390/rs15082048).\n- We ourselves don't consider time series of the sea ice minima themselves, but only plot climatologies of the multi-model mean to see how the differences between the models and the observations are distributed spatially. However, this has been done by a few others, e.g. [Shu et al (2020)](https://doi.org/10.1029/2020GL087965), who found the observed Arctic September sea ice extent (SIE) declining trend between 1979 and 2014 is slightly underestimated in CMIP6 models, while the observed weak but significant upward trend of the Antarctic SIE is not captured.\n```"} {"chunk_id": "climate_projections-cmip6_model-performance_q02__e34228da0115", "report_id": "climate_projections-cmip6_model-performance_q02", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "model-performance_q02", "aspect_base": "model-performance", "category": "Climate_Projections", "match_confidence": "exact", "section": "Historical accuracy of sea ice extent in the CMIP6 experiments > Methodology", "title": "Historical accuracy of sea ice extent in the CMIP6 experiments", "chunk_index": 2, "token_count": 550, "text_raw": "We compare sea ice concentrations from the [CMIP6](https://cds.climate.copernicus.eu/datasets/projections-cmip6?tab=overview) historical experiment with that obtained from [passive microwave satellite products from EUMETSAT OSI- SAF and ESA CCI](https://cds.climate.copernicus.eu/datasets/satellite-sea-ice-concentration?tab=overview). Time series of the evaluation metrics Integrated Ice Edge Error (IIEE) ([Goessling et al, 2016](https://doi.org/10.1002/2015GL067232); [Henke et al, 2023)](https://doi.org/10.1080/15230430.2023.2271592), bias in extent and area, and the root mean square error (RMSE) in sea ice concentration are produced and plotted. All of these statistics are area-weighted averages. We consider all the CMIP6 models that output sea ice concentration in the historical experiment, but remove outliers by only plotting the interquartile limits (IQLs). Other authors have chosen different subsetting approaches - e.g. by scoring the models and only retaining the top-performing ones. We recommend users test as many models as possible, before deciding if a subset adequately represents the uncertainty for their application.\n\nDecadal climatologies for both the models and observations are also produced and compared, for the months March, September and December.\n\nThe “Analysis and results” section is organised as follows:\n\n**[](section-1)**\n\n**[](section-2)**\n\n**[](section-3)**\n\n        **[](section-3.1)**\n\n        Plot time series of the error metrics.\n\n        **[](section-3.2)**\n\n        This section has maps of climatologies and their biases for the [](section-3.2.1), [](section-3.2.2), [](section-3.2.3), [](section-3.2.4). We also plot maps for the [](section-3.2.5) since this was an interesting month for the climate projections.", "text_with_prefix": "EQC Quality Assessment: \"Historical accuracy of sea ice extent in the CMIP6 experiments\"\nDataset: projections-cmip6 [CDS]\nAspect: model-performance_q02 | Category: Climate_Projections\nSection: Historical accuracy of sea ice extent in the CMIP6 experiments > Methodology\n---\nWe compare sea ice concentrations from the [CMIP6](https://cds.climate.copernicus.eu/datasets/projections-cmip6?tab=overview) historical experiment with that obtained from [passive microwave satellite products from EUMETSAT OSI- SAF and ESA CCI](https://cds.climate.copernicus.eu/datasets/satellite-sea-ice-concentration?tab=overview). Time series of the evaluation metrics Integrated Ice Edge Error (IIEE) ([Goessling et al, 2016](https://doi.org/10.1002/2015GL067232); [Henke et al, 2023)](https://doi.org/10.1080/15230430.2023.2271592), bias in extent and area, and the root mean square error (RMSE) in sea ice concentration are produced and plotted. All of these statistics are area-weighted averages. We consider all the CMIP6 models that output sea ice concentration in the historical experiment, but remove outliers by only plotting the interquartile limits (IQLs). Other authors have chosen different subsetting approaches - e.g. by scoring the models and only retaining the top-performing ones. We recommend users test as many models as possible, before deciding if a subset adequately represents the uncertainty for their application.\n\nDecadal climatologies for both the models and observations are also produced and compared, for the months March, September and December.\n\nThe “Analysis and results” section is organised as follows:\n\n**[](section-1)**\n\n**[](section-2)**\n\n**[](section-3)**\n\n        **[](section-3.1)**\n\n        Plot time series of the error metrics.\n\n        **[](section-3.2)**\n\n        This section has maps of climatologies and their biases for the [](section-3.2.1), [](section-3.2.2), [](section-3.2.3), [](section-3.2.4). We also plot maps for the [](section-3.2.5) since this was an interesting month for the climate projections."} {"chunk_id": "climate_projections-cmip6_model-performance_q02__10bf8f4c204e", "report_id": "climate_projections-cmip6_model-performance_q02", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "model-performance_q02", "aspect_base": "model-performance", "category": "Climate_Projections", "match_confidence": "exact", "section": "Historical accuracy of sea ice extent in the CMIP6 experiments > Analysis and results > 1. Import libraries, set parameters and definitions of functions > 1.2 Set parameters", "title": "Historical accuracy of sea ice extent in the CMIP6 experiments", "chunk_index": 3, "token_count": 312, "text_raw": "- Set the time period to be analysed with `year_start` and `year_stop`.\n- Set the concentration threshold `sic_threshold` for determining sea ice extent (we use 30% to be consistent with the ice edge product).\n- `clim_months` is a list of the months to plot the climatologies for.\n- `decades_historical` of starting years of decades to calculate climatologies for.\n- `projections` is a dictionary of projections to plot the climatology maps in.\n- `map_slices` is a dictionary of slices in the x and y directions (on observation grid) to zoom in on the climatology maps.\n- Set the regions to be analysed with the list `regions`.\n- Set the models to be evaluated with the list `models`.\n\nSea Ice Concentration Threshold\nmonths to plot the climatologies for\nstarting years of decades to calculate climatologies for\nprojections to plot the climatology maps in\nslices in x and y directions (on observation grid) to zoom in on the climatology maps\nSelect regions\nChoose CMIP6 historical models", "text_with_prefix": "EQC Quality Assessment: \"Historical accuracy of sea ice extent in the CMIP6 experiments\"\nDataset: projections-cmip6 [CDS]\nAspect: model-performance_q02 | Category: Climate_Projections\nSection: Historical accuracy of sea ice extent in the CMIP6 experiments > Analysis and results > 1. Import libraries, set parameters and definitions of functions > 1.2 Set parameters\n---\n- Set the time period to be analysed with `year_start` and `year_stop`.\n- Set the concentration threshold `sic_threshold` for determining sea ice extent (we use 30% to be consistent with the ice edge product).\n- `clim_months` is a list of the months to plot the climatologies for.\n- `decades_historical` of starting years of decades to calculate climatologies for.\n- `projections` is a dictionary of projections to plot the climatology maps in.\n- `map_slices` is a dictionary of slices in the x and y directions (on observation grid) to zoom in on the climatology maps.\n- Set the regions to be analysed with the list `regions`.\n- Set the models to be evaluated with the list `models`.\n\nSea Ice Concentration Threshold\nmonths to plot the climatologies for\nstarting years of decades to calculate climatologies for\nprojections to plot the climatology maps in\nslices in x and y directions (on observation grid) to zoom in on the climatology maps\nSelect regions\nChoose CMIP6 historical models"} {"chunk_id": "climate_projections-cmip6_model-performance_q02__1dca0b7155ae", "report_id": "climate_projections-cmip6_model-performance_q02", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "model-performance_q02", "aspect_base": "model-performance", "category": "Climate_Projections", "match_confidence": "exact", "section": "Historical accuracy of sea ice extent in the CMIP6 experiments > Analysis and results > 1. Import libraries, set parameters and definitions of functions > 1.4 Functions to create the time series", "title": "Historical accuracy of sea ice extent in the CMIP6 experiments", "chunk_index": 4, "token_count": 315, "text_raw": "These functions are all applied to a single CMIP6 model.\n- `interpolate_to_satellite_grid` interpolates the model to the satellite grid.\n- `get_monthly_interpolated_data` is applied to both the model and satellite data, to take the monthly mean (the satellite data is daily, but this also forces all the models to have the same time coordinate) and interpolate to the satellite grid (only for model data). It also calculates the RMS error (only for satellite data).\n- `get_satellite_data` downloads the daily satellite data and applies `get_monthly_interpolated_data` to get the monthly mean and the RMS error in the data.\n- `compare_model_vs_obs` compares the sea ice concentration from a single CMIP6 model with satellite estimates, calculating error metrics like biases in concentration and extent, RMSE, and IIEE.\n- `compute_sea_ice_evaluation_diagnostics` downloads the satellite and model data and calls `compare_model_vs_obs` to get a time series of error metrics.\n\nRemove nan columns\nGet variables\nCompute useful variables\nCompute output", "text_with_prefix": "EQC Quality Assessment: \"Historical accuracy of sea ice extent in the CMIP6 experiments\"\nDataset: projections-cmip6 [CDS]\nAspect: model-performance_q02 | Category: Climate_Projections\nSection: Historical accuracy of sea ice extent in the CMIP6 experiments > Analysis and results > 1. Import libraries, set parameters and definitions of functions > 1.4 Functions to create the time series\n---\nThese functions are all applied to a single CMIP6 model.\n- `interpolate_to_satellite_grid` interpolates the model to the satellite grid.\n- `get_monthly_interpolated_data` is applied to both the model and satellite data, to take the monthly mean (the satellite data is daily, but this also forces all the models to have the same time coordinate) and interpolate to the satellite grid (only for model data). It also calculates the RMS error (only for satellite data).\n- `get_satellite_data` downloads the daily satellite data and applies `get_monthly_interpolated_data` to get the monthly mean and the RMS error in the data.\n- `compare_model_vs_obs` compares the sea ice concentration from a single CMIP6 model with satellite estimates, calculating error metrics like biases in concentration and extent, RMSE, and IIEE.\n- `compute_sea_ice_evaluation_diagnostics` downloads the satellite and model data and calls `compare_model_vs_obs` to get a time series of error metrics.\n\nRemove nan columns\nGet variables\nCompute useful variables\nCompute output"} {"chunk_id": "climate_projections-cmip6_model-performance_q02__8bde81c4189b", "report_id": "climate_projections-cmip6_model-performance_q02", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "model-performance_q02", "aspect_base": "model-performance", "category": "Climate_Projections", "match_confidence": "exact", "section": "Historical accuracy of sea ice extent in the CMIP6 experiments > Analysis and results > 1. Import libraries, set parameters and definitions of functions > 1.5 Post-processing and plotting of the time series", "title": "Historical accuracy of sea ice extent in the CMIP6 experiments", "chunk_index": 5, "token_count": 211, "text_raw": "- `postprocess_dataset` is applied after loading from the cache and renames some variables and dimensions for easier use.\n- `plot_timeseries` plots time series for each error metric (sea ice concentration bias and RMSE, sea ice extent bias and IIEE), and for each region (Arctic or Antarctic).\n\nerr_name = \"RMS obs error\"\nget median and interquartile limits to show the spread of the models\nplot the median\nPlot spread and add observation errors\nadd obs errors to plots\nadd obs errors to plots\nfix time labels for last variable", "text_with_prefix": "EQC Quality Assessment: \"Historical accuracy of sea ice extent in the CMIP6 experiments\"\nDataset: projections-cmip6 [CDS]\nAspect: model-performance_q02 | Category: Climate_Projections\nSection: Historical accuracy of sea ice extent in the CMIP6 experiments > Analysis and results > 1. Import libraries, set parameters and definitions of functions > 1.5 Post-processing and plotting of the time series\n---\n- `postprocess_dataset` is applied after loading from the cache and renames some variables and dimensions for easier use.\n- `plot_timeseries` plots time series for each error metric (sea ice concentration bias and RMSE, sea ice extent bias and IIEE), and for each region (Arctic or Antarctic).\n\nerr_name = \"RMS obs error\"\nget median and interquartile limits to show the spread of the models\nplot the median\nPlot spread and add observation errors\nadd obs errors to plots\nadd obs errors to plots\nfix time labels for last variable"} {"chunk_id": "climate_projections-cmip6_model-performance_q02__0d5f80e383c2", "report_id": "climate_projections-cmip6_model-performance_q02", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "model-performance_q02", "aspect_base": "model-performance", "category": "Climate_Projections", "match_confidence": "exact", "section": "Historical accuracy of sea ice extent in the CMIP6 experiments > Analysis and results > 1. Import libraries, set parameters and definitions of functions > 1.6 Functions to compute the climatologies", "title": "Historical accuracy of sea ice extent in the CMIP6 experiments", "chunk_index": 6, "token_count": 249, "text_raw": "- `compute_monthly_climatology` transforms a dataset (which can contain model or observation data) by sorting according to month, taking the time average, and interpolating to the satellite grid (model data only).\n- For a specified time interval, `get_monthly_climatology_eumetsat` downloads the satellite data and calculates the climatology for the selected region and months.\n- For a specified time interval, `get_monthly_climatology_model` downloads and calculates the climatology for the selected region, months and model.\n- `get_monthly_climatologies_cmip6` calls `get_monthly_climatology_model` once for each model and calculates the climatology of the ensemble mean.\n\nsome models produce extra variables so drop any that are not needed", "text_with_prefix": "EQC Quality Assessment: \"Historical accuracy of sea ice extent in the CMIP6 experiments\"\nDataset: projections-cmip6 [CDS]\nAspect: model-performance_q02 | Category: Climate_Projections\nSection: Historical accuracy of sea ice extent in the CMIP6 experiments > Analysis and results > 1. Import libraries, set parameters and definitions of functions > 1.6 Functions to compute the climatologies\n---\n- `compute_monthly_climatology` transforms a dataset (which can contain model or observation data) by sorting according to month, taking the time average, and interpolating to the satellite grid (model data only).\n- For a specified time interval, `get_monthly_climatology_eumetsat` downloads the satellite data and calculates the climatology for the selected region and months.\n- For a specified time interval, `get_monthly_climatology_model` downloads and calculates the climatology for the selected region, months and model.\n- `get_monthly_climatologies_cmip6` calls `get_monthly_climatology_model` once for each model and calculates the climatology of the ensemble mean.\n\nsome models produce extra variables so drop any that are not needed"} {"chunk_id": "climate_projections-cmip6_model-performance_q02__f3ade0d43dc9", "report_id": "climate_projections-cmip6_model-performance_q02", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "model-performance_q02", "aspect_base": "model-performance", "category": "Climate_Projections", "match_confidence": "exact", "section": "Historical accuracy of sea ice extent in the CMIP6 experiments > Analysis and results > 1. Import libraries, set parameters and definitions of functions > 1.7 Post-processing and plotting of climatologies", "title": "Historical accuracy of sea ice extent in the CMIP6 experiments", "chunk_index": 7, "token_count": 237, "text_raw": "- `postprocess_climatology` is applied after loading a climatology from the cache. It makes the name and units of the sea ice concentration variable consistent between datasets.\n- `make_sic_maps` plots the climatology of the CMIP6 ensemble mean next to the observed climatology for the three decades considered (1985-1994, 1995-2004, 2005-2014), which are shown in rows.\n- `make_sic_bias_maps` plots the bias (CMIP6 ensemble mean minus the observations) for the same decades.\n- `compare_sic_maps` is a wrapper for those functions.\n\nrename month\nrename SIC and convert to %\n\n(section-2)=", "text_with_prefix": "EQC Quality Assessment: \"Historical accuracy of sea ice extent in the CMIP6 experiments\"\nDataset: projections-cmip6 [CDS]\nAspect: model-performance_q02 | Category: Climate_Projections\nSection: Historical accuracy of sea ice extent in the CMIP6 experiments > Analysis and results > 1. Import libraries, set parameters and definitions of functions > 1.7 Post-processing and plotting of climatologies\n---\n- `postprocess_climatology` is applied after loading a climatology from the cache. It makes the name and units of the sea ice concentration variable consistent between datasets.\n- `make_sic_maps` plots the climatology of the CMIP6 ensemble mean next to the observed climatology for the three decades considered (1985-1994, 1995-2004, 2005-2014), which are shown in rows.\n- `make_sic_bias_maps` plots the bias (CMIP6 ensemble mean minus the observations) for the same decades.\n- `compare_sic_maps` is a wrapper for those functions.\n\nrename month\nrename SIC and convert to %\n\n(section-2)="} {"chunk_id": "climate_projections-cmip6_model-performance_q02__5be98cc0a361", "report_id": "climate_projections-cmip6_model-performance_q02", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "model-performance_q02", "aspect_base": "model-performance", "category": "Climate_Projections", "match_confidence": "exact", "section": "Historical accuracy of sea ice extent in the CMIP6 experiments > Analysis and results > 2. Download and transform data", "title": "Historical accuracy of sea ice extent in the CMIP6 experiments", "chunk_index": 8, "token_count": 108, "text_raw": "Download and transform the data and save it to disk. The second time `download_and_transform` is run it will save time by loading the transformed data from the disk.", "text_with_prefix": "EQC Quality Assessment: \"Historical accuracy of sea ice extent in the CMIP6 experiments\"\nDataset: projections-cmip6 [CDS]\nAspect: model-performance_q02 | Category: Climate_Projections\nSection: Historical accuracy of sea ice extent in the CMIP6 experiments > Analysis and results > 2. Download and transform data\n---\nDownload and transform the data and save it to disk. The second time `download_and_transform` is run it will save time by loading the transformed data from the disk."} {"chunk_id": "climate_projections-cmip6_model-performance_q02__b8ad5c228d7b", "report_id": "climate_projections-cmip6_model-performance_q02", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "model-performance_q02", "aspect_base": "model-performance", "category": "Climate_Projections", "match_confidence": "exact", "section": "Historical accuracy of sea ice extent in the CMIP6 experiments > Analysis and results > 2. Download and transform data > 2.1 Set some parameters for downloading", "title": "Historical accuracy of sea ice extent in the CMIP6 experiments", "chunk_index": 9, "token_count": 144, "text_raw": "- `interpolation_kwargs` sets some interpolation options.\n- `io_kwargs` sets some parameters to speed up the download process.\n- `eval_kwargs` contains all the options that will be passed to `download.download_and_transform` when creating the time series of evaluation metrics.\n\nParameters to speed up IO", "text_with_prefix": "EQC Quality Assessment: \"Historical accuracy of sea ice extent in the CMIP6 experiments\"\nDataset: projections-cmip6 [CDS]\nAspect: model-performance_q02 | Category: Climate_Projections\nSection: Historical accuracy of sea ice extent in the CMIP6 experiments > Analysis and results > 2. Download and transform data > 2.1 Set some parameters for downloading\n---\n- `interpolation_kwargs` sets some interpolation options.\n- `io_kwargs` sets some parameters to speed up the download process.\n- `eval_kwargs` contains all the options that will be passed to `download.download_and_transform` when creating the time series of evaluation metrics.\n\nParameters to speed up IO"} {"chunk_id": "climate_projections-cmip6_model-performance_q02__05e5d787dcea", "report_id": "climate_projections-cmip6_model-performance_q02", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "model-performance_q02", "aspect_base": "model-performance", "category": "Climate_Projections", "match_confidence": "exact", "section": "Historical accuracy of sea ice extent in the CMIP6 experiments > Analysis and results > 2. Download and transform data > 2.2 Download and transform CMIP6 data for time series", "title": "Historical accuracy of sea ice extent in the CMIP6 experiments", "chunk_index": 10, "token_count": 158, "text_raw": "For each model and region, we download the data, transform it with `compute_sea_ice_evaluation_diagnostics`, save it to disk and post-process with `postprocess_dataset`. The time series are then combined into one dataset with extra dimensions `model` and `region`.\n\noptions to be passed to compute_sea_ice_evaluation_diagnostics", "text_with_prefix": "EQC Quality Assessment: \"Historical accuracy of sea ice extent in the CMIP6 experiments\"\nDataset: projections-cmip6 [CDS]\nAspect: model-performance_q02 | Category: Climate_Projections\nSection: Historical accuracy of sea ice extent in the CMIP6 experiments > Analysis and results > 2. Download and transform data > 2.2 Download and transform CMIP6 data for time series\n---\nFor each model and region, we download the data, transform it with `compute_sea_ice_evaluation_diagnostics`, save it to disk and post-process with `postprocess_dataset`. The time series are then combined into one dataset with extra dimensions `model` and `region`.\n\noptions to be passed to compute_sea_ice_evaluation_diagnostics"} {"chunk_id": "climate_projections-cmip6_model-performance_q02__1b88c6181525", "report_id": "climate_projections-cmip6_model-performance_q02", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "model-performance_q02", "aspect_base": "model-performance", "category": "Climate_Projections", "match_confidence": "exact", "section": "Historical accuracy of sea ice extent in the CMIP6 experiments > Analysis and results > 2. Download and transform data > 2.4 Download the mean climatology for the historical CMIP6 experiment", "title": "Historical accuracy of sea ice extent in the CMIP6 experiments", "chunk_index": 11, "token_count": 155, "text_raw": "We use all models in the historical experiment to get the ensemble mean and get the climatology for one region and decade at a time to see the spatial distribution of the bias when compared to satellite estimates, and to see how the bias develops with time.\n\nset some options for plotting the climatologies\n\n(section-3)=", "text_with_prefix": "EQC Quality Assessment: \"Historical accuracy of sea ice extent in the CMIP6 experiments\"\nDataset: projections-cmip6 [CDS]\nAspect: model-performance_q02 | Category: Climate_Projections\nSection: Historical accuracy of sea ice extent in the CMIP6 experiments > Analysis and results > 2. Download and transform data > 2.4 Download the mean climatology for the historical CMIP6 experiment\n---\nWe use all models in the historical experiment to get the ensemble mean and get the climatology for one region and decade at a time to see the spatial distribution of the bias when compared to satellite estimates, and to see how the bias develops with time.\n\nset some options for plotting the climatologies\n\n(section-3)="} {"chunk_id": "climate_projections-cmip6_model-performance_q02__af2101332776", "report_id": "climate_projections-cmip6_model-performance_q02", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "model-performance_q02", "aspect_base": "model-performance", "category": "Climate_Projections", "match_confidence": "exact", "section": "Historical accuracy of sea ice extent in the CMIP6 experiments > Analysis and results > 3. Results > 3.1 Time series of evaluation metrics", "title": "Historical accuracy of sea ice extent in the CMIP6 experiments", "chunk_index": 12, "token_count": 587, "text_raw": "The evalution metrics that will be the most important here are the IIEE and the bias in extent, as the determination of the presence or absence of ice from passive microwave is more reliable than the actual value of the concentration derived from it. In particular, there is systematic underestimation of the concentration. This is especially true in summer due to meltponds for example, but it is also true in winter where the EUMETSAT OSI-SAF product usually estimates about 90% concentration in the Arctic pack (when it is very close to 100%, as can be seen from SAR images for example). The ESA CCI product which has a smaller footprint usually estimates a higher concentration than EUMETSAF OSI-SAF in the Arctic pack. Nevertheless it is still interesting to look at the area-weighted bias and RMSE in concentration itself - for example an underestimation of concentration would generally be undesirable for a model. The plots show the median and inter-quartile limits (IQL) to show the spread of our ensemble of CMIP6 models (all those that output sea ice concentration in the historical experiment).\n\nThe IIEE for the CMIP6 models is quite high, being about $2\\times10^6$km$^2$ in the Arctic and twice this in the Antarctic. (Refer to the sea ice diagnostics notebook which shows the historical Arctic minimum extent is about $8\\times10^6$km$^2$.) The bias in Arctic extent is not too high, although it is quite variable. For Antarctica however it is quite high, underestimating it by about $4\\times10^6$km$^2$ by 2015 (growing from around $1\\times10^6$km$^2$ in 1980).\n\nThe bias for the concentration itself is quite low for CMIP6, being well within the limits of the observation error. There is a clear tendency for underestimation by CMIP6 in Antarctica though, with the bias dropping linearly. The RMS error for CMIP6 is also very high, indicating a different spatial distribution to the satellite data, since the bias was not too high.\n\nWe will take a look at the spatial distribution more closely in the next section.\n\nFinally we note that when the ESA-CCI and EUMETSAT OSI-SAF products overlap in time (after 2002), the metrics for CMIP6 are quite similar.\n\n(section-3.2)=", "text_with_prefix": "EQC Quality Assessment: \"Historical accuracy of sea ice extent in the CMIP6 experiments\"\nDataset: projections-cmip6 [CDS]\nAspect: model-performance_q02 | Category: Climate_Projections\nSection: Historical accuracy of sea ice extent in the CMIP6 experiments > Analysis and results > 3. Results > 3.1 Time series of evaluation metrics\n---\nThe evalution metrics that will be the most important here are the IIEE and the bias in extent, as the determination of the presence or absence of ice from passive microwave is more reliable than the actual value of the concentration derived from it. In particular, there is systematic underestimation of the concentration. This is especially true in summer due to meltponds for example, but it is also true in winter where the EUMETSAT OSI-SAF product usually estimates about 90% concentration in the Arctic pack (when it is very close to 100%, as can be seen from SAR images for example). The ESA CCI product which has a smaller footprint usually estimates a higher concentration than EUMETSAF OSI-SAF in the Arctic pack. Nevertheless it is still interesting to look at the area-weighted bias and RMSE in concentration itself - for example an underestimation of concentration would generally be undesirable for a model. The plots show the median and inter-quartile limits (IQL) to show the spread of our ensemble of CMIP6 models (all those that output sea ice concentration in the historical experiment).\n\nThe IIEE for the CMIP6 models is quite high, being about $2\\times10^6$km$^2$ in the Arctic and twice this in the Antarctic. (Refer to the sea ice diagnostics notebook which shows the historical Arctic minimum extent is about $8\\times10^6$km$^2$.) The bias in Arctic extent is not too high, although it is quite variable. For Antarctica however it is quite high, underestimating it by about $4\\times10^6$km$^2$ by 2015 (growing from around $1\\times10^6$km$^2$ in 1980).\n\nThe bias for the concentration itself is quite low for CMIP6, being well within the limits of the observation error. There is a clear tendency for underestimation by CMIP6 in Antarctica though, with the bias dropping linearly. The RMS error for CMIP6 is also very high, indicating a different spatial distribution to the satellite data, since the bias was not too high.\n\nWe will take a look at the spatial distribution more closely in the next section.\n\nFinally we note that when the ESA-CCI and EUMETSAT OSI-SAF products overlap in time (after 2002), the metrics for CMIP6 are quite similar.\n\n(section-3.2)="} {"chunk_id": "climate_projections-cmip6_model-performance_q02__4329f7c1b055", "report_id": "climate_projections-cmip6_model-performance_q02", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "model-performance_q02", "aspect_base": "model-performance", "category": "Climate_Projections", "match_confidence": "exact", "section": "Historical accuracy of sea ice extent in the CMIP6 experiments > Analysis and results > 3. Results > 3.2 Spatial distribution of errors", "title": "Historical accuracy of sea ice extent in the CMIP6 experiments", "chunk_index": 13, "token_count": 202, "text_raw": "We compare 10-year-averaged maps from 1985-2015 which is roughly the overlap period between the historical experiment and the EUMETSAT OSI-SAF product. In the Arctic, the mean concentration is quite close to the mean from observations, but there is disagreement in the seas with seaonal ice - e.g. Greenland, Barents, Labrador, Bering and Okhotsk Seas and Hudson Bay, and (in the summer) off the Russian and Alaskan coasts. For the Antarctic there is systematic underestimation of concentration.\n\n(section-3.2.1)=", "text_with_prefix": "EQC Quality Assessment: \"Historical accuracy of sea ice extent in the CMIP6 experiments\"\nDataset: projections-cmip6 [CDS]\nAspect: model-performance_q02 | Category: Climate_Projections\nSection: Historical accuracy of sea ice extent in the CMIP6 experiments > Analysis and results > 3. Results > 3.2 Spatial distribution of errors\n---\nWe compare 10-year-averaged maps from 1985-2015 which is roughly the overlap period between the historical experiment and the EUMETSAT OSI-SAF product. In the Arctic, the mean concentration is quite close to the mean from observations, but there is disagreement in the seas with seaonal ice - e.g. Greenland, Barents, Labrador, Bering and Okhotsk Seas and Hudson Bay, and (in the summer) off the Russian and Alaskan coasts. For the Antarctic there is systematic underestimation of concentration.\n\n(section-3.2.1)="} {"chunk_id": "climate_projections-cmip6_model-performance_q02__a44cd9b5a7d5", "report_id": "climate_projections-cmip6_model-performance_q02", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "model-performance_q02", "aspect_base": "model-performance", "category": "Climate_Projections", "match_confidence": "exact", "section": "Historical accuracy of sea ice extent in the CMIP6 experiments > Analysis and results > 3. Results > 3.2.1 Arctic minimum", "title": "Historical accuracy of sea ice extent in the CMIP6 experiments", "chunk_index": 14, "token_count": 177, "text_raw": "Looking at the maps below for September, we can see there is a general underestimation in the pack ice and an overestimation in the MIZ and near the coasts. This pattern persists with time although the locations move as the Arctic ice cover shrinks with time. This may be a resolution effect, since roughly the same amount of ice (low bias in sea ice concentration) may just be being spread out over a larger area.\n\n(section-3.2.2)=", "text_with_prefix": "EQC Quality Assessment: \"Historical accuracy of sea ice extent in the CMIP6 experiments\"\nDataset: projections-cmip6 [CDS]\nAspect: model-performance_q02 | Category: Climate_Projections\nSection: Historical accuracy of sea ice extent in the CMIP6 experiments > Analysis and results > 3. Results > 3.2.1 Arctic minimum\n---\nLooking at the maps below for September, we can see there is a general underestimation in the pack ice and an overestimation in the MIZ and near the coasts. This pattern persists with time although the locations move as the Arctic ice cover shrinks with time. This may be a resolution effect, since roughly the same amount of ice (low bias in sea ice concentration) may just be being spread out over a larger area.\n\n(section-3.2.2)="} {"chunk_id": "climate_projections-cmip6_model-performance_q02__ed37b9823008", "report_id": "climate_projections-cmip6_model-performance_q02", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "model-performance_q02", "aspect_base": "model-performance", "category": "Climate_Projections", "match_confidence": "exact", "section": "Historical accuracy of sea ice extent in the CMIP6 experiments > Analysis and results > 3. Results > 3.2.2 Arctic maximum", "title": "Historical accuracy of sea ice extent in the CMIP6 experiments", "chunk_index": 15, "token_count": 176, "text_raw": "Looking at the maps below for March, we can see the ice in the central Arctic is generally unbiased, but there are some areas where there is strong underestimation - the Bering Sea and the Sea of Okhotsk - and there is strong overestimation in the Greenland Sea and the North Atlantic Ocean near the entrance to Hudson Bay. Overall, the total ice area is roughly the same in CMIP6 and the satellite observations.\n\n(section-3.2.3)=", "text_with_prefix": "EQC Quality Assessment: \"Historical accuracy of sea ice extent in the CMIP6 experiments\"\nDataset: projections-cmip6 [CDS]\nAspect: model-performance_q02 | Category: Climate_Projections\nSection: Historical accuracy of sea ice extent in the CMIP6 experiments > Analysis and results > 3. Results > 3.2.2 Arctic maximum\n---\nLooking at the maps below for March, we can see the ice in the central Arctic is generally unbiased, but there are some areas where there is strong underestimation - the Bering Sea and the Sea of Okhotsk - and there is strong overestimation in the Greenland Sea and the North Atlantic Ocean near the entrance to Hudson Bay. Overall, the total ice area is roughly the same in CMIP6 and the satellite observations.\n\n(section-3.2.3)="} {"chunk_id": "climate_projections-cmip6_model-performance_q02__35ef0499cc2d", "report_id": "climate_projections-cmip6_model-performance_q02", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "model-performance_q02", "aspect_base": "model-performance", "category": "Climate_Projections", "match_confidence": "exact", "section": "Historical accuracy of sea ice extent in the CMIP6 experiments > Analysis and results > 3. Results > 3.2.3 Antarctic minimum", "title": "Historical accuracy of sea ice extent in the CMIP6 experiments", "chunk_index": 16, "token_count": 183, "text_raw": "There is consistently too little ice on the Atlantic side of Antarctica (in the Weddell, Bellingshausen and Amundsen Seas). The amount of sea ice in those areas is also decreasing with time, while the concentration from satellite is staying relatively constant. On the Pacific and Indian ocean sides there is more of a dipole pattern (too little at the coast and too much away from it), probably due to the effect of low resolution in the CMIP6 models.\n\n(section-3.2.4)=", "text_with_prefix": "EQC Quality Assessment: \"Historical accuracy of sea ice extent in the CMIP6 experiments\"\nDataset: projections-cmip6 [CDS]\nAspect: model-performance_q02 | Category: Climate_Projections\nSection: Historical accuracy of sea ice extent in the CMIP6 experiments > Analysis and results > 3. Results > 3.2.3 Antarctic minimum\n---\nThere is consistently too little ice on the Atlantic side of Antarctica (in the Weddell, Bellingshausen and Amundsen Seas). The amount of sea ice in those areas is also decreasing with time, while the concentration from satellite is staying relatively constant. On the Pacific and Indian ocean sides there is more of a dipole pattern (too little at the coast and too much away from it), probably due to the effect of low resolution in the CMIP6 models.\n\n(section-3.2.4)="} {"chunk_id": "climate_projections-cmip6_model-performance_q02__24654b97a955", "report_id": "climate_projections-cmip6_model-performance_q02", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "model-performance_q02", "aspect_base": "model-performance", "category": "Climate_Projections", "match_confidence": "exact", "section": "Historical accuracy of sea ice extent in the CMIP6 experiments > Analysis and results > 3. Results > 3.2.4 Antarctic maximum", "title": "Historical accuracy of sea ice extent in the CMIP6 experiments", "chunk_index": 17, "token_count": 163, "text_raw": "For the month of September there is in general too little ice everywhere, although away from the coast at longitude about 140$^\\circ$W in the Atlantic Ocean, and in the Indian Ocean the underestimation is most pronounced. The underestimation on those areas is also increasing with time, while the concentration from satellite is staying relatively constant in this month as well.\n\n(section-3.2.5)=", "text_with_prefix": "EQC Quality Assessment: \"Historical accuracy of sea ice extent in the CMIP6 experiments\"\nDataset: projections-cmip6 [CDS]\nAspect: model-performance_q02 | Category: Climate_Projections\nSection: Historical accuracy of sea ice extent in the CMIP6 experiments > Analysis and results > 3. Results > 3.2.4 Antarctic maximum\n---\nFor the month of September there is in general too little ice everywhere, although away from the coast at longitude about 140$^\\circ$W in the Atlantic Ocean, and in the Indian Ocean the underestimation is most pronounced. The underestimation on those areas is also increasing with time, while the concentration from satellite is staying relatively constant in this month as well.\n\n(section-3.2.5)="} {"chunk_id": "climate_projections-cmip6_model-performance_q02__7e57cf70b7f1", "report_id": "climate_projections-cmip6_model-performance_q02", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "model-performance_q02", "aspect_base": "model-performance", "category": "Climate_Projections", "match_confidence": "exact", "section": "Historical accuracy of sea ice extent in the CMIP6 experiments > Analysis and results > 3. Results > 3.2.5 Arctic December (relevant to the Arctic shipping route assessment)", "title": "Historical accuracy of sea ice extent in the CMIP6 experiments", "chunk_index": 18, "token_count": 197, "text_raw": "One of the other CMIP6 quality assessments looks at projections of accessibility of Arctic shipping routes, where sea ice in December is particularly interesting. Hence we also plot some maps for this month here. The CMIP6 ensemble mean is quite close to the satellite data in the Arctic Ocean, with the main feature of the bias maps being the underestimation in the Hudson Bay. Note that not all models that provide sea ice concentration in the historical experiment also provide it in the different projection experiments, so that adds some additional uncertainty to these conclusions.", "text_with_prefix": "EQC Quality Assessment: \"Historical accuracy of sea ice extent in the CMIP6 experiments\"\nDataset: projections-cmip6 [CDS]\nAspect: model-performance_q02 | Category: Climate_Projections\nSection: Historical accuracy of sea ice extent in the CMIP6 experiments > Analysis and results > 3. Results > 3.2.5 Arctic December (relevant to the Arctic shipping route assessment)\n---\nOne of the other CMIP6 quality assessments looks at projections of accessibility of Arctic shipping routes, where sea ice in December is particularly interesting. Hence we also plot some maps for this month here. The CMIP6 ensemble mean is quite close to the satellite data in the Arctic Ocean, with the main feature of the bias maps being the underestimation in the Hudson Bay. Note that not all models that provide sea ice concentration in the historical experiment also provide it in the different projection experiments, so that adds some additional uncertainty to these conclusions."} {"chunk_id": "climate_projections-cmip6_model-performance_q02__0c8191faba11", "report_id": "climate_projections-cmip6_model-performance_q02", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "model-performance_q02", "aspect_base": "model-performance", "category": "Climate_Projections", "match_confidence": "exact", "section": "Historical accuracy of sea ice extent in the CMIP6 experiments > ℹ️ If you want to know more > Key resources", "title": "Historical accuracy of sea ice extent in the CMIP6 experiments", "chunk_index": 19, "token_count": 363, "text_raw": "Introductory sea ice materials:\n- [Role of sea ice in the climate](https://marine.copernicus.eu/explainers/why-ocean-important/sea-ice)\n- [Sea ice as an indicator of climate change](https://climate.copernicus.eu/climate-indicators/sea-ice#)\n- [Observing sea ice with satellites](https://www.metoffice.gov.uk/research/climate/cryosphere-oceans/sea-ice/measure)\n\nIntroductory CMIP6 materials:\n- [A short introduction to CMIP and CMIP6](https://www.wcrp-climate.org/wgcm-cmip/cmip-video)\n- [CMIP6: the next generation of climate models explained](https://www.carbonbrief.org/cmip6-the-next-generation-of-climate-models-explained/)\n\nCode libraries used:\n- [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control) developed by [B-Open](https://www.bopen.eu/)", "text_with_prefix": "EQC Quality Assessment: \"Historical accuracy of sea ice extent in the CMIP6 experiments\"\nDataset: projections-cmip6 [CDS]\nAspect: model-performance_q02 | Category: Climate_Projections\nSection: Historical accuracy of sea ice extent in the CMIP6 experiments > ℹ️ If you want to know more > Key resources\n---\nIntroductory sea ice materials:\n- [Role of sea ice in the climate](https://marine.copernicus.eu/explainers/why-ocean-important/sea-ice)\n- [Sea ice as an indicator of climate change](https://climate.copernicus.eu/climate-indicators/sea-ice#)\n- [Observing sea ice with satellites](https://www.metoffice.gov.uk/research/climate/cryosphere-oceans/sea-ice/measure)\n\nIntroductory CMIP6 materials:\n- [A short introduction to CMIP and CMIP6](https://www.wcrp-climate.org/wgcm-cmip/cmip-video)\n- [CMIP6: the next generation of climate models explained](https://www.carbonbrief.org/cmip6-the-next-generation-of-climate-models-explained/)\n\nCode libraries used:\n- [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control) developed by [B-Open](https://www.bopen.eu/)"} {"chunk_id": "climate_projections-cmip6_model-performance_q02__ec567574f597", "report_id": "climate_projections-cmip6_model-performance_q02", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "model-performance_q02", "aspect_base": "model-performance", "category": "Climate_Projections", "match_confidence": "exact", "section": "Historical accuracy of sea ice extent in the CMIP6 experiments > ℹ️ If you want to know more > References", "title": "Historical accuracy of sea ice extent in the CMIP6 experiments", "chunk_index": 20, "token_count": 977, "text_raw": "1. Davy, R., and S. Outten (2020). The Arctic Surface Climate in CMIP6: Status and Developments since CMIP5. J. Climate, 33, 8047–8068, https://doi.org/10.1175/JCLI-D-19-0990.1 \n\n2. Frankignoul, C., L. Raillard, B. Ferster, and Y. Kwon (2024). Arctic September sea ice concentration biases in CMIP6 models and their relationships with other model variables. J. Climate, https://doi.org/10.1175/JCLI-D-23-0452.1 , in press.\n\n3. Goessling, H. F., S. Tietsche, J. J. Day, E. Hawkins, and T. Jung (2016). Predictability of the Arctic sea ice edge, Geophys. Res. Lett., 43, 1642–1650, https://doi.org/10.1002/2015GL067232 \n\n4. Henke, M., F. Cassalho, T. Miesse, C. M. Ferreira, J. Zhang and T. M. Ravens (2023). Assessment of Arctic sea ice and surface climate conditions in nine CMIP6 climate models. Arctic, Antarctic, and Alpine Research, 55(1). https://doi.org/10.1080/15230430.2023.2271592 \n\n5. Nie, Y., X. Lin, Q. Yang, J. Liu, D. Chen and P. Uotila (2023). Differences between the CMIP5 and CMIP6 Antarctic sea ice concentration budgets. Geophysical Research Letters, 50, e2023GL105265. https://doi.org/10.1029/2023GL105265 \n\n6. Li, S., Y. Zhang, C. Chen, Y. Zhang, D. Xu, and S. Hu (2023). Assessment of Antarctic Sea Ice Cover in CMIP6 Prediction with Comparison to AMSR2 during 2015–2021. Remote Sensing 15(8): 2048, https://doi.org/10.3390/rs15082048 \n\n7. Roach, L. A., J. Dörr, C. R. Holmes, F. Massonnet, E. W. Blockley, D. Notz, T. Rackow, M. N. Raphael, S. P. O'Farrell, D. A. Bailey, and C. M. Bitz (2020). Antarctic sea ice area in CMIP6. Geophysical Research Letters, 47, e2019GL086729, https://doi.org/10.1029/2019GL086729 \n\n8. Shu, Q., Q. Wang, Z. Song, F. Qiao, J. Zhao, M. Chu and X. Li (2020). Assessment of sea ice extent in CMIP6 with comparison to observations and CMIP5. Geophysical Research Letters, 47, e2020GL087965, https://doi.org/10.1029/2020GL087965 \n\n9. SIMIP Community (2020). Arctic sea ice in CMIP6. Geophysical Research Letters, 47, e2019GL086749, https://doi.org/10.1029/2019GL086749 \n\n10. Watts, M., W. Maslowski, Y. J. Lee, J. C. Kinney, and R. Osinski (2021). A Spatial Evaluation of Arctic Sea Ice and Regional Limitations in CMIP6 Historical Simulations. J. Climate, 34, 6399–6420, https://doi.org/10.1175/JCLI-D-20-0491.1 ", "text_with_prefix": "EQC Quality Assessment: \"Historical accuracy of sea ice extent in the CMIP6 experiments\"\nDataset: projections-cmip6 [CDS]\nAspect: model-performance_q02 | Category: Climate_Projections\nSection: Historical accuracy of sea ice extent in the CMIP6 experiments > ℹ️ If you want to know more > References\n---\n1. Davy, R., and S. Outten (2020). The Arctic Surface Climate in CMIP6: Status and Developments since CMIP5. J. Climate, 33, 8047–8068, https://doi.org/10.1175/JCLI-D-19-0990.1 \n\n2. Frankignoul, C., L. Raillard, B. Ferster, and Y. Kwon (2024). Arctic September sea ice concentration biases in CMIP6 models and their relationships with other model variables. J. Climate, https://doi.org/10.1175/JCLI-D-23-0452.1 , in press.\n\n3. Goessling, H. F., S. Tietsche, J. J. Day, E. Hawkins, and T. Jung (2016). Predictability of the Arctic sea ice edge, Geophys. Res. Lett., 43, 1642–1650, https://doi.org/10.1002/2015GL067232 \n\n4. Henke, M., F. Cassalho, T. Miesse, C. M. Ferreira, J. Zhang and T. M. Ravens (2023). Assessment of Arctic sea ice and surface climate conditions in nine CMIP6 climate models. Arctic, Antarctic, and Alpine Research, 55(1). https://doi.org/10.1080/15230430.2023.2271592 \n\n5. Nie, Y., X. Lin, Q. Yang, J. Liu, D. Chen and P. Uotila (2023). Differences between the CMIP5 and CMIP6 Antarctic sea ice concentration budgets. Geophysical Research Letters, 50, e2023GL105265. https://doi.org/10.1029/2023GL105265 \n\n6. Li, S., Y. Zhang, C. Chen, Y. Zhang, D. Xu, and S. Hu (2023). Assessment of Antarctic Sea Ice Cover in CMIP6 Prediction with Comparison to AMSR2 during 2015–2021. Remote Sensing 15(8): 2048, https://doi.org/10.3390/rs15082048 \n\n7. Roach, L. A., J. Dörr, C. R. Holmes, F. Massonnet, E. W. Blockley, D. Notz, T. Rackow, M. N. Raphael, S. P. O'Farrell, D. A. Bailey, and C. M. Bitz (2020). Antarctic sea ice area in CMIP6. Geophysical Research Letters, 47, e2019GL086729, https://doi.org/10.1029/2019GL086729 \n\n8. Shu, Q., Q. Wang, Z. Song, F. Qiao, J. Zhao, M. Chu and X. Li (2020). Assessment of sea ice extent in CMIP6 with comparison to observations and CMIP5. Geophysical Research Letters, 47, e2020GL087965, https://doi.org/10.1029/2020GL087965 \n\n9. SIMIP Community (2020). Arctic sea ice in CMIP6. Geophysical Research Letters, 47, e2019GL086749, https://doi.org/10.1029/2019GL086749 \n\n10. Watts, M., W. Maslowski, Y. J. Lee, J. C. Kinney, and R. Osinski (2021). A Spatial Evaluation of Arctic Sea Ice and Regional Limitations in CMIP6 Historical Simulations. J. Climate, 34, 6399–6420, https://doi.org/10.1175/JCLI-D-20-0491.1 "} {"chunk_id": "climate_projections-cmip6_model-performance_q03__40ec278b07ae", "report_id": "climate_projections-cmip6_model-performance_q03", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "model-performance_q03", "aspect_base": "model-performance", "category": "Climate_Projections", "match_confidence": "exact", "section": "Evaluation of the Arctic sea ice thickness in the CMIP6 historical experiments", "title": "Evaluation of the Arctic sea ice thickness in the CMIP6 historical experiments", "chunk_index": 0, "token_count": 95, "text_raw": "Production date: 2024-05-31\n\nProduced by: Timothy Williams, Nansen Environmental and Remote Sensing Center (NERSC)", "text_with_prefix": "EQC Quality Assessment: \"Evaluation of the Arctic sea ice thickness in the CMIP6 historical experiments\"\nDataset: projections-cmip6 [CDS]\nAspect: model-performance_q03 | Category: Climate_Projections\nSection: Evaluation of the Arctic sea ice thickness in the CMIP6 historical experiments\n---\nProduction date: 2024-05-31\n\nProduced by: Timothy Williams, Nansen Environmental and Remote Sensing Center (NERSC)"} {"chunk_id": "climate_projections-cmip6_model-performance_q03__5450e4361e57", "report_id": "climate_projections-cmip6_model-performance_q03", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "model-performance_q03", "aspect_base": "model-performance", "category": "Climate_Projections", "match_confidence": "exact", "section": "Evaluation of the Arctic sea ice thickness in the CMIP6 historical experiments > Quality assessment statement", "title": "Evaluation of the Arctic sea ice thickness in the CMIP6 historical experiments", "chunk_index": 1, "token_count": 973, "text_raw": "These are the key outcomes of this assessment\n\n- On average the Arctic sea ice thickness (SIT) from the CMIP6 ensemble is fairly unbiased against the altimeter-based estimates, with the exception of October. However the bias starts to become negative towards the end of the historical experiment in 2013 and 2014.\n\n- There are some spatial differences between the CMIP6 models and their ensemble mean. This is shown by the RMSE staying relatively constant throughout the whole evaluation period despite the low bias, indicating that the spatial biases are cancelling each other out. It is also shown by spatial maps of climatologies of the bias, which show that in the ENVISAT period the Beaufort Sea has a strong positive bias, while the Laptev, Kara and Barents Seas have more of a negative bias. In the CryoSat-2 period, the spatial distribution of the bias changes with the Beafort, Laptev, Kara and Barents Seas all have a positive bias, while the Greenland Sea, and the area extending from the north of Greenland to the north pole have a strong negative bias, making the mean bias negative.\n\n- We also compared maps of the climatology of the ensemble mean of the CMIP6 models during the CryoSat-2 period (2010-2020) to the climatology of the ensemble mean in the ENVISAT period (2002-2010). Spatially, the CMIP6 model mean has a fairly uniform drop in thickness when we move from the earlier satellite period to the later one, while the difference in the altimeter climatologies is more variable spatially.\n\n- Other authors have also looked at evaluating CMIP6 models sea ice thickness or volume. [West & Blockley (2024)](https://doi.org/10.5194/gmd-2024-121) compared a subset of the models (those that output heat flux diagnostics) to ice mass-balance buoys, finding that the CMIP6 models they looked at overestimated heat fluxes and thus sea ice melt and growth. A number of authors ([Davy & Outten, 2020](https://doi.org/10.1175/JCLI-D-19-0990.1); [Xu et al. , 2023](https://doi.org/10.3389/fmars.2023.1223772); [Watts et al. 2021](https://doi.org/10.1175/JCLI-D-20-0491.1)) found that the CMIP6 ensemble mean was similar to the PIOMAS (Pan-Arctic Ice-Ocean Modeling and Assimilation System) reanalysis, although the spread was quite large. [Davy & Outten (2020)](https://doi.org/10.1175/JCLI-D-19-0990.1) also included CMIP5 models in their comparison, noting that both the bias and spread have improved in CMIP6. The spatial patterns in the differences between altimeter and the CMIP6 ensemble mean that we found have also been found by others ([Watts et al., 2021](https://doi.org/10.1175/JCLI-D-20-0491.1); [Henke et al (2023)](https://doi.org/10.1080/15230430.2023.2271592)). In addition, [Kuang et al (2024)](https://doi.org/10.1016/j.accre.2024.06.008) calculated sea ice volume fluxes through six Arctic gateways (the most important being the Fram Strait), finding the multi-model mean also gave a reasonable representation of this flux.\n\n- A number of authors have also used CMIP6 sea ice thickness and volume when considering projections of sea ice. An interesting example of this is the article by [Zhou et al (2022)](https://doi.org/10.1088/1748-9326/ac9d4d), who found that subsetting the CMIP6 models according to how two thickness-based quantities (mean April SIT and the response of the minimum sea ice area to the mean April SIT) compared to the PIOMAS values substantially reduced the spread in projected sea ice extent and when the Arctic Ocean might become ice-free (depending on the warming scenario).\n```", "text_with_prefix": "EQC Quality Assessment: \"Evaluation of the Arctic sea ice thickness in the CMIP6 historical experiments\"\nDataset: projections-cmip6 [CDS]\nAspect: model-performance_q03 | Category: Climate_Projections\nSection: Evaluation of the Arctic sea ice thickness in the CMIP6 historical experiments > Quality assessment statement\n---\nThese are the key outcomes of this assessment\n\n- On average the Arctic sea ice thickness (SIT) from the CMIP6 ensemble is fairly unbiased against the altimeter-based estimates, with the exception of October. However the bias starts to become negative towards the end of the historical experiment in 2013 and 2014.\n\n- There are some spatial differences between the CMIP6 models and their ensemble mean. This is shown by the RMSE staying relatively constant throughout the whole evaluation period despite the low bias, indicating that the spatial biases are cancelling each other out. It is also shown by spatial maps of climatologies of the bias, which show that in the ENVISAT period the Beaufort Sea has a strong positive bias, while the Laptev, Kara and Barents Seas have more of a negative bias. In the CryoSat-2 period, the spatial distribution of the bias changes with the Beafort, Laptev, Kara and Barents Seas all have a positive bias, while the Greenland Sea, and the area extending from the north of Greenland to the north pole have a strong negative bias, making the mean bias negative.\n\n- We also compared maps of the climatology of the ensemble mean of the CMIP6 models during the CryoSat-2 period (2010-2020) to the climatology of the ensemble mean in the ENVISAT period (2002-2010). Spatially, the CMIP6 model mean has a fairly uniform drop in thickness when we move from the earlier satellite period to the later one, while the difference in the altimeter climatologies is more variable spatially.\n\n- Other authors have also looked at evaluating CMIP6 models sea ice thickness or volume. [West & Blockley (2024)](https://doi.org/10.5194/gmd-2024-121) compared a subset of the models (those that output heat flux diagnostics) to ice mass-balance buoys, finding that the CMIP6 models they looked at overestimated heat fluxes and thus sea ice melt and growth. A number of authors ([Davy & Outten, 2020](https://doi.org/10.1175/JCLI-D-19-0990.1); [Xu et al. , 2023](https://doi.org/10.3389/fmars.2023.1223772); [Watts et al. 2021](https://doi.org/10.1175/JCLI-D-20-0491.1)) found that the CMIP6 ensemble mean was similar to the PIOMAS (Pan-Arctic Ice-Ocean Modeling and Assimilation System) reanalysis, although the spread was quite large. [Davy & Outten (2020)](https://doi.org/10.1175/JCLI-D-19-0990.1) also included CMIP5 models in their comparison, noting that both the bias and spread have improved in CMIP6. The spatial patterns in the differences between altimeter and the CMIP6 ensemble mean that we found have also been found by others ([Watts et al., 2021](https://doi.org/10.1175/JCLI-D-20-0491.1); [Henke et al (2023)](https://doi.org/10.1080/15230430.2023.2271592)). In addition, [Kuang et al (2024)](https://doi.org/10.1016/j.accre.2024.06.008) calculated sea ice volume fluxes through six Arctic gateways (the most important being the Fram Strait), finding the multi-model mean also gave a reasonable representation of this flux.\n\n- A number of authors have also used CMIP6 sea ice thickness and volume when considering projections of sea ice. An interesting example of this is the article by [Zhou et al (2022)](https://doi.org/10.1088/1748-9326/ac9d4d), who found that subsetting the CMIP6 models according to how two thickness-based quantities (mean April SIT and the response of the minimum sea ice area to the mean April SIT) compared to the PIOMAS values substantially reduced the spread in projected sea ice extent and when the Arctic Ocean might become ice-free (depending on the warming scenario).\n```"} {"chunk_id": "climate_projections-cmip6_model-performance_q03__668692572cac", "report_id": "climate_projections-cmip6_model-performance_q03", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "model-performance_q03", "aspect_base": "model-performance", "category": "Climate_Projections", "match_confidence": "exact", "section": "Evaluation of the Arctic sea ice thickness in the CMIP6 historical experiments > Methodology", "title": "Evaluation of the Arctic sea ice thickness in the CMIP6 historical experiments", "chunk_index": 2, "token_count": 476, "text_raw": "For our ensemble we include each model in the CMIP6 historical experiment that outputs sea ice thickness, although users should note that choosing a different subset of models could give different results.\nThen, for each model in our ensemble, we calculate time series of bias and RMSE for the models compared to sea ice thickness estimates from altimetry - from the ENVISAT product between 2002 and April 2010, and from the CryoSat-2 product after October 2010. We then make plots of the time series showing the median and interquartile limits of the CMIP6 ensemble. There is a separate notebook which looks at these satellite products and addresses whether temporal trends can be seen in these data themselves.\n\nWe also calculate monthly thickness climatologies for the ENVISAT period and the CryoSat-2 period, and plot them to compare the spatial distribution of the thickness fields and the biases.\n\nFinally we calculate the difference between the climatologies for the CryoSat-2 and ENVISAT periods, to compare the change in thickness over time for the model and the satellites.\n\nThe \"Analysis and results\" section is organised as follows:\n\n**[](section-1)**\n\n**[](section-2)**\n\n**[](section-3)**\n\n        **[](section-3.1)**\n\n        Plot time series of the error metrics.\n\n        **[](section-3.2)**\n\n        Plot maps of climatologies and their biases for each winter month (October-April).\n\n        **[](section-3.3)**\n\n        Compare the change in thickness moving from the ENVISAT period to the CryoSat-2 period for both the satellites and the CMIP6 ensemble mean.", "text_with_prefix": "EQC Quality Assessment: \"Evaluation of the Arctic sea ice thickness in the CMIP6 historical experiments\"\nDataset: projections-cmip6 [CDS]\nAspect: model-performance_q03 | Category: Climate_Projections\nSection: Evaluation of the Arctic sea ice thickness in the CMIP6 historical experiments > Methodology\n---\nFor our ensemble we include each model in the CMIP6 historical experiment that outputs sea ice thickness, although users should note that choosing a different subset of models could give different results.\nThen, for each model in our ensemble, we calculate time series of bias and RMSE for the models compared to sea ice thickness estimates from altimetry - from the ENVISAT product between 2002 and April 2010, and from the CryoSat-2 product after October 2010. We then make plots of the time series showing the median and interquartile limits of the CMIP6 ensemble. There is a separate notebook which looks at these satellite products and addresses whether temporal trends can be seen in these data themselves.\n\nWe also calculate monthly thickness climatologies for the ENVISAT period and the CryoSat-2 period, and plot them to compare the spatial distribution of the thickness fields and the biases.\n\nFinally we calculate the difference between the climatologies for the CryoSat-2 and ENVISAT periods, to compare the change in thickness over time for the model and the satellites.\n\nThe \"Analysis and results\" section is organised as follows:\n\n**[](section-1)**\n\n**[](section-2)**\n\n**[](section-3)**\n\n        **[](section-3.1)**\n\n        Plot time series of the error metrics.\n\n        **[](section-3.2)**\n\n        Plot maps of climatologies and their biases for each winter month (October-April).\n\n        **[](section-3.3)**\n\n        Compare the change in thickness moving from the ENVISAT period to the CryoSat-2 period for both the satellites and the CMIP6 ensemble mean."} {"chunk_id": "climate_projections-cmip6_model-performance_q03__b05125c07471", "report_id": "climate_projections-cmip6_model-performance_q03", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "model-performance_q03", "aspect_base": "model-performance", "category": "Climate_Projections", "match_confidence": "exact", "section": "Evaluation of the Arctic sea ice thickness in the CMIP6 historical experiments > Analysis and results > 1. Import libraries, set parameters and definitions of functions > 1.2 Set parameters", "title": "Evaluation of the Arctic sea ice thickness in the CMIP6 historical experiments", "chunk_index": 3, "token_count": 137, "text_raw": "- Set the time period to be analysed with `year_start` and `year_stop`.\n- Set the models to be evaluated with the list `models`.\n\nChoose CMIP6 historical models\n\"taiesm1\", # very high values", "text_with_prefix": "EQC Quality Assessment: \"Evaluation of the Arctic sea ice thickness in the CMIP6 historical experiments\"\nDataset: projections-cmip6 [CDS]\nAspect: model-performance_q03 | Category: Climate_Projections\nSection: Evaluation of the Arctic sea ice thickness in the CMIP6 historical experiments > Analysis and results > 1. Import libraries, set parameters and definitions of functions > 1.2 Set parameters\n---\n- Set the time period to be analysed with `year_start` and `year_stop`.\n- Set the models to be evaluated with the list `models`.\n\nChoose CMIP6 historical models\n\"taiesm1\", # very high values"} {"chunk_id": "climate_projections-cmip6_model-performance_q03__d6f4ba6a912f", "report_id": "climate_projections-cmip6_model-performance_q03", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "model-performance_q03", "aspect_base": "model-performance", "category": "Climate_Projections", "match_confidence": "exact", "section": "Evaluation of the Arctic sea ice thickness in the CMIP6 historical experiments > Analysis and results > 1. Import libraries, set parameters and definitions of functions > 1.4 Functions to create the time series", "title": "Evaluation of the Arctic sea ice thickness in the CMIP6 historical experiments", "chunk_index": 4, "token_count": 302, "text_raw": "These functions are all applied to a single CMIP6 model. They are all applied to the downloaded data before caching.\n- `get_satellite_data` downloads the monthly satellite data.\n- `regrid` interpolates the model to the satellite grid.\n- `get_monthly_interpolated_data` is applied to both the model and satellite data, to take the monthly mean (the satellite data is daily, but this also forces all the models to have the same time coordinate) and interpolate to the satellite grid (only for model data). It also calculates the RMS error (only for satellite data).\n- `compare_model_vs_satellite` compares the sea ice thickness from a single CMIP6 model with satellite estimates, calculating error metrics like bias and RMSE in thickness.\n- `compute_sea_ice_thickness_diagnostics` downloads the satellite and model data and calls `compare_model_vs_satellite` to get a time series of error metrics.\n\nRemove nan columns\nHomogenize time\nGet variables\nCompute useful variables\nCompute output", "text_with_prefix": "EQC Quality Assessment: \"Evaluation of the Arctic sea ice thickness in the CMIP6 historical experiments\"\nDataset: projections-cmip6 [CDS]\nAspect: model-performance_q03 | Category: Climate_Projections\nSection: Evaluation of the Arctic sea ice thickness in the CMIP6 historical experiments > Analysis and results > 1. Import libraries, set parameters and definitions of functions > 1.4 Functions to create the time series\n---\nThese functions are all applied to a single CMIP6 model. They are all applied to the downloaded data before caching.\n- `get_satellite_data` downloads the monthly satellite data.\n- `regrid` interpolates the model to the satellite grid.\n- `get_monthly_interpolated_data` is applied to both the model and satellite data, to take the monthly mean (the satellite data is daily, but this also forces all the models to have the same time coordinate) and interpolate to the satellite grid (only for model data). It also calculates the RMS error (only for satellite data).\n- `compare_model_vs_satellite` compares the sea ice thickness from a single CMIP6 model with satellite estimates, calculating error metrics like bias and RMSE in thickness.\n- `compute_sea_ice_thickness_diagnostics` downloads the satellite and model data and calls `compare_model_vs_satellite` to get a time series of error metrics.\n\nRemove nan columns\nHomogenize time\nGet variables\nCompute useful variables\nCompute output"} {"chunk_id": "climate_projections-cmip6_model-performance_q03__bd11cb1603c4", "report_id": "climate_projections-cmip6_model-performance_q03", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "model-performance_q03", "aspect_base": "model-performance", "category": "Climate_Projections", "match_confidence": "exact", "section": "Evaluation of the Arctic sea ice thickness in the CMIP6 historical experiments > Analysis and results > 1. Import libraries, set parameters and definitions of functions > 1.5 Function to plot error time series", "title": "Evaluation of the Arctic sea ice thickness in the CMIP6 historical experiments", "chunk_index": 5, "token_count": 151, "text_raw": "- `plot_error_time_series` plots the spread of the model errors (compared to the observed sea ice concentration), and also the root mean square observation error for comparison.\n\nda = da.resample(time=\"1MS\").mean()\nmark approximate time when CryoSat-2 becomes available", "text_with_prefix": "EQC Quality Assessment: \"Evaluation of the Arctic sea ice thickness in the CMIP6 historical experiments\"\nDataset: projections-cmip6 [CDS]\nAspect: model-performance_q03 | Category: Climate_Projections\nSection: Evaluation of the Arctic sea ice thickness in the CMIP6 historical experiments > Analysis and results > 1. Import libraries, set parameters and definitions of functions > 1.5 Function to plot error time series\n---\n- `plot_error_time_series` plots the spread of the model errors (compared to the observed sea ice concentration), and also the root mean square observation error for comparison.\n\nda = da.resample(time=\"1MS\").mean()\nmark approximate time when CryoSat-2 becomes available"} {"chunk_id": "climate_projections-cmip6_model-performance_q03__282f6b567a43", "report_id": "climate_projections-cmip6_model-performance_q03", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "model-performance_q03", "aspect_base": "model-performance", "category": "Climate_Projections", "match_confidence": "exact", "section": "Evaluation of the Arctic sea ice thickness in the CMIP6 historical experiments > Analysis and results > 1. Import libraries, set parameters and definitions of functions > 1.6 Functions to create climatologies for plotting", "title": "Evaluation of the Arctic sea ice thickness in the CMIP6 historical experiments", "chunk_index": 6, "token_count": 186, "text_raw": "- `regridded_monthly_weighted_mean` gets monthly climatologies and then regrids them to the satellite grid. This is applied to the downloaded data before caching.\n- `postprocess_map_datasets` is applied after loading the cached data to rename some variables and dimensions in order to make them easier to use and plot.\n\nget monthly climatology\nget satellite grid\ninterpolate the model data onto the satellite grid\nreorder months", "text_with_prefix": "EQC Quality Assessment: \"Evaluation of the Arctic sea ice thickness in the CMIP6 historical experiments\"\nDataset: projections-cmip6 [CDS]\nAspect: model-performance_q03 | Category: Climate_Projections\nSection: Evaluation of the Arctic sea ice thickness in the CMIP6 historical experiments > Analysis and results > 1. Import libraries, set parameters and definitions of functions > 1.6 Functions to create climatologies for plotting\n---\n- `regridded_monthly_weighted_mean` gets monthly climatologies and then regrids them to the satellite grid. This is applied to the downloaded data before caching.\n- `postprocess_map_datasets` is applied after loading the cached data to rename some variables and dimensions in order to make them easier to use and plot.\n\nget monthly climatology\nget satellite grid\ninterpolate the model data onto the satellite grid\nreorder months"} {"chunk_id": "climate_projections-cmip6_model-performance_q03__0ecfb68f4bb7", "report_id": "climate_projections-cmip6_model-performance_q03", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "model-performance_q03", "aspect_base": "model-performance", "category": "Climate_Projections", "match_confidence": "exact", "section": "Evaluation of the Arctic sea ice thickness in the CMIP6 historical experiments > Analysis and results > 1. Import libraries, set parameters and definitions of functions > 1.7 Functions to plot climatologies", "title": "Evaluation of the Arctic sea ice thickness in the CMIP6 historical experiments", "chunk_index": 7, "token_count": 187, "text_raw": "- `plot_maps` plots a general climatology.\n- `plot_sithick_maps` plots thickness climatologies.\n- `plot_bias_maps` plots differences between the climatologies of the satellites and the CMIP6 ensemble mean.\n- `plot_change_maps` plots the difference between a climatology for the Cryosat-2 period (2010-2014) and the Envisat period (2002-2010).\n\n(section-2)=", "text_with_prefix": "EQC Quality Assessment: \"Evaluation of the Arctic sea ice thickness in the CMIP6 historical experiments\"\nDataset: projections-cmip6 [CDS]\nAspect: model-performance_q03 | Category: Climate_Projections\nSection: Evaluation of the Arctic sea ice thickness in the CMIP6 historical experiments > Analysis and results > 1. Import libraries, set parameters and definitions of functions > 1.7 Functions to plot climatologies\n---\n- `plot_maps` plots a general climatology.\n- `plot_sithick_maps` plots thickness climatologies.\n- `plot_bias_maps` plots differences between the climatologies of the satellites and the CMIP6 ensemble mean.\n- `plot_change_maps` plots the difference between a climatology for the Cryosat-2 period (2010-2014) and the Envisat period (2002-2010).\n\n(section-2)="} {"chunk_id": "climate_projections-cmip6_model-performance_q03__7afa6e0d2d00", "report_id": "climate_projections-cmip6_model-performance_q03", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "model-performance_q03", "aspect_base": "model-performance", "category": "Climate_Projections", "match_confidence": "exact", "section": "Evaluation of the Arctic sea ice thickness in the CMIP6 historical experiments > Analysis and results > 2. Download and transform the data > 2.2 Create the climatologies", "title": "Evaluation of the Arctic sea ice thickness in the CMIP6 historical experiments", "chunk_index": 8, "token_count": 139, "text_raw": "We first save monthly climatologies for the satellite data.\nThen we loop over all the models we are considering and apply `regridded_monthly_weighted_mean` to create climatologies, before taking their ensemble mean.\n\n(section-3)=", "text_with_prefix": "EQC Quality Assessment: \"Evaluation of the Arctic sea ice thickness in the CMIP6 historical experiments\"\nDataset: projections-cmip6 [CDS]\nAspect: model-performance_q03 | Category: Climate_Projections\nSection: Evaluation of the Arctic sea ice thickness in the CMIP6 historical experiments > Analysis and results > 2. Download and transform the data > 2.2 Create the climatologies\n---\nWe first save monthly climatologies for the satellite data.\nThen we loop over all the models we are considering and apply `regridded_monthly_weighted_mean` to create climatologies, before taking their ensemble mean.\n\n(section-3)="} {"chunk_id": "climate_projections-cmip6_model-performance_q03__76e8a5fb52be", "report_id": "climate_projections-cmip6_model-performance_q03", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "model-performance_q03", "aspect_base": "model-performance", "category": "Climate_Projections", "match_confidence": "exact", "section": "Evaluation of the Arctic sea ice thickness in the CMIP6 historical experiments > Analysis and results > 3. Results > 3.1 Time series of evaluation metrics", "title": "Evaluation of the Arctic sea ice thickness in the CMIP6 historical experiments", "chunk_index": 9, "token_count": 961, "text_raw": "Below we plot time series of the bias and RMSE in the sea ice thickness of the CMIP6 models compared to the altimetry data.\nWe have taken an ensemble of models (all those that output sea ice thickness), interpolated them to the altimeter grid, and calculated time series of area-weighted error metrics for each model. We have then plotted the median and the inter-quartile limits for the ensemble for each metric. Also note that the altimeter data is only available in the winter months (October - April), which is why the time series have gaps. The vertical red lines (mid-2010) divide the ENVISAT period (2002-2010) from the CryoSat-2 one (2010-2020 for the CDR, but the historical CMIP6 experiments finish in 2014).\n\nThe bias in the models are generally about the same level as the observation error level ($\\pm\\sigma_{\\rm obs}$, where $\\sigma_{\\rm obs}$ is the RMS error in the observations), and the ensemble median is close to zero for most of the time interval considered. Around 2014, however, the bias drops to about -0.25m compared to CryoSat-2.\n\nWithin each winter the bias is generally highest in October - about 0.3m - 0.5m for ENVISAT and also in the 2011 winter. The intra-winter variability in the bias seems to reduce after this however.\n\nIt is worth remembering that ENVISAT has its own bias which has been estimated to be about -0.6m so the models are probably also underestimating the thickness by that amount in the ENVISAT period [(CDS SIT Quality Assessment Report)](https://dast.copernicus-climate.eu/documents/satellite-sea-ice-thickness/level-3/v3-0/WP2-FDDP-2022-09_C3S2-Lot3_PQAR-of-v3.0-SeaIceThickness-products_v3.2_Final.pdf). The bias in CryoSat-2 has been estimated to be about -0.11m [(CDS SIT Quality Assessment Report)](https://dast.copernicus-climate.eu/documents/satellite-sea-ice-thickness/level-3/v3-0/WP2-FDDP-2022-09_C3S2-Lot3_PQAR-of-v3.0-SeaIceThickness-products_v3.2_Final.pdf) so it is a little surprising that the model bias does not become negative immediately upon the change in satellite but instead stays around zero from 2010 to 2013. However, the validation data which was used to estimate the bias was limited in space to 3 or 4 drilling sites in the Beafort Sea and some airborne electromagnetic tracks in the Beafort Sea and north of Greenland [(CDS SIT Quality Assessment Report)](https://dast.copernicus-climate.eu/documents/satellite-sea-ice-thickness/level-3/v3-0/WP2-FDDP-2022-09_C3S2-Lot3_PQAR-of-v3.0-SeaIceThickness-products_v3.2_Final.pdf), so the satellite biases may not apply to the whole Arctic.\n\nThe RMSE is also relatively constant over the entire period and can be about 50% above the plotted error range which is taken to be $(\\sigma_{\\rm obs}, \\sqrt{2}\\sigma_{\\rm obs})$. (This approximate error level comes from considering uncorrelated errors in the model and the observation time series when going through the calculation of the MSE - then the MSE is comparable to the error level if it is around $\\sigma_{\\rm obs}^2 + \\sigma_{\\rm mod}^2$, which becomes $2\\sigma_{\\rm obs}^2$ if we assume the model and observation errors are similar.) However in some years the median is near the error range and there is decent overlap between the IQL from the models and the error range. There is not such a clear intra-winter variability in this statistic as in the bias.\n\n(section-3.2)=", "text_with_prefix": "EQC Quality Assessment: \"Evaluation of the Arctic sea ice thickness in the CMIP6 historical experiments\"\nDataset: projections-cmip6 [CDS]\nAspect: model-performance_q03 | Category: Climate_Projections\nSection: Evaluation of the Arctic sea ice thickness in the CMIP6 historical experiments > Analysis and results > 3. Results > 3.1 Time series of evaluation metrics\n---\nBelow we plot time series of the bias and RMSE in the sea ice thickness of the CMIP6 models compared to the altimetry data.\nWe have taken an ensemble of models (all those that output sea ice thickness), interpolated them to the altimeter grid, and calculated time series of area-weighted error metrics for each model. We have then plotted the median and the inter-quartile limits for the ensemble for each metric. Also note that the altimeter data is only available in the winter months (October - April), which is why the time series have gaps. The vertical red lines (mid-2010) divide the ENVISAT period (2002-2010) from the CryoSat-2 one (2010-2020 for the CDR, but the historical CMIP6 experiments finish in 2014).\n\nThe bias in the models are generally about the same level as the observation error level ($\\pm\\sigma_{\\rm obs}$, where $\\sigma_{\\rm obs}$ is the RMS error in the observations), and the ensemble median is close to zero for most of the time interval considered. Around 2014, however, the bias drops to about -0.25m compared to CryoSat-2.\n\nWithin each winter the bias is generally highest in October - about 0.3m - 0.5m for ENVISAT and also in the 2011 winter. The intra-winter variability in the bias seems to reduce after this however.\n\nIt is worth remembering that ENVISAT has its own bias which has been estimated to be about -0.6m so the models are probably also underestimating the thickness by that amount in the ENVISAT period [(CDS SIT Quality Assessment Report)](https://dast.copernicus-climate.eu/documents/satellite-sea-ice-thickness/level-3/v3-0/WP2-FDDP-2022-09_C3S2-Lot3_PQAR-of-v3.0-SeaIceThickness-products_v3.2_Final.pdf). The bias in CryoSat-2 has been estimated to be about -0.11m [(CDS SIT Quality Assessment Report)](https://dast.copernicus-climate.eu/documents/satellite-sea-ice-thickness/level-3/v3-0/WP2-FDDP-2022-09_C3S2-Lot3_PQAR-of-v3.0-SeaIceThickness-products_v3.2_Final.pdf) so it is a little surprising that the model bias does not become negative immediately upon the change in satellite but instead stays around zero from 2010 to 2013. However, the validation data which was used to estimate the bias was limited in space to 3 or 4 drilling sites in the Beafort Sea and some airborne electromagnetic tracks in the Beafort Sea and north of Greenland [(CDS SIT Quality Assessment Report)](https://dast.copernicus-climate.eu/documents/satellite-sea-ice-thickness/level-3/v3-0/WP2-FDDP-2022-09_C3S2-Lot3_PQAR-of-v3.0-SeaIceThickness-products_v3.2_Final.pdf), so the satellite biases may not apply to the whole Arctic.\n\nThe RMSE is also relatively constant over the entire period and can be about 50% above the plotted error range which is taken to be $(\\sigma_{\\rm obs}, \\sqrt{2}\\sigma_{\\rm obs})$. (This approximate error level comes from considering uncorrelated errors in the model and the observation time series when going through the calculation of the MSE - then the MSE is comparable to the error level if it is around $\\sigma_{\\rm obs}^2 + \\sigma_{\\rm mod}^2$, which becomes $2\\sigma_{\\rm obs}^2$ if we assume the model and observation errors are similar.) However in some years the median is near the error range and there is decent overlap between the IQL from the models and the error range. There is not such a clear intra-winter variability in this statistic as in the bias.\n\n(section-3.2)="} {"chunk_id": "climate_projections-cmip6_model-performance_q03__f7a267ac7f9f", "report_id": "climate_projections-cmip6_model-performance_q03", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "model-performance_q03", "aspect_base": "model-performance", "category": "Climate_Projections", "match_confidence": "exact", "section": "Evaluation of the Arctic sea ice thickness in the CMIP6 historical experiments > Analysis and results > 3. Results > 3.2 Spatial distribution of errors", "title": "Evaluation of the Arctic sea ice thickness in the CMIP6 historical experiments", "chunk_index": 10, "token_count": 523, "text_raw": "Below we plot monthly climatologies calculated from 2002 to 2010, of the thicknesses from ENVISAT and CMIP6 (the ensemble mean over all models), and the bias (model - observations) in the CMIP6 mean. The ice in the Canadian Archipelago is consistently very thick in the models for all months. In October there is a strong positive bias in the Beaufort sea, which, while present in all other months, is slightly reduced. This bias also extends spatially into the Siberian Sea for all months, although it is not as strong there. In the Eastern Arctic (the Laptev, Kara and Barents Seas), the Labrador Sea and Hudson Bay there are consistent negative biases. In the Greenland Sea the thickness is mostly negatively biased but in October there is a reasonably sized region of positive bias as well. In other months there is some positive bias there as well but the regions are smaller. Bear in mind, however, that considering such a dynamic region on a monthly time-scale may limit the conclusions that can be made about the Greenland Sea.\n\nBelow we plot monthly climatologies calculated from 2010 to 2020, of the thicknesses from CryoSat-2 and CMIP6 (the ensemble mean over all models), and the bias (model - observations) in the CMIP6 mean. The ice in the Canadian Archipelago is still very thick in the models for all months. In October, there is a region of positive bias in the Beaufort, Siberian and Laptev seas. The extent of this region decreases as the months progress. There is a region of negative bias to the north of Greenland and the Canadian Archipelago, since the thicker ice off these coasts extends to the north pole in CryoSat-2 but is quite localised at the coasts in the models.\n\nThe ice in the Hudson Bay has a moderate negative bias, while the bias in the northern part of the Labrador Sea grows from around zero in November to about 25-50cm in April. Further south in the Labrador Sea there is a slight negative bias.\n\n(section-3.3)=", "text_with_prefix": "EQC Quality Assessment: \"Evaluation of the Arctic sea ice thickness in the CMIP6 historical experiments\"\nDataset: projections-cmip6 [CDS]\nAspect: model-performance_q03 | Category: Climate_Projections\nSection: Evaluation of the Arctic sea ice thickness in the CMIP6 historical experiments > Analysis and results > 3. Results > 3.2 Spatial distribution of errors\n---\nBelow we plot monthly climatologies calculated from 2002 to 2010, of the thicknesses from ENVISAT and CMIP6 (the ensemble mean over all models), and the bias (model - observations) in the CMIP6 mean. The ice in the Canadian Archipelago is consistently very thick in the models for all months. In October there is a strong positive bias in the Beaufort sea, which, while present in all other months, is slightly reduced. This bias also extends spatially into the Siberian Sea for all months, although it is not as strong there. In the Eastern Arctic (the Laptev, Kara and Barents Seas), the Labrador Sea and Hudson Bay there are consistent negative biases. In the Greenland Sea the thickness is mostly negatively biased but in October there is a reasonably sized region of positive bias as well. In other months there is some positive bias there as well but the regions are smaller. Bear in mind, however, that considering such a dynamic region on a monthly time-scale may limit the conclusions that can be made about the Greenland Sea.\n\nBelow we plot monthly climatologies calculated from 2010 to 2020, of the thicknesses from CryoSat-2 and CMIP6 (the ensemble mean over all models), and the bias (model - observations) in the CMIP6 mean. The ice in the Canadian Archipelago is still very thick in the models for all months. In October, there is a region of positive bias in the Beaufort, Siberian and Laptev seas. The extent of this region decreases as the months progress. There is a region of negative bias to the north of Greenland and the Canadian Archipelago, since the thicker ice off these coasts extends to the north pole in CryoSat-2 but is quite localised at the coasts in the models.\n\nThe ice in the Hudson Bay has a moderate negative bias, while the bias in the northern part of the Labrador Sea grows from around zero in November to about 25-50cm in April. Further south in the Labrador Sea there is a slight negative bias.\n\n(section-3.3)="} {"chunk_id": "climate_projections-cmip6_model-performance_q03__5352b378e0be", "report_id": "climate_projections-cmip6_model-performance_q03", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "model-performance_q03", "aspect_base": "model-performance", "category": "Climate_Projections", "match_confidence": "exact", "section": "Evaluation of the Arctic sea ice thickness in the CMIP6 historical experiments > Analysis and results > 3. Results > 3.3 Change in thickness over time", "title": "Evaluation of the Arctic sea ice thickness in the CMIP6 historical experiments", "chunk_index": 11, "token_count": 530, "text_raw": "Below we compare the differences (later - earlier) in the two climatologies for the ENVISAT period (2002-2010) and the overlap between the CryoSat-2 period and the historical experiment (2010-2024). The satellite products show a general decrease, with the exception of the Beaufort Sea in October, while the CMIP6 models show a near-uniform drop between the two periods. The magnitude of the drop is higher for the satellites however.\nIf we follow the estimated biases for each satellite, the change in the satellite thickness should drop by about 40cm, since ENVISAT has a larger negative bias than CryoSat-2. Hence we can theoretically trust decreases in the satellite thickness more than increases. However, as mentioned earlier, the validation data which was used to estimate the satellite biases was quite limited spatially [(CDS SIT Quality Assessment Report)](https://dast.copernicus-climate.eu/documents/satellite-sea-ice-thickness/level-3/v3-0/WP2-FDDP-2022-09_C3S2-Lot3_PQAR-of-v3.0-SeaIceThickness-products_v3.2_Final.pdf).\n\nThe short length of the time periods makes it difficult to see trends, which is also the conclusion of the notebook about the satellite sea ice thickness data, and the advice in the [CDS SIT Quality Assessment Report](https://dast.copernicus-climate.eu/documents/satellite-sea-ice-thickness/level-3/v3-0/WP2-FDDP-2022-09_C3S2-Lot3_PQAR-of-v3.0-SeaIceThickness-products_v3.2_Final.pdf). This is also why many authors use the PIOMAS reanalysis as a reference sea ice thickness product. [Davy & Outten (2020)](https://doi.org/10.1175/JCLI-D-19-0990.1) found that the CMIP6 ensemble mean had a similar trend to PIOMAS, showing thinning of sea ice and calculated over longer time intervals.", "text_with_prefix": "EQC Quality Assessment: \"Evaluation of the Arctic sea ice thickness in the CMIP6 historical experiments\"\nDataset: projections-cmip6 [CDS]\nAspect: model-performance_q03 | Category: Climate_Projections\nSection: Evaluation of the Arctic sea ice thickness in the CMIP6 historical experiments > Analysis and results > 3. Results > 3.3 Change in thickness over time\n---\nBelow we compare the differences (later - earlier) in the two climatologies for the ENVISAT period (2002-2010) and the overlap between the CryoSat-2 period and the historical experiment (2010-2024). The satellite products show a general decrease, with the exception of the Beaufort Sea in October, while the CMIP6 models show a near-uniform drop between the two periods. The magnitude of the drop is higher for the satellites however.\nIf we follow the estimated biases for each satellite, the change in the satellite thickness should drop by about 40cm, since ENVISAT has a larger negative bias than CryoSat-2. Hence we can theoretically trust decreases in the satellite thickness more than increases. However, as mentioned earlier, the validation data which was used to estimate the satellite biases was quite limited spatially [(CDS SIT Quality Assessment Report)](https://dast.copernicus-climate.eu/documents/satellite-sea-ice-thickness/level-3/v3-0/WP2-FDDP-2022-09_C3S2-Lot3_PQAR-of-v3.0-SeaIceThickness-products_v3.2_Final.pdf).\n\nThe short length of the time periods makes it difficult to see trends, which is also the conclusion of the notebook about the satellite sea ice thickness data, and the advice in the [CDS SIT Quality Assessment Report](https://dast.copernicus-climate.eu/documents/satellite-sea-ice-thickness/level-3/v3-0/WP2-FDDP-2022-09_C3S2-Lot3_PQAR-of-v3.0-SeaIceThickness-products_v3.2_Final.pdf). This is also why many authors use the PIOMAS reanalysis as a reference sea ice thickness product. [Davy & Outten (2020)](https://doi.org/10.1175/JCLI-D-19-0990.1) found that the CMIP6 ensemble mean had a similar trend to PIOMAS, showing thinning of sea ice and calculated over longer time intervals."} {"chunk_id": "climate_projections-cmip6_model-performance_q03__b2b58cb75d5d", "report_id": "climate_projections-cmip6_model-performance_q03", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "model-performance_q03", "aspect_base": "model-performance", "category": "Climate_Projections", "match_confidence": "exact", "section": "Evaluation of the Arctic sea ice thickness in the CMIP6 historical experiments > ℹ️ If you want to know more > Key resources", "title": "Evaluation of the Arctic sea ice thickness in the CMIP6 historical experiments", "chunk_index": 12, "token_count": 382, "text_raw": "Introductory sea ice materials:\n- [Role of sea ice in the climate](https://marine.copernicus.eu/explainers/why-ocean-important/sea-ice)\n- [Sea ice as an indicator of climate change](https://climate.copernicus.eu/climate-indicators/sea-ice#)\n- [Observing sea ice with satellites](https://www.metoffice.gov.uk/research/climate/cryosphere-oceans/sea-ice/measure)\n\nIntroductory CMIP6 materials:\n- [A short introduction to CMIP and CMIP6](https://www.wcrp-climate.org/wgcm-cmip/cmip-video)\n- [CMIP6: the next generation of climate models explained](https://www.carbonbrief.org/cmip6-the-next-generation-of-climate-models-explained/)\n\nCode libraries used:\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)", "text_with_prefix": "EQC Quality Assessment: \"Evaluation of the Arctic sea ice thickness in the CMIP6 historical experiments\"\nDataset: projections-cmip6 [CDS]\nAspect: model-performance_q03 | Category: Climate_Projections\nSection: Evaluation of the Arctic sea ice thickness in the CMIP6 historical experiments > ℹ️ If you want to know more > Key resources\n---\nIntroductory sea ice materials:\n- [Role of sea ice in the climate](https://marine.copernicus.eu/explainers/why-ocean-important/sea-ice)\n- [Sea ice as an indicator of climate change](https://climate.copernicus.eu/climate-indicators/sea-ice#)\n- [Observing sea ice with satellites](https://www.metoffice.gov.uk/research/climate/cryosphere-oceans/sea-ice/measure)\n\nIntroductory CMIP6 materials:\n- [A short introduction to CMIP and CMIP6](https://www.wcrp-climate.org/wgcm-cmip/cmip-video)\n- [CMIP6: the next generation of climate models explained](https://www.carbonbrief.org/cmip6-the-next-generation-of-climate-models-explained/)\n\nCode libraries used:\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)"} {"chunk_id": "climate_projections-cmip6_model-performance_q03__dfaa5d60d212", "report_id": "climate_projections-cmip6_model-performance_q03", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "model-performance_q03", "aspect_base": "model-performance", "category": "Climate_Projections", "match_confidence": "exact", "section": "Evaluation of the Arctic sea ice thickness in the CMIP6 historical experiments > ℹ️ If you want to know more > References", "title": "Evaluation of the Arctic sea ice thickness in the CMIP6 historical experiments", "chunk_index": 13, "token_count": 798, "text_raw": "1. [CDS Satellite Sea Ice Thickness Product Quality Assessment Report](https://dast.copernicus-climate.eu/documents/satellite-sea-ice-thickness/level-3/v3-0/WP2-FDDP-2022-09_C3S2-Lot3_PQAR-of-v3.0-SeaIceThickness-products_v3.2_Final.pdf)\n\n1. Davy, R., and S. Outten (2020). The Arctic Surface Climate in CMIP6: Status and Developments since CMIP5. J. Climate, 33, 8047–8068, https://doi.org/10.1175/JCLI-D-19-0990.1 \n\n1. Henke, M., F. Cassalho, T. Miesse, C. M. Ferreira, J. Zhang and T. M. Ravens (2023). Assessment of Arctic sea ice and surface climate conditions in nine CMIP6 climate models. Arctic, Antarctic, and Alpine Research, 55(1), https://doi.org/10.1080/15230430.2023.2271592 \n\n1. Kuang, H. Y., Sun, S. Z., Ye, Y. F., Wang, S. Y., Bi, H. B., Chen, Z. Q., & Cheng, X. (2024). An assessment of the CMIP6 performance in simulating Arctic sea ice volume flux via Fram Strait. Advances in Climate Change Research, 15(4), 584-595, https://doi.org/10.1016/j.accre.2024.06.008 \n\n1. Watts, M., W. Maslowski, Y. J. Lee, J. C. Kinney, and R. Osinski (2021). A Spatial Evaluation of Arctic Sea Ice and Regional Limitations in CMIP6 Historical Simulations. J. Climate, 34, 6399–6420, https://doi.org/10.1175/JCLI-D-20-0491.1 \n\n1. West, A. E., & Blockley, E. W. (2024). CMIP6 models overestimate sea ice melt, growth & conduction relative to ice mass balance buoy estimates. Geoscientific Model Development Discussions, 2024, 1-33, https://doi.org/10.5194/gmd-2024-121 \n\n1. Xu, M., & Li, J. (2023). Assessment of sea ice thickness simulations in the CMIP6 models with CICE components. Frontiers in Marine Science, 10, 1223772, https://doi.org/10.3389/fmars.2023.1223772 \n\n1. Zhou, X., Wang, B., & Huang, F. (2022). Evaluating sea ice thickness simulation is critical for projecting a summer ice-free Arctic Ocean. Environmental Research Letters, 17(11), 114033, https://doi.org/10.1088/1748-9326/ac9d4d ", "text_with_prefix": "EQC Quality Assessment: \"Evaluation of the Arctic sea ice thickness in the CMIP6 historical experiments\"\nDataset: projections-cmip6 [CDS]\nAspect: model-performance_q03 | Category: Climate_Projections\nSection: Evaluation of the Arctic sea ice thickness in the CMIP6 historical experiments > ℹ️ If you want to know more > References\n---\n1. [CDS Satellite Sea Ice Thickness Product Quality Assessment Report](https://dast.copernicus-climate.eu/documents/satellite-sea-ice-thickness/level-3/v3-0/WP2-FDDP-2022-09_C3S2-Lot3_PQAR-of-v3.0-SeaIceThickness-products_v3.2_Final.pdf)\n\n1. Davy, R., and S. Outten (2020). The Arctic Surface Climate in CMIP6: Status and Developments since CMIP5. J. Climate, 33, 8047–8068, https://doi.org/10.1175/JCLI-D-19-0990.1 \n\n1. Henke, M., F. Cassalho, T. Miesse, C. M. Ferreira, J. Zhang and T. M. Ravens (2023). Assessment of Arctic sea ice and surface climate conditions in nine CMIP6 climate models. Arctic, Antarctic, and Alpine Research, 55(1), https://doi.org/10.1080/15230430.2023.2271592 \n\n1. Kuang, H. Y., Sun, S. Z., Ye, Y. F., Wang, S. Y., Bi, H. B., Chen, Z. Q., & Cheng, X. (2024). An assessment of the CMIP6 performance in simulating Arctic sea ice volume flux via Fram Strait. Advances in Climate Change Research, 15(4), 584-595, https://doi.org/10.1016/j.accre.2024.06.008 \n\n1. Watts, M., W. Maslowski, Y. J. Lee, J. C. Kinney, and R. Osinski (2021). A Spatial Evaluation of Arctic Sea Ice and Regional Limitations in CMIP6 Historical Simulations. J. Climate, 34, 6399–6420, https://doi.org/10.1175/JCLI-D-20-0491.1 \n\n1. West, A. E., & Blockley, E. W. (2024). CMIP6 models overestimate sea ice melt, growth & conduction relative to ice mass balance buoy estimates. Geoscientific Model Development Discussions, 2024, 1-33, https://doi.org/10.5194/gmd-2024-121 \n\n1. Xu, M., & Li, J. (2023). Assessment of sea ice thickness simulations in the CMIP6 models with CICE components. Frontiers in Marine Science, 10, 1223772, https://doi.org/10.3389/fmars.2023.1223772 \n\n1. Zhou, X., Wang, B., & Huang, F. (2022). Evaluating sea ice thickness simulation is critical for projecting a summer ice-free Arctic Ocean. Environmental Research Letters, 17(11), 114033, https://doi.org/10.1088/1748-9326/ac9d4d "} {"chunk_id": "climate_projections-cmip6_validation_q10__f8161f4413ae", "report_id": "climate_projections-cmip6_validation_q10", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q10", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability", "title": "Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability", "chunk_index": 0, "token_count": 104, "text_raw": "Production date: 19-01-2025\n\nProduced by: CMCC foundation - Euro-Mediterranean Center on Climate Change. Albert Martinez Boti.", "text_with_prefix": "EQC Quality Assessment: \"Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q10 | Category: Climate_Projections\nSection: Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability\n---\nProduction date: 19-01-2025\n\nProduced by: CMCC foundation - Euro-Mediterranean Center on Climate Change. Albert Martinez Boti."} {"chunk_id": "climate_projections-cmip6_validation_q10__a9cc5c8c3261", "report_id": "climate_projections-cmip6_validation_q10", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q10", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability > Quality assessment question", "title": "Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability", "chunk_index": 1, "token_count": 511, "text_raw": "* **How well do CMIP6 projections represent historic inter-annual variability of precipitation across the Indian sub-continent? Are there trends affecting precipitation inter-annual variability?**\n\nMost research has focused on projected long-term changes in mean climate and extremes, while changes in variability have received less attention. However, understanding the evolution of precipitation variability is crucial for accurately modelling climate phenomena and quantify climate change impacts, particularly on water resources and related socioeconomic factors [[1]](https://doi.org/10.1111/gcb.12581),[[2]](https://doi.org/10.1016/j.wace.2020.100277). This notebook evaluates the capability of a subset of 16 [CMIP6](https://cds.climate.copernicus.eu/datasets/projections-cmip6?tab=overview) Global Climate Models (GCMs) to represent inter-annual variability in seasonal mean precipitation over the Indian subcontinent, employing variance and the coefficient of variation (standard deviation divided by the mean) as measures of variability.\n\nSystematic errors are assessed by comparing model outputs against the [ERA5](https://cds.climate.copernicus.eu/datasets/reanalysis-era5-complete?tab=doc) reanalysis, which serves as the reference dataset. The analysis focuses on JJAS (June, July, August and September), traditionally regarded as the Indian Summer Monsoon period, though the methodology is adaptable for annual or other seasonal analyses. Spatial patterns of the variance and the coefficient of variation are examined for JJAS mean precipitation over the historical period from 1940 to 2014. Additionally, the temporal evolution of the spatially-averaged variance and coefficient of variation, including trends, is analysed using a 30-year moving window and presented as time series.\n\nThe analyses presented in this notebook provide valuable insights into the reliability of a subset of CMIP6 models in capturing precipitation inter-annual variability. Understanding the performance of these models can help the water management sector to make more informed decisions and supports the use of CMIP6 Global Climate Models (GCMs) in designing strategies to adapt to projected changes in variability.", "text_with_prefix": "EQC Quality Assessment: \"Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q10 | Category: Climate_Projections\nSection: Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability > Quality assessment question\n---\n* **How well do CMIP6 projections represent historic inter-annual variability of precipitation across the Indian sub-continent? Are there trends affecting precipitation inter-annual variability?**\n\nMost research has focused on projected long-term changes in mean climate and extremes, while changes in variability have received less attention. However, understanding the evolution of precipitation variability is crucial for accurately modelling climate phenomena and quantify climate change impacts, particularly on water resources and related socioeconomic factors [[1]](https://doi.org/10.1111/gcb.12581),[[2]](https://doi.org/10.1016/j.wace.2020.100277). This notebook evaluates the capability of a subset of 16 [CMIP6](https://cds.climate.copernicus.eu/datasets/projections-cmip6?tab=overview) Global Climate Models (GCMs) to represent inter-annual variability in seasonal mean precipitation over the Indian subcontinent, employing variance and the coefficient of variation (standard deviation divided by the mean) as measures of variability.\n\nSystematic errors are assessed by comparing model outputs against the [ERA5](https://cds.climate.copernicus.eu/datasets/reanalysis-era5-complete?tab=doc) reanalysis, which serves as the reference dataset. The analysis focuses on JJAS (June, July, August and September), traditionally regarded as the Indian Summer Monsoon period, though the methodology is adaptable for annual or other seasonal analyses. Spatial patterns of the variance and the coefficient of variation are examined for JJAS mean precipitation over the historical period from 1940 to 2014. Additionally, the temporal evolution of the spatially-averaged variance and coefficient of variation, including trends, is analysed using a 30-year moving window and presented as time series.\n\nThe analyses presented in this notebook provide valuable insights into the reliability of a subset of CMIP6 models in capturing precipitation inter-annual variability. Understanding the performance of these models can help the water management sector to make more informed decisions and supports the use of CMIP6 Global Climate Models (GCMs) in designing strategies to adapt to projected changes in variability."} {"chunk_id": "climate_projections-cmip6_validation_q10__c1ae9cbb44b2", "report_id": "climate_projections-cmip6_validation_q10", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q10", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability > Quality assessment statement", "title": "Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability", "chunk_index": 2, "token_count": 566, "text_raw": "These are the key outcomes of this assessment\n\n* The choice of variability formulation is crucial when describing inter-annual variability during the summer Monsoon. The two formulations considered within this assessment reveal distinct spatial patterns.\n\n* Variance analyses reveal model-dependent differences in bias magnitude, sign, and spatial distribution. Despite these variations, all models show strong links to orography and high-precipitation regions, where biases remain significant due to high precipitation, even though they are not necessarily large relative to the mean. The coefficient of variation removes this effect by normalising the standard deviation (i.e., the square root of the variance) by the mean.\n\n* Model differences are less pronounced for the coefficient of variation than for variance. Most models underestimate it in the inland north-west, where seasonal mean precipitation is very low or near zero.\n\n* The 30-year time series of spatially averaged variance and coefficient of variation show a decreasing trend for ERA5, suggesting a decline in inter-annual variability over the historical period. This is evident from both variance magnitude—potentially influenced by the mean's decline [[3]](https://doi.org/10.1007/978-981-15-4327-2_3)—and the coefficient of variation, which removes the effect of the mean. However, the CMIP6 ensemble median for the subset of considered models shows the opposite trend.\n\n* Understanding and correcting the biases of CMIP6 projections of precipitation variability may increase confidence when analysing projections assessing the future inter-annual variability which can help on planning for water resource allocation, flood mitigation, and agricultural management, ensuring resilience against the challenges posed by a changing climate.\n\n```\n\nattachment:a920aca4-0f57-49cc-9f87-296209edb822.png\n---\nwidth: 1200px\nalt: Mean Bias variance \n---\nFig A. i. Variance of JJAS mean precipitation over the historical period (1940–2014). The layout includes data corresponding to: (a) ERA5, (b) the ensemble median (defined as the median of the variance values for the selected subset of models, calculated for each grid cell), (c) the bias of the ensemble median, and (d) the ensemble spread (calculated as the standard deviation of the variance values for the selected subset of models). Fig A.ii. Same as Fig. A.i. but for the coefficient of variation.\n```", "text_with_prefix": "EQC Quality Assessment: \"Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q10 | Category: Climate_Projections\nSection: Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability > Quality assessment statement\n---\nThese are the key outcomes of this assessment\n\n* The choice of variability formulation is crucial when describing inter-annual variability during the summer Monsoon. The two formulations considered within this assessment reveal distinct spatial patterns.\n\n* Variance analyses reveal model-dependent differences in bias magnitude, sign, and spatial distribution. Despite these variations, all models show strong links to orography and high-precipitation regions, where biases remain significant due to high precipitation, even though they are not necessarily large relative to the mean. The coefficient of variation removes this effect by normalising the standard deviation (i.e., the square root of the variance) by the mean.\n\n* Model differences are less pronounced for the coefficient of variation than for variance. Most models underestimate it in the inland north-west, where seasonal mean precipitation is very low or near zero.\n\n* The 30-year time series of spatially averaged variance and coefficient of variation show a decreasing trend for ERA5, suggesting a decline in inter-annual variability over the historical period. This is evident from both variance magnitude—potentially influenced by the mean's decline [[3]](https://doi.org/10.1007/978-981-15-4327-2_3)—and the coefficient of variation, which removes the effect of the mean. However, the CMIP6 ensemble median for the subset of considered models shows the opposite trend.\n\n* Understanding and correcting the biases of CMIP6 projections of precipitation variability may increase confidence when analysing projections assessing the future inter-annual variability which can help on planning for water resource allocation, flood mitigation, and agricultural management, ensuring resilience against the challenges posed by a changing climate.\n\n```\n\nattachment:a920aca4-0f57-49cc-9f87-296209edb822.png\n---\nwidth: 1200px\nalt: Mean Bias variance \n---\nFig A. i. Variance of JJAS mean precipitation over the historical period (1940–2014). The layout includes data corresponding to: (a) ERA5, (b) the ensemble median (defined as the median of the variance values for the selected subset of models, calculated for each grid cell), (c) the bias of the ensemble median, and (d) the ensemble spread (calculated as the standard deviation of the variance values for the selected subset of models). Fig A.ii. Same as Fig. A.i. but for the coefficient of variation.\n```"} {"chunk_id": "climate_projections-cmip6_validation_q10__2d058679145a", "report_id": "climate_projections-cmip6_validation_q10", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q10", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability > Methodology", "title": "Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability", "chunk_index": 3, "token_count": 802, "text_raw": "A subset of 16 models from the [CMIP6](https://cds.climate.copernicus.eu/datasets/projections-cmip6?tab=overview) project is used to calculate the variance and the coefficient of variation (standard deviation divided by the mean), which serve as formulations to characterise precipitation inter-annual variability in this notebook. Spatial patterns of variance and the coefficient of variation for seasonal mean precipitation, along with their associated biases (using [ERA5](https://cds.climate.copernicus.eu/datasets/reanalysis-era5-complete?tab=doc) as reference), are displayed for each model and the ensemble median (per grid cell). Additionally, the spatially averaged temporal evolution of the variance and coefficient of variation, including trends, is analysed using a 30-year moving window and presented as time series, following a methodology similar to [[4]](https://doi.org/10.1007/s00704-022-03972-2).\n\nThe analysis focuses on the Indian subcontinent during the historical period from 1940 to 2014 and examines the JJAS period, traditionally regarded as the Indian Summer Monsoon (or rainy) season, although the methodology can be adapted for other seasonal or annual periods.\n\nThe analysis and results follow the next outline:\n\n**[](climate_projections-cmip6_climate-and-weather-extremes_q01:section-1)**\n * [](climate_projections-cmip6_climate-and-weather-extremes_q01:section-1.1)\n * [](climate_projections-cmip6_climate-and-weather-extremes_q01:section-1.2)\n * [](climate_projections-cmip6_climate-and-weather-extremes_q01:section-1.3)\n * [](climate_projections-cmip6_climate-and-weather-extremes_q01:section-1.4)\n * [](climate_projections-cmip6_climate-and-weather-extremes_q01:section-1.5)\n * [](climate_projections-cmip6_climate-and-weather-extremes_q01:section-1.6)\n\n**[](climate_projections-cmip6_climate-and-weather-extremes_q01:section-2)**\n * [](climate_projections-cmip6_climate-and-weather-extremes_q01:section-2.1)\n * [](climate_projections-cmip6_climate-and-weather-extremes_q01:section-2.2)\n * [](climate_projections-cmip6_climate-and-weather-extremes_q01:section-2.3)\n\n**[](climate_projections-cmip6_climate-and-weather-extremes_q01:section-3)**\n * [](climate_projections-cmip6_climate-and-weather-extremes_q01:section-3.1)\n * [](climate_projections-cmip6_climate-and-weather-extremes_q01:section-3.2)\n * [](climate_projections-cmip6_climate-and-weather-extremes_q01:section-3.3)\n * [](climate_projections-cmip6_climate-and-weather-extremes_q01:section-3.4)\n * [](climate_projections-cmip6_climate-and-weather-extremes_q01:section-3.5)\n * [](climate_projections-cmip6_climate-and-weather-extremes_q01:section-3.6)", "text_with_prefix": "EQC Quality Assessment: \"Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q10 | Category: Climate_Projections\nSection: Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability > Methodology\n---\nA subset of 16 models from the [CMIP6](https://cds.climate.copernicus.eu/datasets/projections-cmip6?tab=overview) project is used to calculate the variance and the coefficient of variation (standard deviation divided by the mean), which serve as formulations to characterise precipitation inter-annual variability in this notebook. Spatial patterns of variance and the coefficient of variation for seasonal mean precipitation, along with their associated biases (using [ERA5](https://cds.climate.copernicus.eu/datasets/reanalysis-era5-complete?tab=doc) as reference), are displayed for each model and the ensemble median (per grid cell). Additionally, the spatially averaged temporal evolution of the variance and coefficient of variation, including trends, is analysed using a 30-year moving window and presented as time series, following a methodology similar to [[4]](https://doi.org/10.1007/s00704-022-03972-2).\n\nThe analysis focuses on the Indian subcontinent during the historical period from 1940 to 2014 and examines the JJAS period, traditionally regarded as the Indian Summer Monsoon (or rainy) season, although the methodology can be adapted for other seasonal or annual periods.\n\nThe analysis and results follow the next outline:\n\n**[](climate_projections-cmip6_climate-and-weather-extremes_q01:section-1)**\n * [](climate_projections-cmip6_climate-and-weather-extremes_q01:section-1.1)\n * [](climate_projections-cmip6_climate-and-weather-extremes_q01:section-1.2)\n * [](climate_projections-cmip6_climate-and-weather-extremes_q01:section-1.3)\n * [](climate_projections-cmip6_climate-and-weather-extremes_q01:section-1.4)\n * [](climate_projections-cmip6_climate-and-weather-extremes_q01:section-1.5)\n * [](climate_projections-cmip6_climate-and-weather-extremes_q01:section-1.6)\n\n**[](climate_projections-cmip6_climate-and-weather-extremes_q01:section-2)**\n * [](climate_projections-cmip6_climate-and-weather-extremes_q01:section-2.1)\n * [](climate_projections-cmip6_climate-and-weather-extremes_q01:section-2.2)\n * [](climate_projections-cmip6_climate-and-weather-extremes_q01:section-2.3)\n\n**[](climate_projections-cmip6_climate-and-weather-extremes_q01:section-3)**\n * [](climate_projections-cmip6_climate-and-weather-extremes_q01:section-3.1)\n * [](climate_projections-cmip6_climate-and-weather-extremes_q01:section-3.2)\n * [](climate_projections-cmip6_climate-and-weather-extremes_q01:section-3.3)\n * [](climate_projections-cmip6_climate-and-weather-extremes_q01:section-3.4)\n * [](climate_projections-cmip6_climate-and-weather-extremes_q01:section-3.5)\n * [](climate_projections-cmip6_climate-and-weather-extremes_q01:section-3.6)"} {"chunk_id": "climate_projections-cmip6_validation_q10__0c624b27bc03", "report_id": "climate_projections-cmip6_validation_q10", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q10", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability > Analysis and results > 1. Parameters, requests and functions definition > 1.2. Define Parameters", "title": "Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability", "chunk_index": 4, "token_count": 371, "text_raw": "In the \"Define Parameters\" section, various customisable options for the notebook are specified:\n\n- The initial and ending year used for the historical period can be specified by changing the parameters `year_start` and `year_stop` (1940-2014 is chosen).\n- The `timeseries` set the temporal period. For instance, selecting \"JJAS\" implies considering only the June, July, August, September season.\n- `area` allows specifying the geographical domain of interest. A domain including the Indian sub-continent has been selected.\n- The `interpolation_method` parameter allows selecting the interpolation method when regridding is performed. \n- The `chunk` selection allows the user to define if dividing into chunks when downloading the data on their local machine. Although it does not significantly affect the analysis, it is recommended to keep the default value for optimal performance.\n- `variable` and `collection_id` are not customisable for this assessment and are set to 'precipitation' and 'CMIP6'. Expert users can use this notebook as a guide to create similar analyses for other variables or model sets (such as CORDEX).\n\nChoose annual or seasonal timeseries\nInterpolation method\nArea to show\nChunks for download\nCollection id\n\n(climate_projections-cmip6_climate-and-weather-extremes_q01:section-1.3)=", "text_with_prefix": "EQC Quality Assessment: \"Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q10 | Category: Climate_Projections\nSection: Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability > Analysis and results > 1. Parameters, requests and functions definition > 1.2. Define Parameters\n---\nIn the \"Define Parameters\" section, various customisable options for the notebook are specified:\n\n- The initial and ending year used for the historical period can be specified by changing the parameters `year_start` and `year_stop` (1940-2014 is chosen).\n- The `timeseries` set the temporal period. For instance, selecting \"JJAS\" implies considering only the June, July, August, September season.\n- `area` allows specifying the geographical domain of interest. A domain including the Indian sub-continent has been selected.\n- The `interpolation_method` parameter allows selecting the interpolation method when regridding is performed. \n- The `chunk` selection allows the user to define if dividing into chunks when downloading the data on their local machine. Although it does not significantly affect the analysis, it is recommended to keep the default value for optimal performance.\n- `variable` and `collection_id` are not customisable for this assessment and are set to 'precipitation' and 'CMIP6'. Expert users can use this notebook as a guide to create similar analyses for other variables or model sets (such as CORDEX).\n\nChoose annual or seasonal timeseries\nInterpolation method\nArea to show\nChunks for download\nCollection id\n\n(climate_projections-cmip6_climate-and-weather-extremes_q01:section-1.3)="} {"chunk_id": "climate_projections-cmip6_validation_q10__2e8219ccd733", "report_id": "climate_projections-cmip6_validation_q10", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q10", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability > Analysis and results > 1. Parameters, requests and functions definition > 1.3. Define models", "title": "Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability", "chunk_index": 5, "token_count": 155, "text_raw": "The following climate analyses are performed considering a subset of GCMs from CMIP6.\n\nThe selected CMIP6 models have available both the historical and SSP5-8.5 experiments.\n\n(climate_projections-cmip6_climate-and-weather-extremes_q01:section-1.4)=", "text_with_prefix": "EQC Quality Assessment: \"Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q10 | Category: Climate_Projections\nSection: Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability > Analysis and results > 1. Parameters, requests and functions definition > 1.3. Define models\n---\nThe following climate analyses are performed considering a subset of GCMs from CMIP6.\n\nThe selected CMIP6 models have available both the historical and SSP5-8.5 experiments.\n\n(climate_projections-cmip6_climate-and-weather-extremes_q01:section-1.4)="} {"chunk_id": "climate_projections-cmip6_validation_q10__0849f835f608", "report_id": "climate_projections-cmip6_validation_q10", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q10", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability > Analysis and results > 1. Parameters, requests and functions definition > 1.4. Define ERA5 request", "title": "Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability", "chunk_index": 6, "token_count": 150, "text_raw": "Within this notebook, ERA5 serves as the reference product. In this section, we set the required parameters for the cds-api data-request of ERA5.\n\n(climate_projections-cmip6_climate-and-weather-extremes_q01:section-1.5)=", "text_with_prefix": "EQC Quality Assessment: \"Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q10 | Category: Climate_Projections\nSection: Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability > Analysis and results > 1. Parameters, requests and functions definition > 1.4. Define ERA5 request\n---\nWithin this notebook, ERA5 serves as the reference product. In this section, we set the required parameters for the cds-api data-request of ERA5.\n\n(climate_projections-cmip6_climate-and-weather-extremes_q01:section-1.5)="} {"chunk_id": "climate_projections-cmip6_validation_q10__954a7ac558a0", "report_id": "climate_projections-cmip6_validation_q10", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q10", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability > Analysis and results > 1. Parameters, requests and functions definition > 1.5. Define model requests", "title": "Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability", "chunk_index": 7, "token_count": 133, "text_raw": "In this section we set the required parameters for the cds-api data-request.\n\n(climate_projections-cmip6_climate-and-weather-extremes_q01:section-1.6)=", "text_with_prefix": "EQC Quality Assessment: \"Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q10 | Category: Climate_Projections\nSection: Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability > Analysis and results > 1. Parameters, requests and functions definition > 1.5. Define model requests\n---\nIn this section we set the required parameters for the cds-api data-request.\n\n(climate_projections-cmip6_climate-and-weather-extremes_q01:section-1.6)="} {"chunk_id": "climate_projections-cmip6_validation_q10__e01b9a344af1", "report_id": "climate_projections-cmip6_validation_q10", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q10", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability > Analysis and results > 1. Parameters, requests and functions definition > 1.6. Functions to cache", "title": "Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability", "chunk_index": 8, "token_count": 372, "text_raw": "In this section, functions that will be executed in the caching phase are defined. Caching is the process of storing copies of files in a temporary storage location, so that they can be accessed more quickly. This process also checks if the user has already downloaded a file, avoiding redundant downloads.\n\nFunctions description:\n\n- `get_grid_out` and `add_bounds` ensure the regrid is performed properly.\n\n- The `select_timeseries` function subsets the dataset based on the chosen `timeseries` parameter.\n\n- The `compute_rolling_variance` function calculates the rolling variance and coefficient of variation using a 30-year window (which are needed when displaying the time series afterwards). It also computes the variance and coefficient of variation for the entire period.\n\n- The `compute_interannual_variance` function selects the season using the `select_timeseries` function. It then computes the rolling variance (and coefficient of variation) over the historical period (1940–2014) by calling `compute_rolling_variance`. The window width is 30 years, with a step size of one year. The variance and coefficient of variation is also calculated for the whole historical period.\n\nSelect the time series\nGet seasonal/annual means\nChange units\nCompute variances\nHandle grid output if requested\n\n(climate_projections-cmip6_climate-and-weather-extremes_q01:section-2)=", "text_with_prefix": "EQC Quality Assessment: \"Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q10 | Category: Climate_Projections\nSection: Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability > Analysis and results > 1. Parameters, requests and functions definition > 1.6. Functions to cache\n---\nIn this section, functions that will be executed in the caching phase are defined. Caching is the process of storing copies of files in a temporary storage location, so that they can be accessed more quickly. This process also checks if the user has already downloaded a file, avoiding redundant downloads.\n\nFunctions description:\n\n- `get_grid_out` and `add_bounds` ensure the regrid is performed properly.\n\n- The `select_timeseries` function subsets the dataset based on the chosen `timeseries` parameter.\n\n- The `compute_rolling_variance` function calculates the rolling variance and coefficient of variation using a 30-year window (which are needed when displaying the time series afterwards). It also computes the variance and coefficient of variation for the entire period.\n\n- The `compute_interannual_variance` function selects the season using the `select_timeseries` function. It then computes the rolling variance (and coefficient of variation) over the historical period (1940–2014) by calling `compute_rolling_variance`. The window width is 30 years, with a step size of one year. The variance and coefficient of variation is also calculated for the whole historical period.\n\nSelect the time series\nGet seasonal/annual means\nChange units\nCompute variances\nHandle grid output if requested\n\n(climate_projections-cmip6_climate-and-weather-extremes_q01:section-2)="} {"chunk_id": "climate_projections-cmip6_validation_q10__f8865b66ba52", "report_id": "climate_projections-cmip6_validation_q10", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q10", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability > Analysis and results > 2. Downloading and processing > 2.1. Download and transform ERA5", "title": "Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability", "chunk_index": 9, "token_count": 206, "text_raw": "In this section, the `download.download_and_transform` function from the 'c3s_eqc_automatic_quality_control' package is used to download ERA5 reference monthly data, select the season (e.g., \"JJAS\" in this case) and compute the variance and coefficient of variation for the entire historical period, as well as for a 30-year rolling window. The results are then cached to prevent redundant downloads and processing.\n\n(climate_projections-cmip6_climate-and-weather-extremes_q01:section-2.2)=", "text_with_prefix": "EQC Quality Assessment: \"Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q10 | Category: Climate_Projections\nSection: Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability > Analysis and results > 2. Downloading and processing > 2.1. Download and transform ERA5\n---\nIn this section, the `download.download_and_transform` function from the 'c3s_eqc_automatic_quality_control' package is used to download ERA5 reference monthly data, select the season (e.g., \"JJAS\" in this case) and compute the variance and coefficient of variation for the entire historical period, as well as for a 30-year rolling window. The results are then cached to prevent redundant downloads and processing.\n\n(climate_projections-cmip6_climate-and-weather-extremes_q01:section-2.2)="} {"chunk_id": "climate_projections-cmip6_validation_q10__382f796c0b15", "report_id": "climate_projections-cmip6_validation_q10", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q10", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability > Analysis and results > 2. Downloading and processing > 2.2. Download and transform models", "title": "Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability", "chunk_index": 10, "token_count": 369, "text_raw": "In this section, the `download.download_and_transform` function from the 'c3s_eqc_automatic_quality_control' package is employed to download daily data from the CMIP6 models, select the season (\"JJAS\" in this example), compute the variance and coefficient of variation for the entire historical period, as well as for a 30-year rolling windows, interpolate to the ERA5 grid (only when specified; otherwise, the original model grid is maintained), and cache the results to avoid redundant downloads and processing.\n\nOriginal model\nInterpolated model\n\n```text\nmodel='access_cm2'\nmodel='bcc_csm2_mr'\nmodel='cmcc_esm2'\nmodel='cnrm_cm6_1_hr'\nmodel='cnrm_esm2_1'\nmodel='ec_earth3_cc'\nmodel='gfdl_esm4'\nmodel='inm_cm4_8'\nmodel='inm_cm5_0'\nmodel='mpi_esm1_2_lr'\nmodel='miroc6'\nmodel='miroc_es2l'\nmodel='mri_esm2_0'\nmodel='noresm2_mm'\nmodel='nesm3'\nmodel='ukesm1_0_ll'\n```\n\n(climate_projections-cmip6_climate-and-weather-extremes_q01:section-2.3)=", "text_with_prefix": "EQC Quality Assessment: \"Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q10 | Category: Climate_Projections\nSection: Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability > Analysis and results > 2. Downloading and processing > 2.2. Download and transform models\n---\nIn this section, the `download.download_and_transform` function from the 'c3s_eqc_automatic_quality_control' package is employed to download daily data from the CMIP6 models, select the season (\"JJAS\" in this example), compute the variance and coefficient of variation for the entire historical period, as well as for a 30-year rolling windows, interpolate to the ERA5 grid (only when specified; otherwise, the original model grid is maintained), and cache the results to avoid redundant downloads and processing.\n\nOriginal model\nInterpolated model\n\n```text\nmodel='access_cm2'\nmodel='bcc_csm2_mr'\nmodel='cmcc_esm2'\nmodel='cnrm_cm6_1_hr'\nmodel='cnrm_esm2_1'\nmodel='ec_earth3_cc'\nmodel='gfdl_esm4'\nmodel='inm_cm4_8'\nmodel='inm_cm5_0'\nmodel='mpi_esm1_2_lr'\nmodel='miroc6'\nmodel='miroc_es2l'\nmodel='mri_esm2_0'\nmodel='noresm2_mm'\nmodel='nesm3'\nmodel='ukesm1_0_ll'\n```\n\n(climate_projections-cmip6_climate-and-weather-extremes_q01:section-2.3)="} {"chunk_id": "climate_projections-cmip6_validation_q10__37f4ee62693a", "report_id": "climate_projections-cmip6_validation_q10", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q10", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability > Analysis and results > 2. Downloading and processing > 2.3. Change some attributes", "title": "Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability", "chunk_index": 11, "token_count": 132, "text_raw": "Drop 'bnds' dimension if it exists for each dataset and for the interpolated dataset\n\n(climate_projections-cmip6_climate-and-weather-extremes_q01:section-3)=", "text_with_prefix": "EQC Quality Assessment: \"Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q10 | Category: Climate_Projections\nSection: Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability > Analysis and results > 2. Downloading and processing > 2.3. Change some attributes\n---\nDrop 'bnds' dimension if it exists for each dataset and for the interpolated dataset\n\n(climate_projections-cmip6_climate-and-weather-extremes_q01:section-3)="} {"chunk_id": "climate_projections-cmip6_validation_q10__5e3c0143ffd9", "report_id": "climate_projections-cmip6_validation_q10", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q10", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability > Analysis and results > 3. Plot and describe results", "title": "Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability", "chunk_index": 12, "token_count": 344, "text_raw": "This section will display the following results:\n\n- Maps showing the spatial distribution of the **variance of JJAS mean precipitation** calculated for the whole historical period 1940-2014 comparing ERA5 and the ensemble median (defined as the median of the variance values for the selected subset of models, calculated for each grid cell). The layout include ERA5, the ensemble median, the ensemble median bias and the ensemble spread (derived as the standard deviation of the variance for the selected subset of models). The same exact layout of maps is displayed for the coefficient of variation.\n\n- Maps showing the spatial distribution of the **variance of JJAS mean precipitation** calculated for the whole historical period 1940-2014 for each model. The same exact layout of maps is displayed for the coefficient of variation.\n\n- **Bias maps of the variance and coefficient of variation of JJAS mean precipitation** calculated for the whole historical period 1940–2014 for each model.\n\n- **Timeseries** showing the evolution of the 30-year variance and coefficient of variation (standard deviation divided by the mean) of JJAS mean precipitation over the historical period, including trends and inter-model spread.\n\n(climate_projections-cmip6_climate-and-weather-extremes_q01:section-3.1)=", "text_with_prefix": "EQC Quality Assessment: \"Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q10 | Category: Climate_Projections\nSection: Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability > Analysis and results > 3. Plot and describe results\n---\nThis section will display the following results:\n\n- Maps showing the spatial distribution of the **variance of JJAS mean precipitation** calculated for the whole historical period 1940-2014 comparing ERA5 and the ensemble median (defined as the median of the variance values for the selected subset of models, calculated for each grid cell). The layout include ERA5, the ensemble median, the ensemble median bias and the ensemble spread (derived as the standard deviation of the variance for the selected subset of models). The same exact layout of maps is displayed for the coefficient of variation.\n\n- Maps showing the spatial distribution of the **variance of JJAS mean precipitation** calculated for the whole historical period 1940-2014 for each model. The same exact layout of maps is displayed for the coefficient of variation.\n\n- **Bias maps of the variance and coefficient of variation of JJAS mean precipitation** calculated for the whole historical period 1940–2014 for each model.\n\n- **Timeseries** showing the evolution of the 30-year variance and coefficient of variation (standard deviation divided by the mean) of JJAS mean precipitation over the historical period, including trends and inter-model spread.\n\n(climate_projections-cmip6_climate-and-weather-extremes_q01:section-3.1)="} {"chunk_id": "climate_projections-cmip6_validation_q10__8b5c2cf6ebae", "report_id": "climate_projections-cmip6_validation_q10", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q10", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability > Analysis and results > 3. Plot and describe results > 3.1. Define plotting functions", "title": "Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability", "chunk_index": 13, "token_count": 276, "text_raw": "The functions presented here are used to plot layouts of the variance and coefficient of variation of seasonal mean precipitation over the whole historical period.\n\nThree layout types can be displayed, depending on the chosen function:\n\n1. Layout including the reference ERA5 product, the ensemble median, the bias of the ensemble median, and the ensemble spread: `plot_ensemble()` is used.\n2. Layout including every model: `plot_models()` is employed.\n3. Layout including the bias of every model: `plot_models()` is used again.\n\nDefine function to plot the caption of the figures (for the ensemble case)\nAdd each line to the figure\nend captioning\nDefine function to plot the caption of the figures (for the individual models case)\nAdd caption to the figure\nAdd each line to the figure\n\n(climate_projections-cmip6_climate-and-weather-extremes_q01:section-3.2)=", "text_with_prefix": "EQC Quality Assessment: \"Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q10 | Category: Climate_Projections\nSection: Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability > Analysis and results > 3. Plot and describe results > 3.1. Define plotting functions\n---\nThe functions presented here are used to plot layouts of the variance and coefficient of variation of seasonal mean precipitation over the whole historical period.\n\nThree layout types can be displayed, depending on the chosen function:\n\n1. Layout including the reference ERA5 product, the ensemble median, the bias of the ensemble median, and the ensemble spread: `plot_ensemble()` is used.\n2. Layout including every model: `plot_models()` is employed.\n3. Layout including the bias of every model: `plot_models()` is used again.\n\nDefine function to plot the caption of the figures (for the ensemble case)\nAdd each line to the figure\nend captioning\nDefine function to plot the caption of the figures (for the individual models case)\nAdd caption to the figure\nAdd each line to the figure\n\n(climate_projections-cmip6_climate-and-weather-extremes_q01:section-3.2)="} {"chunk_id": "climate_projections-cmip6_validation_q10__c546d6a8842d", "report_id": "climate_projections-cmip6_validation_q10", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q10", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability > Analysis and results > 3. Plot and describe results > 3.2. Plot ensemble maps", "title": "Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability", "chunk_index": 14, "token_count": 264, "text_raw": "In this section, we invoke the `plot_ensemble()` function to visualise the variance values for JJAS mean precipitation over the period 1940–2014 for: (a) the reference ERA5 product, (b) the ensemble median (defined as the median of the variance values for the selected subset of models, calculated for each grid cell), (c) the bias of the ensemble median, and (d) the ensemble spread (calculated as the standard deviation of the variance for the selected subset of models). The same exact layout of maps is displayed for the coefficient of variation.\n\nChange default colorbars\nFig number counter\nCommon title\nget the variance of the whole period\nShow the plot\nIncrement figure number\n\n(climate_projections-cmip6_climate-and-weather-extremes_q01:section-3.3)=", "text_with_prefix": "EQC Quality Assessment: \"Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q10 | Category: Climate_Projections\nSection: Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability > Analysis and results > 3. Plot and describe results > 3.2. Plot ensemble maps\n---\nIn this section, we invoke the `plot_ensemble()` function to visualise the variance values for JJAS mean precipitation over the period 1940–2014 for: (a) the reference ERA5 product, (b) the ensemble median (defined as the median of the variance values for the selected subset of models, calculated for each grid cell), (c) the bias of the ensemble median, and (d) the ensemble spread (calculated as the standard deviation of the variance for the selected subset of models). The same exact layout of maps is displayed for the coefficient of variation.\n\nChange default colorbars\nFig number counter\nCommon title\nget the variance of the whole period\nShow the plot\nIncrement figure number\n\n(climate_projections-cmip6_climate-and-weather-extremes_q01:section-3.3)="} {"chunk_id": "climate_projections-cmip6_validation_q10__e1edf52530dc", "report_id": "climate_projections-cmip6_validation_q10", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q10", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability > Analysis and results > 3. Plot and describe results > 3.3. Plot model maps", "title": "Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability", "chunk_index": 15, "token_count": 188, "text_raw": "In this section, we invoke the `plot_models()` function to visualise the variance and the coefficient of variation of JJAS mean precipitation over the period 1940–2014 for every model individually. Note that the model data used in this section maintains its original grid.\n\nSelect the mean values of the variances\nShow the plot\nIncrement figure number\n\n(climate_projections-cmip6_climate-and-weather-extremes_q01:section-3.4)=", "text_with_prefix": "EQC Quality Assessment: \"Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q10 | Category: Climate_Projections\nSection: Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability > Analysis and results > 3. Plot and describe results > 3.3. Plot model maps\n---\nIn this section, we invoke the `plot_models()` function to visualise the variance and the coefficient of variation of JJAS mean precipitation over the period 1940–2014 for every model individually. Note that the model data used in this section maintains its original grid.\n\nSelect the mean values of the variances\nShow the plot\nIncrement figure number\n\n(climate_projections-cmip6_climate-and-weather-extremes_q01:section-3.4)="} {"chunk_id": "climate_projections-cmip6_validation_q10__1f4175efd770", "report_id": "climate_projections-cmip6_validation_q10", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q10", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability > Analysis and results > 3. Plot and describe results > 3.4. Plot bias maps", "title": "Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability", "chunk_index": 16, "token_count": 179, "text_raw": "In this section, we invoke the `plot_models()` function to visualise the bias of the variance and the coefficient of variation of JJAS mean precipitation over the period 1940–2014 for every model individually. Note that the model data used in this section has previously been interpolated to the ERA5 grid.\n\n(climate_projections-cmip6_climate-and-weather-extremes_q01:section-3.5)=", "text_with_prefix": "EQC Quality Assessment: \"Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q10 | Category: Climate_Projections\nSection: Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability > Analysis and results > 3. Plot and describe results > 3.4. Plot bias maps\n---\nIn this section, we invoke the `plot_models()` function to visualise the bias of the variance and the coefficient of variation of JJAS mean precipitation over the period 1940–2014 for every model individually. Note that the model data used in this section has previously been interpolated to the ERA5 grid.\n\n(climate_projections-cmip6_climate-and-weather-extremes_q01:section-3.5)="} {"chunk_id": "climate_projections-cmip6_validation_q10__93fc6df3093c", "report_id": "climate_projections-cmip6_validation_q10", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q10", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability > Analysis and results > 3. Plot and describe results > 3.5. Timeseries of the coefficient of variation", "title": "Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability", "chunk_index": 17, "token_count": 413, "text_raw": "This section examines the time series of spatially averaged variance and coefficient of variation, calculated from JJAS mean precipitation over the historical period using a 30-year window. The analysis compares the CMIP6 ensemble median, ERA5, and individual models, highlighting trends for both the CMIP6 ensemble median and ERA5. It also displays the ensemble spread, along with the slope values of the trends for ERA5 and the CMIP6 ensemble median.\n\nDefine the colors\nCutout region\nGet the spatial mean for ERA5\nGet the spatial mean for CMIP6, preserving each model\nCalculate the median, ensemble mean, and standard deviation\nInitialize list to store trend information\nPlot individual model series for each variance\n\n
\n
\n

Fig 7. Time series of the spatially averaged values for the chosen variability formulations: (a) variance and (b) coefficient of variation for the JJAS mean precipitation over the period 1940–2014. The variance and coefficient of variation have been calculated using a 30-year window. Each time series displays the CMIP6 ensemble median (green), ERA5 (black), and individual models (grey), with dashed lines indicating the trend for the CMIP6 ensemble median and ERA5. The shaded area represents the ensemble spread. A grey box at the bottom displays the trend slope.

\n
\n\n(climate_projections-cmip6_climate-and-weather-extremes_q01:section-3.6)=", "text_with_prefix": "EQC Quality Assessment: \"Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q10 | Category: Climate_Projections\nSection: Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability > Analysis and results > 3. Plot and describe results > 3.5. Timeseries of the coefficient of variation\n---\nThis section examines the time series of spatially averaged variance and coefficient of variation, calculated from JJAS mean precipitation over the historical period using a 30-year window. The analysis compares the CMIP6 ensemble median, ERA5, and individual models, highlighting trends for both the CMIP6 ensemble median and ERA5. It also displays the ensemble spread, along with the slope values of the trends for ERA5 and the CMIP6 ensemble median.\n\nDefine the colors\nCutout region\nGet the spatial mean for ERA5\nGet the spatial mean for CMIP6, preserving each model\nCalculate the median, ensemble mean, and standard deviation\nInitialize list to store trend information\nPlot individual model series for each variance\n\n
\n
\n

Fig 7. Time series of the spatially averaged values for the chosen variability formulations: (a) variance and (b) coefficient of variation for the JJAS mean precipitation over the period 1940–2014. The variance and coefficient of variation have been calculated using a 30-year window. Each time series displays the CMIP6 ensemble median (green), ERA5 (black), and individual models (grey), with dashed lines indicating the trend for the CMIP6 ensemble median and ERA5. The shaded area represents the ensemble spread. A grey box at the bottom displays the trend slope.

\n
\n\n(climate_projections-cmip6_climate-and-weather-extremes_q01:section-3.6)="} {"chunk_id": "climate_projections-cmip6_validation_q10__feb0501d4ccb", "report_id": "climate_projections-cmip6_validation_q10", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q10", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability > Analysis and results > 3. Plot and describe results > 3.6. Results summary and discussion", "title": "Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability", "chunk_index": 18, "token_count": 1056, "text_raw": "* Substantial differences in spatial patterns can be observed depending on the formulation used to represent inter-annual variability during the Monsoon season (JJAS). This notebook uses two formulations: variance and the coefficient of variation (standard deviation divided by the mean).\n\n* Large differences in sign, magnitude, and spatial patterns of the biase are observed in the analyses using variance to represent inter-annual variability. Despite the differences across the subset of considered CMIP6 models, there is agreement in showing strong dependencies on orography and regions with high precipitation. These regions exhibit high biases due to elevated mean JJAS precipitation, although these biases are not necessarily large relative to the mean. Using the coefficient of variation eliminates this effect and may be more useful for assessing variability normalised to the mean amount of precipitation during the Monsoon period.\n\n* Generally, there are smaller differences depending on the model selection for the coefficient of variation, compared to the differences observed for the variance formulation. Inland north-western regions show large negative biases. These regions have, on average, near-zero or very low seasonal averages. One hypothesis to explain this may be that unusually wet seasons are underestimated in comparison to ERA5.\n\n* The time series displaying the evolution of variance and coefficient of variation (using a 30-year window) show a decline for ERA5 in the considered region. Based on the calculated trend, an approximate 50% reduction is estimated when comparing the start and end of the period for the variance formulation. This decrease is smaller when considering the coefficient of variation (11%). The ensemble median trend for the subset of considered CMIP6 models shows an increase of around 7% for the variance and 6% for the coefficient of variation over the same period, which differs from the trends observed in ERA5. It is important to note that the region studied is extensive. The time series represent spatially averaged values, and their results should be interpreted with care, as focusing on sub-regions within the domain may yield significantly different outcomes. Additionally, CMIP6 models exhibit notable inter-model spread, and a larger ensemble could also influence the results.\n\n* The time series results suggest that year-to-year variability in monsoon precipitation has become less pronounced according to ERA5. This is evident not only from the variance magnitude, which can be influenced by the evolution of the mean (shown to decrease during this period [[3]](https://doi.org/10.1007/978-981-15-4327-2_3)), but also from the coefficient of variation, which removes this dependency by normalising the standard deviation with the mean.\n\n* What do the results mean for users?\n * **Importance of the selection of the variability formulation**. The coefficient of variation has less dependency on orography because it is normalised by the mean. Moreover, the differences across the considered subset of models in the spatial pattern of the bias for the coefficient of variation appear to be smaller than in the case of variance. However, special attention needs to be given when using the coefficient of variation for the inland north-western region (where the seasonal mean precipitation is usually very low or near zero).\n * **ERA5 has biases** [[5]](https://doi.org/10.1002/qj.3616)[[6]](https://doi.org/10.3390/atmos12111462). This assessment uses it because a longer period is available and allows a more robust analysis of the temporal evolution of the variability. However the user may **consider using other reference climate products such as the [GPCP](https://cds.climate.copernicus.eu/datasets/satellite-precipitation?tab=overview)**. This could lead to different results.\n * ERA5 shows a **decrease in inter-annual variability over the historical period for the considered region**, while the CMIP6 ensemble median indicates a slight increase. It is **important to consider a larger ensemble and note that the time series in this assessment represent spatially averaged values**, which may yield different results when focusing on other sub-regions.\n * Given the challenges in replicating the spatial patterns for variance and the coefficient of variation, as well as the difficulties in representing its temporal evolution, **bias correction of CMIP6 projections may be essential for effectively using future projections of inter-annual precipitation variability. This enhancement could encourage the water management sector to adopt CMIP6 future projections, which can be extremely useful when designing strategies to adapt to projected changes in precipitation.**\n * **Assessments using regional climate models (RCMs) may produce different results** (not necessarily better) due to their higher spatial resolution, better representation of local processes, and more detailed treatment of orography and land-atmosphere interactions.", "text_with_prefix": "EQC Quality Assessment: \"Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q10 | Category: Climate_Projections\nSection: Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability > Analysis and results > 3. Plot and describe results > 3.6. Results summary and discussion\n---\n* Substantial differences in spatial patterns can be observed depending on the formulation used to represent inter-annual variability during the Monsoon season (JJAS). This notebook uses two formulations: variance and the coefficient of variation (standard deviation divided by the mean).\n\n* Large differences in sign, magnitude, and spatial patterns of the biase are observed in the analyses using variance to represent inter-annual variability. Despite the differences across the subset of considered CMIP6 models, there is agreement in showing strong dependencies on orography and regions with high precipitation. These regions exhibit high biases due to elevated mean JJAS precipitation, although these biases are not necessarily large relative to the mean. Using the coefficient of variation eliminates this effect and may be more useful for assessing variability normalised to the mean amount of precipitation during the Monsoon period.\n\n* Generally, there are smaller differences depending on the model selection for the coefficient of variation, compared to the differences observed for the variance formulation. Inland north-western regions show large negative biases. These regions have, on average, near-zero or very low seasonal averages. One hypothesis to explain this may be that unusually wet seasons are underestimated in comparison to ERA5.\n\n* The time series displaying the evolution of variance and coefficient of variation (using a 30-year window) show a decline for ERA5 in the considered region. Based on the calculated trend, an approximate 50% reduction is estimated when comparing the start and end of the period for the variance formulation. This decrease is smaller when considering the coefficient of variation (11%). The ensemble median trend for the subset of considered CMIP6 models shows an increase of around 7% for the variance and 6% for the coefficient of variation over the same period, which differs from the trends observed in ERA5. It is important to note that the region studied is extensive. The time series represent spatially averaged values, and their results should be interpreted with care, as focusing on sub-regions within the domain may yield significantly different outcomes. Additionally, CMIP6 models exhibit notable inter-model spread, and a larger ensemble could also influence the results.\n\n* The time series results suggest that year-to-year variability in monsoon precipitation has become less pronounced according to ERA5. This is evident not only from the variance magnitude, which can be influenced by the evolution of the mean (shown to decrease during this period [[3]](https://doi.org/10.1007/978-981-15-4327-2_3)), but also from the coefficient of variation, which removes this dependency by normalising the standard deviation with the mean.\n\n* What do the results mean for users?\n * **Importance of the selection of the variability formulation**. The coefficient of variation has less dependency on orography because it is normalised by the mean. Moreover, the differences across the considered subset of models in the spatial pattern of the bias for the coefficient of variation appear to be smaller than in the case of variance. However, special attention needs to be given when using the coefficient of variation for the inland north-western region (where the seasonal mean precipitation is usually very low or near zero).\n * **ERA5 has biases** [[5]](https://doi.org/10.1002/qj.3616)[[6]](https://doi.org/10.3390/atmos12111462). This assessment uses it because a longer period is available and allows a more robust analysis of the temporal evolution of the variability. However the user may **consider using other reference climate products such as the [GPCP](https://cds.climate.copernicus.eu/datasets/satellite-precipitation?tab=overview)**. This could lead to different results.\n * ERA5 shows a **decrease in inter-annual variability over the historical period for the considered region**, while the CMIP6 ensemble median indicates a slight increase. It is **important to consider a larger ensemble and note that the time series in this assessment represent spatially averaged values**, which may yield different results when focusing on other sub-regions.\n * Given the challenges in replicating the spatial patterns for variance and the coefficient of variation, as well as the difficulties in representing its temporal evolution, **bias correction of CMIP6 projections may be essential for effectively using future projections of inter-annual precipitation variability. This enhancement could encourage the water management sector to adopt CMIP6 future projections, which can be extremely useful when designing strategies to adapt to projected changes in precipitation.**\n * **Assessments using regional climate models (RCMs) may produce different results** (not necessarily better) due to their higher spatial resolution, better representation of local processes, and more detailed treatment of orography and land-atmosphere interactions."} {"chunk_id": "climate_projections-cmip6_validation_q10__107767ae54e8", "report_id": "climate_projections-cmip6_validation_q10", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q10", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability > Analysis and results > 3. Plot and describe results > 3.6. Results summary and discussion", "title": "Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability", "chunk_index": 19, "token_count": 207, "text_raw": "Assessments using regional climate models (RCMs) may produce different results** (not necessarily better) due to their higher spatial resolution, better representation of local processes, and more detailed treatment of orography and land-atmosphere interactions.\n\n\n
\nRESULTS NOTE:
\nIt is important to note that the results presented are specific to the 16 models chosen, and users should aim to assess as wide a range of models as possible before making a sub-selection.", "text_with_prefix": "EQC Quality Assessment: \"Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q10 | Category: Climate_Projections\nSection: Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability > Analysis and results > 3. Plot and describe results > 3.6. Results summary and discussion\n---\nAssessments using regional climate models (RCMs) may produce different results** (not necessarily better) due to their higher spatial resolution, better representation of local processes, and more detailed treatment of orography and land-atmosphere interactions.\n\n\n
\nRESULTS NOTE:
\nIt is important to note that the results presented are specific to the 16 models chosen, and users should aim to assess as wide a range of models as possible before making a sub-selection."} {"chunk_id": "climate_projections-cmip6_validation_q10__d61f696fd576", "report_id": "climate_projections-cmip6_validation_q10", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q10", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability > ℹ️ If you want to know more > Key resources", "title": "Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability", "chunk_index": 20, "token_count": 259, "text_raw": "Some key resources and further reading were linked throughout this assessment.\n\nThe CDS catalogue entries for the data used were:\n* CMIP6 climate projections (Monthly - Precipitation): https://cds.climate.copernicus.eu/datasets/projections-cmip6?tab=overview\n* ERA5 monthly averaged data on single levels from 1940 to present (mean total precipitation rate): https://cds.climate.copernicus.eu/datasets/reanalysis-era5-single-levels-monthly-means?tab=overview\n\nCode libraries used:\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)", "text_with_prefix": "EQC Quality Assessment: \"Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q10 | Category: Climate_Projections\nSection: Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability > ℹ️ If you want to know more > Key resources\n---\nSome key resources and further reading were linked throughout this assessment.\n\nThe CDS catalogue entries for the data used were:\n* CMIP6 climate projections (Monthly - Precipitation): https://cds.climate.copernicus.eu/datasets/projections-cmip6?tab=overview\n* ERA5 monthly averaged data on single levels from 1940 to present (mean total precipitation rate): https://cds.climate.copernicus.eu/datasets/reanalysis-era5-single-levels-monthly-means?tab=overview\n\nCode libraries used:\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)"} {"chunk_id": "climate_projections-cmip6_validation_q10__10f496ff3d51", "report_id": "climate_projections-cmip6_validation_q10", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q10", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability > ℹ️ If you want to know more > References", "title": "Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability", "chunk_index": 21, "token_count": 680, "text_raw": "[[1]](https://doi.org/10.1111/gcb.12581) Thornton, P. K., Ericksen, P. J., Herrero, M., and Challinor, A. J., 2014. Climate variability and vulnerability to climate change: A review. Global Change Biol. 20, https://doi.org/10.1111/gcb.12581\n\n[[2]](https://doi.org/10.1016/j.wace.2020.100277) Masroor, M., Rehman, S., Avtar, R., Sahana, M., Ahmed, R., Sajjad, H., 2020. Exploring climate variability and its impact on drought occurrence: evidence from Godavari Middle sub-basin, India. Weather and Climate Extremes. 30, 100277. https://doi.org/10.1016/j.wace.2020.100277\n\n[[3]](https://doi.org/10.1007/978-981-15-4327-2_3) Kulkarni, A. et al., 2020. Precipitation Changes in India. In: Krishnan, R., Sanjay, J., Gnanaseelan, C., Mujumdar, M., Kulkarni, A., Chakraborty, S. (eds) Assessment of Climate Change over the Indian Region. Springer, Singapore. https://doi.org/10.1007/978-981-15-4327-2_3\n\n[[4]](https://doi.org/10.1007/s00704-022-03972-2) Longobardi, A., Boulariah, O., 2022. Long-term regional changes in inter-annual precipitation variability in the Campania Region, Southern Italy. Theor Appl Climatol 148, 869–879. https://doi.org/10.1007/s00704-022-03972-2\n\n[[5]](https://doi.org/10.1002/qj.3616) Ramon, J., Lledó L., Torralba, V., Soret, A., Doblas-Reyes, F.J., 2019. What global reanalysis best represents near-surface winds?. Q J R Meteorol Soc. 2019; 145: 3236–3251. https://doi.org/10.1002/qj.3616\n\n[[6]](https://doi.org/10.3390/atmos12111462) Hassler, B., and Lauer, A., 2021. Comparison of Reanalysis and Observational Precipitation Datasets Including ERA5 and WFDE5. Atmosphere, 12, 1462. https://doi.org/10.3390/atmos12111462", "text_with_prefix": "EQC Quality Assessment: \"Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q10 | Category: Climate_Projections\nSection: Testing the capability of CMIP6 GCMs to represent precipitation inter-annual variability > ℹ️ If you want to know more > References\n---\n[[1]](https://doi.org/10.1111/gcb.12581) Thornton, P. K., Ericksen, P. J., Herrero, M., and Challinor, A. J., 2014. Climate variability and vulnerability to climate change: A review. Global Change Biol. 20, https://doi.org/10.1111/gcb.12581\n\n[[2]](https://doi.org/10.1016/j.wace.2020.100277) Masroor, M., Rehman, S., Avtar, R., Sahana, M., Ahmed, R., Sajjad, H., 2020. Exploring climate variability and its impact on drought occurrence: evidence from Godavari Middle sub-basin, India. Weather and Climate Extremes. 30, 100277. https://doi.org/10.1016/j.wace.2020.100277\n\n[[3]](https://doi.org/10.1007/978-981-15-4327-2_3) Kulkarni, A. et al., 2020. Precipitation Changes in India. In: Krishnan, R., Sanjay, J., Gnanaseelan, C., Mujumdar, M., Kulkarni, A., Chakraborty, S. (eds) Assessment of Climate Change over the Indian Region. Springer, Singapore. https://doi.org/10.1007/978-981-15-4327-2_3\n\n[[4]](https://doi.org/10.1007/s00704-022-03972-2) Longobardi, A., Boulariah, O., 2022. Long-term regional changes in inter-annual precipitation variability in the Campania Region, Southern Italy. Theor Appl Climatol 148, 869–879. https://doi.org/10.1007/s00704-022-03972-2\n\n[[5]](https://doi.org/10.1002/qj.3616) Ramon, J., Lledó L., Torralba, V., Soret, A., Doblas-Reyes, F.J., 2019. What global reanalysis best represents near-surface winds?. Q J R Meteorol Soc. 2019; 145: 3236–3251. https://doi.org/10.1002/qj.3616\n\n[[6]](https://doi.org/10.3390/atmos12111462) Hassler, B., and Lauer, A., 2021. Comparison of Reanalysis and Observational Precipitation Datasets Including ERA5 and WFDE5. Atmosphere, 12, 1462. https://doi.org/10.3390/atmos12111462"} {"chunk_id": "climate_projections-cmip6_validation_q11__52d12e6cd142", "report_id": "climate_projections-cmip6_validation_q11", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q11", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability", "title": "Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability", "chunk_index": 0, "token_count": 104, "text_raw": "Production date: 31-03-2025\n\nProduced by: CMCC foundation - Euro-Mediterranean Center on Climate Change. Albert Martinez Boti.", "text_with_prefix": "EQC Quality Assessment: \"Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q11 | Category: Climate_Projections\nSection: Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability\n---\nProduction date: 31-03-2025\n\nProduced by: CMCC foundation - Euro-Mediterranean Center on Climate Change. Albert Martinez Boti."} {"chunk_id": "climate_projections-cmip6_validation_q11__18597c2983da", "report_id": "climate_projections-cmip6_validation_q11", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q11", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability > Quality assessment question", "title": "Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability", "chunk_index": 1, "token_count": 967, "text_raw": "* **How well do CMIP6 projections represent historical intra-seasonal variability of Monsoon precipitation across the Indian subcontinent? Are there trends affecting the intra-seasonal variability of Monsoon precipitation?​**\n\nMost research has focused on projected long-term changes in mean climate and extremes, while changes in variability have received less attention. However, understanding the evolution of precipitation variability is crucial for accurately modelling climate phenomena and quantifying the impacts of climate change, particularly on water resources and related socioeconomic factors [[1]](https://doi.org/10.1111/gcb.12581), [[2]](https://doi.org/10.1016/j.wace.2020.100277). This notebook evaluates the capability of a subset of 16 [CMIP6](https://cds.climate.copernicus.eu/datasets/projections-cmip6?tab=overview) Global Climate Models (GCMs) to represent monsoon intra-seasonal variability over the Indian subcontinent. Specifically, it evaluates the High-Frequency Intra-Seasonal Oscillations (HFISOs) with a characteristic period of 10–20 days, generally associated with the Quasi-Biweekly Oscillation (QBWO) [[3]](https://doi.org/10.1111/j.2153-3490.1980.tb01717.x), [[4]](https://doi.org/10.1175/2007MWR1991.1), [[5]](https://doi.org/10.1175/2011JCLI3916.1). The Low-Frequency Intra-Seasonal Oscillations (LFISOs) — which have a characteristic period of 30–60 days [[6]](https://doi.org/10.1175/1520-0469%281971%29028%3C0702:DOADOI%3E2.0.CO;2), [[7]](https://doi.org/10.1175/1520-0469%281972%29029<1109:DOGSCC>2.0.CO;2) and are usually linked to the well-known Madden-Julian Oscillation (MJO) mode — are also studied.\n\nSystematic errors are obtained by comparing LFISOs and HFISOs — calculated from model daily outputs — with those from the ERA5 reanalysis, which serves as the reference dataset. These oscillations are extracted by applying a pass-band filter (10–20 days for HFISOs and 30–60 days for LFISOs) to daily precipitation anomalies. To assess the intensity of these oscillations [[8]](https://doi.org/10.1038/s41612-022-00253-7) , variance and the coefficient of variation of the filtered data are calculated for the summer monsoon period (JJAS) from 1940 to 2014. This assessment presents the spatial patterns of the variance and coefficient of variation of LFISOs and HFISOs, as well as their evolution. The latter is evaluated through time series of spatially averaged variance and coefficient of variation of the filtered data using 30-year moving windows.\n\nBiases and future changes in LFISOs and HFISOs are highly relevant to users, as these modes are often associated with: (1) active-break cycles (i.e., alternating ‘active’ phases of heavy rainfall and quiescent phases during the peak monsoon season in July and August [[9]](https://doi.org/10.1002/jgrd.50422), [[10]](https://doi.org/10.1175/JCLI-D-18-0319.1)), (2) the activity of synoptic systems such as monsoon lows (e.g., [[11]](https://doi.org/10.1175/JCLI-D-16-0360.1)) and (3) late monsoon rainfall [[12]](https://doi.org/10.1007/s00382-019-04946-3). This evaluation complements another [assessment](https://ecmwf-projects.github.io/c3s2-eqc-quality-assessment/Climate_Projections/CMIP6/climate_projections-cmip6_validation_q10.html#) that examined the representation of historic inter-annual precipitation variability across the Indian subcontinent.", "text_with_prefix": "EQC Quality Assessment: \"Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q11 | Category: Climate_Projections\nSection: Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability > Quality assessment question\n---\n* **How well do CMIP6 projections represent historical intra-seasonal variability of Monsoon precipitation across the Indian subcontinent? Are there trends affecting the intra-seasonal variability of Monsoon precipitation?​**\n\nMost research has focused on projected long-term changes in mean climate and extremes, while changes in variability have received less attention. However, understanding the evolution of precipitation variability is crucial for accurately modelling climate phenomena and quantifying the impacts of climate change, particularly on water resources and related socioeconomic factors [[1]](https://doi.org/10.1111/gcb.12581), [[2]](https://doi.org/10.1016/j.wace.2020.100277). This notebook evaluates the capability of a subset of 16 [CMIP6](https://cds.climate.copernicus.eu/datasets/projections-cmip6?tab=overview) Global Climate Models (GCMs) to represent monsoon intra-seasonal variability over the Indian subcontinent. Specifically, it evaluates the High-Frequency Intra-Seasonal Oscillations (HFISOs) with a characteristic period of 10–20 days, generally associated with the Quasi-Biweekly Oscillation (QBWO) [[3]](https://doi.org/10.1111/j.2153-3490.1980.tb01717.x), [[4]](https://doi.org/10.1175/2007MWR1991.1), [[5]](https://doi.org/10.1175/2011JCLI3916.1). The Low-Frequency Intra-Seasonal Oscillations (LFISOs) — which have a characteristic period of 30–60 days [[6]](https://doi.org/10.1175/1520-0469%281971%29028%3C0702:DOADOI%3E2.0.CO;2), [[7]](https://doi.org/10.1175/1520-0469%281972%29029<1109:DOGSCC>2.0.CO;2) and are usually linked to the well-known Madden-Julian Oscillation (MJO) mode — are also studied.\n\nSystematic errors are obtained by comparing LFISOs and HFISOs — calculated from model daily outputs — with those from the ERA5 reanalysis, which serves as the reference dataset. These oscillations are extracted by applying a pass-band filter (10–20 days for HFISOs and 30–60 days for LFISOs) to daily precipitation anomalies. To assess the intensity of these oscillations [[8]](https://doi.org/10.1038/s41612-022-00253-7) , variance and the coefficient of variation of the filtered data are calculated for the summer monsoon period (JJAS) from 1940 to 2014. This assessment presents the spatial patterns of the variance and coefficient of variation of LFISOs and HFISOs, as well as their evolution. The latter is evaluated through time series of spatially averaged variance and coefficient of variation of the filtered data using 30-year moving windows.\n\nBiases and future changes in LFISOs and HFISOs are highly relevant to users, as these modes are often associated with: (1) active-break cycles (i.e., alternating ‘active’ phases of heavy rainfall and quiescent phases during the peak monsoon season in July and August [[9]](https://doi.org/10.1002/jgrd.50422), [[10]](https://doi.org/10.1175/JCLI-D-18-0319.1)), (2) the activity of synoptic systems such as monsoon lows (e.g., [[11]](https://doi.org/10.1175/JCLI-D-16-0360.1)) and (3) late monsoon rainfall [[12]](https://doi.org/10.1007/s00382-019-04946-3). This evaluation complements another [assessment](https://ecmwf-projects.github.io/c3s2-eqc-quality-assessment/Climate_Projections/CMIP6/climate_projections-cmip6_validation_q10.html#) that examined the representation of historic inter-annual precipitation variability across the Indian subcontinent."} {"chunk_id": "climate_projections-cmip6_validation_q11__5df7385b4ea3", "report_id": "climate_projections-cmip6_validation_q11", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q11", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability > Quality assessment statement", "title": "Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability", "chunk_index": 2, "token_count": 582, "text_raw": "These are the key outcomes of this assessment\n\n* This assessment complements a previous analysis of [historical inter-annual precipitation variability across the Indian subcontinent](https://ecmwf-projects.github.io/c3s2-eqc-quality-assessment/Climate_Projections/CMIP6/climate_projections-cmip6_validation_q10.html#).\n\n* Intra-seasonal and inter-annual variability assessments show similar spatial patterns, suggesting a connection between the two [[10]](https://doi.org/10.1175/JCLI-D-18-0319.1)\n\n* Variability formulation matters: the two metrics used highlight different spatial features of intra-seasonal variability.\n\n* Variance shows model-dependent biases linked to orography and wet regions. The Coefficient of Variation (CV) normalises for mean precipitation but can lead to artificially high values in extremely dry areas.\n\n* CV differences between models are smaller than those for variance. Most models underestimate CV in dry inland regions.\n\n* HFISOs and LFISOs variance and CV exhibit similar spatial patterns.\n\n* CMIP6 models poorly capture observed trends in HFISO and LFISO variability. ERA5 shows decreasing variance and increasing CV, which are not reproduced by the ensemble median.\n\n* Bias correction may be necessary before using CMIP6 precipitation projections in water management applications. Similarities in inter-annual and intra-seasonal variability biases suggest corrections may be required at the daily level.\n\n```\n\nattachment:8aac0bed-aa00-4ba0-9343-cbc722fc0ca8.png\n---\nwidth: 900px\nalt: Mean Bias variance \n---\nFig A. i. Variance of HFISOs in daily precipitation anomalies over the historical period (1940–2014). The layout includes data corresponding to: (a) ERA5, (b) the ensemble median (defined as the median of the Variance of HFISOs values for the selected subset of models, calculated for each grid cell), (c) the bias of the ensemble median, and (d) the ensemble spread (calculated as the standard deviation of the Variance of HFISOs values for the selected subset of models). Before calculating the variance, a Lanczos band-pass filter has been applied to the daily precipitation anomalies to isolate the 10–20-day high-frequency intraseasonal oscillations (HFISOs). Fig A.ii. Same as Fig. A.i. but for the coefficient of variation of HFISOs.\n```", "text_with_prefix": "EQC Quality Assessment: \"Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q11 | Category: Climate_Projections\nSection: Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability > Quality assessment statement\n---\nThese are the key outcomes of this assessment\n\n* This assessment complements a previous analysis of [historical inter-annual precipitation variability across the Indian subcontinent](https://ecmwf-projects.github.io/c3s2-eqc-quality-assessment/Climate_Projections/CMIP6/climate_projections-cmip6_validation_q10.html#).\n\n* Intra-seasonal and inter-annual variability assessments show similar spatial patterns, suggesting a connection between the two [[10]](https://doi.org/10.1175/JCLI-D-18-0319.1)\n\n* Variability formulation matters: the two metrics used highlight different spatial features of intra-seasonal variability.\n\n* Variance shows model-dependent biases linked to orography and wet regions. The Coefficient of Variation (CV) normalises for mean precipitation but can lead to artificially high values in extremely dry areas.\n\n* CV differences between models are smaller than those for variance. Most models underestimate CV in dry inland regions.\n\n* HFISOs and LFISOs variance and CV exhibit similar spatial patterns.\n\n* CMIP6 models poorly capture observed trends in HFISO and LFISO variability. ERA5 shows decreasing variance and increasing CV, which are not reproduced by the ensemble median.\n\n* Bias correction may be necessary before using CMIP6 precipitation projections in water management applications. Similarities in inter-annual and intra-seasonal variability biases suggest corrections may be required at the daily level.\n\n```\n\nattachment:8aac0bed-aa00-4ba0-9343-cbc722fc0ca8.png\n---\nwidth: 900px\nalt: Mean Bias variance \n---\nFig A. i. Variance of HFISOs in daily precipitation anomalies over the historical period (1940–2014). The layout includes data corresponding to: (a) ERA5, (b) the ensemble median (defined as the median of the Variance of HFISOs values for the selected subset of models, calculated for each grid cell), (c) the bias of the ensemble median, and (d) the ensemble spread (calculated as the standard deviation of the Variance of HFISOs values for the selected subset of models). Before calculating the variance, a Lanczos band-pass filter has been applied to the daily precipitation anomalies to isolate the 10–20-day high-frequency intraseasonal oscillations (HFISOs). Fig A.ii. Same as Fig. A.i. but for the coefficient of variation of HFISOs.\n```"} {"chunk_id": "climate_projections-cmip6_validation_q11__8d2b7a867fc4", "report_id": "climate_projections-cmip6_validation_q11", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q11", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability > Methodology", "title": "Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability", "chunk_index": 3, "token_count": 945, "text_raw": "A subset of 16 models from the [CMIP6](https://cds.climate.copernicus.eu/datasets/projections-cmip6?tab=overview) project is used to evaluate the representation of intra-seasonal precipitation variability over the Indian subcontinent. The analysis focuses on two intra-seasonal timescales: High-Frequency Intra-Seasonal Oscillations (HFISOs) with periods of 10–20 days and Low-Frequency Intra-Seasonal Oscillations (LFISOs) with periods of 30–60 days. LFISOs and HFISOs are identified using a pass-band filter (10–20 days for HFISOs and 30–60 days for LFISOs) applied to daily precipitation anomalies, computed relative to the climatological precipitation for each day of the year. The intensity of these intra-seasonal oscillations is assessed by calculating the variance and Coefficient of Variation (CV) of the filtered data, following the approach in [[8]](https://doi.org/10.1038/s41612-022-00253-7), for the summer monsoon period (JJAS) from 1940 to 2014.\n\nThe analyses performed within this notebook examine the spatial patterns of variance and the Coefficient of Variation for LFISOs and HFISOs, along with the associated biases for each model and the ensemble median (per grid cell). [ERA5](https://cds.climate.copernicus.eu/datasets/reanalysis-era5-complete?tab=doc) serves as the reference dataset. Note that although HFISOs are generally associated with the Quasi-Biweekly Oscillation (QBWO) [[3]](https://doi.org/10.1111/j.2153-3490.1980.tb01717.x), [[4]](https://doi.org/10.1175/2007MWR1991.1), [[5]](https://doi.org/10.1175/2011JCLI3916.1), and LFISOs are typically linked to the Madden–Julian Oscillation (MJO) [[6]](https://doi.org/10.1175/1520-0469%281971%29028%3C0702:DOADOI%3E2.0.CO;2), [[7]](https://doi.org/10.1175/1520-0469%281972%29029<1109:DOGSCC>2.0.CO;2), the filtering method is subject to limitations, including signal contamination and overlap. Therefore, the variability captured in each band is not exclusively attributable to the QBWO or MJO.\n\nFinally, the temporal evolution of spatially averaged intra-seasonal variability is analysed using time series of variance and the coefficient of variation, computed with a 30-year moving window. This approach allows for assessing long-term changes in the intensity of HFISOs and LFISOs over the historical period from 1940 to 2014.\n\nThe analysis focuses on the JJAS period, traditionally regarded as the Indian Summer Monsoon season, but the methodology can be adapted for other seasonal or annual periods.\n\nThe analysis and results follow the next outline:\n\n**[](climate_projections-cmip6_validation_q11:section-1)**\n * [](climate_projections-cmip6_validation_q11:section-1.1)\n * [](climate_projections-cmip6_validation_q11:section-1.2)\n * [](climate_projections-cmip6_validation_q11:section-1.3)\n * [](climate_projections-cmip6_validation_q11:section-1.4)\n * [](climate_projections-cmip6_validation_q11:section-1.5)\n * [](climate_projections-cmip6_validation_q11:section-1.6)\n\n**[](climate_projections-cmip6_validation_q11:section-2)**\n * [](climate_projections-cmip6_validation_q11:section-2.1)\n * [](climate_projections-cmip6_validation_q11:section-2.2)\n * [](climate_projections-cmip6_validation_q11:section-2.3)", "text_with_prefix": "EQC Quality Assessment: \"Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q11 | Category: Climate_Projections\nSection: Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability > Methodology\n---\nA subset of 16 models from the [CMIP6](https://cds.climate.copernicus.eu/datasets/projections-cmip6?tab=overview) project is used to evaluate the representation of intra-seasonal precipitation variability over the Indian subcontinent. The analysis focuses on two intra-seasonal timescales: High-Frequency Intra-Seasonal Oscillations (HFISOs) with periods of 10–20 days and Low-Frequency Intra-Seasonal Oscillations (LFISOs) with periods of 30–60 days. LFISOs and HFISOs are identified using a pass-band filter (10–20 days for HFISOs and 30–60 days for LFISOs) applied to daily precipitation anomalies, computed relative to the climatological precipitation for each day of the year. The intensity of these intra-seasonal oscillations is assessed by calculating the variance and Coefficient of Variation (CV) of the filtered data, following the approach in [[8]](https://doi.org/10.1038/s41612-022-00253-7), for the summer monsoon period (JJAS) from 1940 to 2014.\n\nThe analyses performed within this notebook examine the spatial patterns of variance and the Coefficient of Variation for LFISOs and HFISOs, along with the associated biases for each model and the ensemble median (per grid cell). [ERA5](https://cds.climate.copernicus.eu/datasets/reanalysis-era5-complete?tab=doc) serves as the reference dataset. Note that although HFISOs are generally associated with the Quasi-Biweekly Oscillation (QBWO) [[3]](https://doi.org/10.1111/j.2153-3490.1980.tb01717.x), [[4]](https://doi.org/10.1175/2007MWR1991.1), [[5]](https://doi.org/10.1175/2011JCLI3916.1), and LFISOs are typically linked to the Madden–Julian Oscillation (MJO) [[6]](https://doi.org/10.1175/1520-0469%281971%29028%3C0702:DOADOI%3E2.0.CO;2), [[7]](https://doi.org/10.1175/1520-0469%281972%29029<1109:DOGSCC>2.0.CO;2), the filtering method is subject to limitations, including signal contamination and overlap. Therefore, the variability captured in each band is not exclusively attributable to the QBWO or MJO.\n\nFinally, the temporal evolution of spatially averaged intra-seasonal variability is analysed using time series of variance and the coefficient of variation, computed with a 30-year moving window. This approach allows for assessing long-term changes in the intensity of HFISOs and LFISOs over the historical period from 1940 to 2014.\n\nThe analysis focuses on the JJAS period, traditionally regarded as the Indian Summer Monsoon season, but the methodology can be adapted for other seasonal or annual periods.\n\nThe analysis and results follow the next outline:\n\n**[](climate_projections-cmip6_validation_q11:section-1)**\n * [](climate_projections-cmip6_validation_q11:section-1.1)\n * [](climate_projections-cmip6_validation_q11:section-1.2)\n * [](climate_projections-cmip6_validation_q11:section-1.3)\n * [](climate_projections-cmip6_validation_q11:section-1.4)\n * [](climate_projections-cmip6_validation_q11:section-1.5)\n * [](climate_projections-cmip6_validation_q11:section-1.6)\n\n**[](climate_projections-cmip6_validation_q11:section-2)**\n * [](climate_projections-cmip6_validation_q11:section-2.1)\n * [](climate_projections-cmip6_validation_q11:section-2.2)\n * [](climate_projections-cmip6_validation_q11:section-2.3)"} {"chunk_id": "climate_projections-cmip6_validation_q11__7693a9394599", "report_id": "climate_projections-cmip6_validation_q11", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q11", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability > Methodology", "title": "Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability", "chunk_index": 4, "token_count": 267, "text_raw": "-2.1)\n * [](climate_projections-cmip6_validation_q11:section-2.2)\n * [](climate_projections-cmip6_validation_q11:section-2.3)\n\n**[](climate_projections-cmip6_validation_q11:section-3)**\n * [](climate_projections-cmip6_validation_q11:section-3.1)\n * [](climate_projections-cmip6_validation_q11:section-3.2)\n * [](climate_projections-cmip6_validation_q11:section-3.3)\n * [](climate_projections-cmip6_validation_q11:section-3.4)\n * [](climate_projections-cmip6_validation_q11:section-3.5)\n * [](climate_projections-cmip6_validation_q11:section-3.6)\n * [](climate_projections-cmip6_validation_q11:section-3.7)", "text_with_prefix": "EQC Quality Assessment: \"Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q11 | Category: Climate_Projections\nSection: Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability > Methodology\n---\n-2.1)\n * [](climate_projections-cmip6_validation_q11:section-2.2)\n * [](climate_projections-cmip6_validation_q11:section-2.3)\n\n**[](climate_projections-cmip6_validation_q11:section-3)**\n * [](climate_projections-cmip6_validation_q11:section-3.1)\n * [](climate_projections-cmip6_validation_q11:section-3.2)\n * [](climate_projections-cmip6_validation_q11:section-3.3)\n * [](climate_projections-cmip6_validation_q11:section-3.4)\n * [](climate_projections-cmip6_validation_q11:section-3.5)\n * [](climate_projections-cmip6_validation_q11:section-3.6)\n * [](climate_projections-cmip6_validation_q11:section-3.7)"} {"chunk_id": "climate_projections-cmip6_validation_q11__58b83098d426", "report_id": "climate_projections-cmip6_validation_q11", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q11", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability > Analysis and results > 1. Parameters, requests and functions definition > 1.2. Define Parameters", "title": "Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability", "chunk_index": 5, "token_count": 421, "text_raw": "In the \"Define Parameters\" section, various customisable options for the notebook are specified:\n\n- The initial and ending year used for the historical period can be specified by changing the parameters `year_start` and `year_stop` (1940-2014 is chosen).\n- `timeseries` sets the temporal period. For instance, selecting \"JJAS\" implies considering only the June, July, August, September season.\n- `area` allows specifying the geographical domain of interest (NWSE). A domain including the Indian sub-continent has been selected.\n- `area_timeseries` specifies the region of interest when displaying the timeseries. We have chosen the Indian Peninsula for this assessment. This domain should be equal or smaller than `area`. \n- The `interpolation_method` parameter allows selecting the interpolation method when regridding is performed. \n- The `chunk` selection allows the user to define if dividing into chunks when downloading the data on their local machine. Although it does not significantly affect the analysis, it is recommended to keep the default value for optimal performance.\n- `variable` and `collection_id` are not customisable for this assessment and are set to 'precipitation' and 'CMIP6'. Expert users can use this notebook as a guide to create similar analyses for other variables or model sets (such as CORDEX).\n\nChoose annual or seasonal timeseries\nInterpolation method\nArea to show (NWSE)\nArea for the timeseries analysis (equal or smaller than area)\nChunks for download\nCollection id\n\n(climate_projections-cmip6_validation_q11:section-1.3)=", "text_with_prefix": "EQC Quality Assessment: \"Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q11 | Category: Climate_Projections\nSection: Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability > Analysis and results > 1. Parameters, requests and functions definition > 1.2. Define Parameters\n---\nIn the \"Define Parameters\" section, various customisable options for the notebook are specified:\n\n- The initial and ending year used for the historical period can be specified by changing the parameters `year_start` and `year_stop` (1940-2014 is chosen).\n- `timeseries` sets the temporal period. For instance, selecting \"JJAS\" implies considering only the June, July, August, September season.\n- `area` allows specifying the geographical domain of interest (NWSE). A domain including the Indian sub-continent has been selected.\n- `area_timeseries` specifies the region of interest when displaying the timeseries. We have chosen the Indian Peninsula for this assessment. This domain should be equal or smaller than `area`. \n- The `interpolation_method` parameter allows selecting the interpolation method when regridding is performed. \n- The `chunk` selection allows the user to define if dividing into chunks when downloading the data on their local machine. Although it does not significantly affect the analysis, it is recommended to keep the default value for optimal performance.\n- `variable` and `collection_id` are not customisable for this assessment and are set to 'precipitation' and 'CMIP6'. Expert users can use this notebook as a guide to create similar analyses for other variables or model sets (such as CORDEX).\n\nChoose annual or seasonal timeseries\nInterpolation method\nArea to show (NWSE)\nArea for the timeseries analysis (equal or smaller than area)\nChunks for download\nCollection id\n\n(climate_projections-cmip6_validation_q11:section-1.3)="} {"chunk_id": "climate_projections-cmip6_validation_q11__28282f4b8ef5", "report_id": "climate_projections-cmip6_validation_q11", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q11", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability > Analysis and results > 1. Parameters, requests and functions definition > 1.3. Define input models and output parameters", "title": "Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability", "chunk_index": 6, "token_count": 174, "text_raw": "The following climate analyses are performed considering a subset of GCMs from CMIP6.\n\nThe selected CMIP6 models have available both the historical and SSP5-8.5 experiments.\n\nRanges of the bands used to filter and get the ISOs (Intra-Seasonal Oscillations)\n\n(climate_projections-cmip6_validation_q11:section-1.4)=", "text_with_prefix": "EQC Quality Assessment: \"Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q11 | Category: Climate_Projections\nSection: Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability > Analysis and results > 1. Parameters, requests and functions definition > 1.3. Define input models and output parameters\n---\nThe following climate analyses are performed considering a subset of GCMs from CMIP6.\n\nThe selected CMIP6 models have available both the historical and SSP5-8.5 experiments.\n\nRanges of the bands used to filter and get the ISOs (Intra-Seasonal Oscillations)\n\n(climate_projections-cmip6_validation_q11:section-1.4)="} {"chunk_id": "climate_projections-cmip6_validation_q11__d850d7afa3ac", "report_id": "climate_projections-cmip6_validation_q11", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q11", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability > Analysis and results > 1. Parameters, requests and functions definition > 1.4. Define ERA5 request", "title": "Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability", "chunk_index": 7, "token_count": 143, "text_raw": "Within this notebook, ERA5 serves as the reference product. In this section, we set the required parameters for the cds-api data-request of ERA5.\n\n(climate_projections-cmip6_validation_q11:section-1.5)=", "text_with_prefix": "EQC Quality Assessment: \"Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q11 | Category: Climate_Projections\nSection: Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability > Analysis and results > 1. Parameters, requests and functions definition > 1.4. Define ERA5 request\n---\nWithin this notebook, ERA5 serves as the reference product. In this section, we set the required parameters for the cds-api data-request of ERA5.\n\n(climate_projections-cmip6_validation_q11:section-1.5)="} {"chunk_id": "climate_projections-cmip6_validation_q11__bf39047dc736", "report_id": "climate_projections-cmip6_validation_q11", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q11", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability > Analysis and results > 1. Parameters, requests and functions definition > 1.5. Define model requests", "title": "Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability", "chunk_index": 8, "token_count": 126, "text_raw": "In this section we set the required parameters for the cds-api data-request.\n\n(climate_projections-cmip6_validation_q11:section-1.6)=", "text_with_prefix": "EQC Quality Assessment: \"Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q11 | Category: Climate_Projections\nSection: Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability > Analysis and results > 1. Parameters, requests and functions definition > 1.5. Define model requests\n---\nIn this section we set the required parameters for the cds-api data-request.\n\n(climate_projections-cmip6_validation_q11:section-1.6)="} {"chunk_id": "climate_projections-cmip6_validation_q11__301d28357922", "report_id": "climate_projections-cmip6_validation_q11", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q11", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability > Analysis and results > 1. Parameters, requests and functions definition > 1.6. Functions to cache", "title": "Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability", "chunk_index": 9, "token_count": 512, "text_raw": "In this section, functions that will be executed in the caching phase are defined. Caching is the process of storing copies of files in a temporary storage location, so that they can be accessed more quickly. This process also checks if the user has already downloaded a file, avoiding redundant downloads.\n\nFunctions description:\n\n- `get_grid_out` and `add_bounds` ensure the regrid is performed properly.\n\n- The `select_timeseries` function subsets the dataset based on the chosen `timeseries` parameter.\n\n- The `lanczos_bandpass_filter` uses the band-pass filter Lanczos to isolate the 10–20-day (30-60 day for the LFISOs) component of the precipitation.\n\n- The `compute_rolling_variance` function calculates the rolling variance and coefficient of variation of the filtered data using a 30-year window (which are needed when displaying the time series afterwards). It also computes the variance and coefficient of variation for the entire period.\n\n- The `compute_intraseasonal_variance` function calculates daily anomalies, filters the low and high frequency intra seasonal oscillations (by calling `lanczos_bandpass_filter`) and selects the season using the `select_timeseries` function. It then computes the rolling variance (and coefficient of variation) of the filtered data over the historical period (1940–2014) by calling `compute_rolling_variance`. The window width is 30 years, with a step size of one year. The variance and coefficient of variation is also calculated for the whole historical period.\n\nDesign FIR filter using `firwin`\nApply the filter using filtfilt for zero-phase distortion\nStore results with unique names\nMerge into a Dataset\nApply resample reduction if specified\nChange units\nget the anomalies\nget the filtered data\nSelect the time series for both ds and ds_filtered\nCompute the variance and coefficient of variation\nCreate the new dataset with NaN values for the calculating the variance and CV for the whole period\nHandle grid output if requested\n\n(climate_projections-cmip6_validation_q11:section-2)=", "text_with_prefix": "EQC Quality Assessment: \"Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q11 | Category: Climate_Projections\nSection: Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability > Analysis and results > 1. Parameters, requests and functions definition > 1.6. Functions to cache\n---\nIn this section, functions that will be executed in the caching phase are defined. Caching is the process of storing copies of files in a temporary storage location, so that they can be accessed more quickly. This process also checks if the user has already downloaded a file, avoiding redundant downloads.\n\nFunctions description:\n\n- `get_grid_out` and `add_bounds` ensure the regrid is performed properly.\n\n- The `select_timeseries` function subsets the dataset based on the chosen `timeseries` parameter.\n\n- The `lanczos_bandpass_filter` uses the band-pass filter Lanczos to isolate the 10–20-day (30-60 day for the LFISOs) component of the precipitation.\n\n- The `compute_rolling_variance` function calculates the rolling variance and coefficient of variation of the filtered data using a 30-year window (which are needed when displaying the time series afterwards). It also computes the variance and coefficient of variation for the entire period.\n\n- The `compute_intraseasonal_variance` function calculates daily anomalies, filters the low and high frequency intra seasonal oscillations (by calling `lanczos_bandpass_filter`) and selects the season using the `select_timeseries` function. It then computes the rolling variance (and coefficient of variation) of the filtered data over the historical period (1940–2014) by calling `compute_rolling_variance`. The window width is 30 years, with a step size of one year. The variance and coefficient of variation is also calculated for the whole historical period.\n\nDesign FIR filter using `firwin`\nApply the filter using filtfilt for zero-phase distortion\nStore results with unique names\nMerge into a Dataset\nApply resample reduction if specified\nChange units\nget the anomalies\nget the filtered data\nSelect the time series for both ds and ds_filtered\nCompute the variance and coefficient of variation\nCreate the new dataset with NaN values for the calculating the variance and CV for the whole period\nHandle grid output if requested\n\n(climate_projections-cmip6_validation_q11:section-2)="} {"chunk_id": "climate_projections-cmip6_validation_q11__92981ca41a5b", "report_id": "climate_projections-cmip6_validation_q11", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q11", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability > Analysis and results > 2. Downloading and processing > 2.1. Download and transform ERA5", "title": "Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability", "chunk_index": 10, "token_count": 222, "text_raw": "In this section, the `download.download_and_transform` function from the 'c3s_eqc_automatic_quality_control' package is used to download ERA5 reference dailly data, calculate daily anomalies, filter the low and high fequency intra seasonal oscillations, select the season (e.g., \"JJAS\" in this case), and compute the variance and coefficient of variation of the filtered data for the entire historical period, as well as for a 30-year rolling window. The results are then cached to prevent redundant downloads and processing.\n\n(climate_projections-cmip6_validation_q11:section-2.2)=", "text_with_prefix": "EQC Quality Assessment: \"Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q11 | Category: Climate_Projections\nSection: Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability > Analysis and results > 2. Downloading and processing > 2.1. Download and transform ERA5\n---\nIn this section, the `download.download_and_transform` function from the 'c3s_eqc_automatic_quality_control' package is used to download ERA5 reference dailly data, calculate daily anomalies, filter the low and high fequency intra seasonal oscillations, select the season (e.g., \"JJAS\" in this case), and compute the variance and coefficient of variation of the filtered data for the entire historical period, as well as for a 30-year rolling window. The results are then cached to prevent redundant downloads and processing.\n\n(climate_projections-cmip6_validation_q11:section-2.2)="} {"chunk_id": "climate_projections-cmip6_validation_q11__c037a4eea8ab", "report_id": "climate_projections-cmip6_validation_q11", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q11", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability > Analysis and results > 2. Downloading and processing > 2.2. Download and transform models", "title": "Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability", "chunk_index": 11, "token_count": 381, "text_raw": "In this section, the `download.download_and_transform` function from the 'c3s_eqc_automatic_quality_control' package is employed to download daily data from the CMIP6 models, calculate daily anomalies, filter the low and high fequency intra seasonal oscillations, select the season (\"JJAS\" in this example), compute the variance and coefficient of variation of the filtered data for the entire historical period (as well as for a 30-year rolling windows), interpolate to the ERA5 grid (only when specified; otherwise, the original model grid is maintained), and cache the results to avoid redundant downloads and processing.\n\nOriginal model\nInterpolated model\n\n```text\nmodel='access_cm2'\nmodel='bcc_csm2_mr'\nmodel='canesm5'\nmodel='cmcc_esm2'\nmodel='cnrm_cm6_1_hr'\nmodel='cnrm_esm2_1'\nmodel='ec_earth3_cc'\nmodel='gfdl_esm4'\nmodel='inm_cm4_8'\nmodel='inm_cm5_0'\nmodel='mpi_esm1_2_lr'\nmodel='miroc6'\nmodel='miroc_es2l'\nmodel='mri_esm2_0'\nmodel='noresm2_mm'\nmodel='nesm3'\n```\n\n(climate_projections-cmip6_validation_q11:section-2.3)=", "text_with_prefix": "EQC Quality Assessment: \"Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q11 | Category: Climate_Projections\nSection: Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability > Analysis and results > 2. Downloading and processing > 2.2. Download and transform models\n---\nIn this section, the `download.download_and_transform` function from the 'c3s_eqc_automatic_quality_control' package is employed to download daily data from the CMIP6 models, calculate daily anomalies, filter the low and high fequency intra seasonal oscillations, select the season (\"JJAS\" in this example), compute the variance and coefficient of variation of the filtered data for the entire historical period (as well as for a 30-year rolling windows), interpolate to the ERA5 grid (only when specified; otherwise, the original model grid is maintained), and cache the results to avoid redundant downloads and processing.\n\nOriginal model\nInterpolated model\n\n```text\nmodel='access_cm2'\nmodel='bcc_csm2_mr'\nmodel='canesm5'\nmodel='cmcc_esm2'\nmodel='cnrm_cm6_1_hr'\nmodel='cnrm_esm2_1'\nmodel='ec_earth3_cc'\nmodel='gfdl_esm4'\nmodel='inm_cm4_8'\nmodel='inm_cm5_0'\nmodel='mpi_esm1_2_lr'\nmodel='miroc6'\nmodel='miroc_es2l'\nmodel='mri_esm2_0'\nmodel='noresm2_mm'\nmodel='nesm3'\n```\n\n(climate_projections-cmip6_validation_q11:section-2.3)="} {"chunk_id": "climate_projections-cmip6_validation_q11__bfb4893e4b17", "report_id": "climate_projections-cmip6_validation_q11", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q11", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability > Analysis and results > 2. Downloading and processing > 2.3. Change some attributes", "title": "Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability", "chunk_index": 12, "token_count": 125, "text_raw": "Drop 'bnds' dimension if it exists for each dataset and for the interpolated dataset\n\n(climate_projections-cmip6_validation_q11:section-3)=", "text_with_prefix": "EQC Quality Assessment: \"Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q11 | Category: Climate_Projections\nSection: Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability > Analysis and results > 2. Downloading and processing > 2.3. Change some attributes\n---\nDrop 'bnds' dimension if it exists for each dataset and for the interpolated dataset\n\n(climate_projections-cmip6_validation_q11:section-3)="} {"chunk_id": "climate_projections-cmip6_validation_q11__848e215b7911", "report_id": "climate_projections-cmip6_validation_q11", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q11", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability > Analysis and results > 3. Plot and describe results", "title": "Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability", "chunk_index": 13, "token_count": 387, "text_raw": "This section will display the following results:\n\n- Maps showing the spatial distribution of the **variance of filtered daily precipitation anomalies**, calculated for the historical period 1940–2014. The results compare ERA5 and the ensemble median, defined as the median of the variance values for the selected subset of models, calculated for each grid cell. The layout includes ERA5, the ensemble median, the ensemble median bias, and the ensemble spread (derived as the standard deviation of the variance for the selected subset of models). The same layout is displayed for the coefficient of variation.\n\n- Maps showing the spatial distribution of the **variance of filtered daily precipitation anomalies**, calculated for the historical period 1940–2014 for each model. The same layout is displayed for the coefficient of variation.\n\n- **Bias maps of the variance of filtered daily precipitation anomalies**, calculated for the historical period 1940–2014 for each model. The same layout is displayed for the coefficient of variation.\n\n- **Time series** showing the evolution of the 30-year variance and coefficient of variation (standard deviation divided by the mean) of filtered daily precipitation anomalies over the historical period, including trends and inter-model spread.\n\n**Note**: A Lanczos band-pass filter is applied to isolate the 10–20-day (HFISOs) and 30–60-day (LFISOs) components of daily precipitation anomalies. The seasonal period is JJAS.\n\n(climate_projections-cmip6_validation_q11:section-3.1)=", "text_with_prefix": "EQC Quality Assessment: \"Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q11 | Category: Climate_Projections\nSection: Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability > Analysis and results > 3. Plot and describe results\n---\nThis section will display the following results:\n\n- Maps showing the spatial distribution of the **variance of filtered daily precipitation anomalies**, calculated for the historical period 1940–2014. The results compare ERA5 and the ensemble median, defined as the median of the variance values for the selected subset of models, calculated for each grid cell. The layout includes ERA5, the ensemble median, the ensemble median bias, and the ensemble spread (derived as the standard deviation of the variance for the selected subset of models). The same layout is displayed for the coefficient of variation.\n\n- Maps showing the spatial distribution of the **variance of filtered daily precipitation anomalies**, calculated for the historical period 1940–2014 for each model. The same layout is displayed for the coefficient of variation.\n\n- **Bias maps of the variance of filtered daily precipitation anomalies**, calculated for the historical period 1940–2014 for each model. The same layout is displayed for the coefficient of variation.\n\n- **Time series** showing the evolution of the 30-year variance and coefficient of variation (standard deviation divided by the mean) of filtered daily precipitation anomalies over the historical period, including trends and inter-model spread.\n\n**Note**: A Lanczos band-pass filter is applied to isolate the 10–20-day (HFISOs) and 30–60-day (LFISOs) components of daily precipitation anomalies. The seasonal period is JJAS.\n\n(climate_projections-cmip6_validation_q11:section-3.1)="} {"chunk_id": "climate_projections-cmip6_validation_q11__d041c7efc937", "report_id": "climate_projections-cmip6_validation_q11", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q11", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability > Analysis and results > 3. Plot and describe results > 3.1. Define plotting functions", "title": "Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability", "chunk_index": 14, "token_count": 304, "text_raw": "The functions presented here are used to plot layouts of the variance and coefficient of variation of filtered daily precipitation anomalies over the whole historical period.\n\nThree layout types can be displayed, depending on the chosen function:\n\n1. Layout including the reference ERA5 product, the ensemble median, the bias of the ensemble median, and the ensemble spread: `plot_ensemble()` is used.\n2. Layout including every model: `plot_models()` is employed.\n3. Layout including the bias of every model: `plot_models()` is used again.\n\nDefine function to plot the caption of the figures (for the ensemble case)\nAdd each line to the figure\nend captioning\nDefine function to plot the caption of the figures (for the individual models case)\nAdd caption to the figure\nAdd each line to the figure\nDefault behaviour\nCreate figure and axes\n--- ERA5 plot ---\n--- Ensemble Median ---\n--- Ensemble Bias (Median - ERA5) ---\n--- Ensemble Standard Deviation ---\n\n(climate_projections-cmip6_validation_q11:section-3.2)=", "text_with_prefix": "EQC Quality Assessment: \"Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q11 | Category: Climate_Projections\nSection: Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability > Analysis and results > 3. Plot and describe results > 3.1. Define plotting functions\n---\nThe functions presented here are used to plot layouts of the variance and coefficient of variation of filtered daily precipitation anomalies over the whole historical period.\n\nThree layout types can be displayed, depending on the chosen function:\n\n1. Layout including the reference ERA5 product, the ensemble median, the bias of the ensemble median, and the ensemble spread: `plot_ensemble()` is used.\n2. Layout including every model: `plot_models()` is employed.\n3. Layout including the bias of every model: `plot_models()` is used again.\n\nDefine function to plot the caption of the figures (for the ensemble case)\nAdd each line to the figure\nend captioning\nDefine function to plot the caption of the figures (for the individual models case)\nAdd caption to the figure\nAdd each line to the figure\nDefault behaviour\nCreate figure and axes\n--- ERA5 plot ---\n--- Ensemble Median ---\n--- Ensemble Bias (Median - ERA5) ---\n--- Ensemble Standard Deviation ---\n\n(climate_projections-cmip6_validation_q11:section-3.2)="} {"chunk_id": "climate_projections-cmip6_validation_q11__84233d7b558a", "report_id": "climate_projections-cmip6_validation_q11", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q11", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability > Analysis and results > 3. Plot and describe results > 3.2. Plot ensemble maps", "title": "Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability", "chunk_index": 15, "token_count": 333, "text_raw": "In this section, we invoke the `plot_ensemble()` function to visualise the variance values of filtered daily precipitation anomalies over the period 1940–2014 for the following: (a) the reference ERA5 product, (b) the ensemble median (defined as the median of the variance values for the selected subset of models, calculated for each grid cell), (c) the bias of the ensemble median, and (d) the ensemble spread (calculated as the standard deviation of the variance for the selected subset of models). The same layout of maps is displayed for the coefficient of variation. As mentioned throughout this notebook, before calculating the variance and coefficient of variation, a Lanczos band-pass filter has been applied to the daily precipitation anomalies to isolate the 10–20-day high-frequency intraseasonal oscillations (HFISOs) and the 30–60-day low-frequency intraseasonal oscillations (LFISOs).\n\nChange default colorbars\nColourbar limits\nFig number counter\nCommon title\nColourbar kwargs for coefficient of variation\nSelect colourbar kwargs based on variable type\n\n(climate_projections-cmip6_validation_q11:section-3.3)=", "text_with_prefix": "EQC Quality Assessment: \"Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q11 | Category: Climate_Projections\nSection: Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability > Analysis and results > 3. Plot and describe results > 3.2. Plot ensemble maps\n---\nIn this section, we invoke the `plot_ensemble()` function to visualise the variance values of filtered daily precipitation anomalies over the period 1940–2014 for the following: (a) the reference ERA5 product, (b) the ensemble median (defined as the median of the variance values for the selected subset of models, calculated for each grid cell), (c) the bias of the ensemble median, and (d) the ensemble spread (calculated as the standard deviation of the variance for the selected subset of models). The same layout of maps is displayed for the coefficient of variation. As mentioned throughout this notebook, before calculating the variance and coefficient of variation, a Lanczos band-pass filter has been applied to the daily precipitation anomalies to isolate the 10–20-day high-frequency intraseasonal oscillations (HFISOs) and the 30–60-day low-frequency intraseasonal oscillations (LFISOs).\n\nChange default colorbars\nColourbar limits\nFig number counter\nCommon title\nColourbar kwargs for coefficient of variation\nSelect colourbar kwargs based on variable type\n\n(climate_projections-cmip6_validation_q11:section-3.3)="} {"chunk_id": "climate_projections-cmip6_validation_q11__bfd0d526138c", "report_id": "climate_projections-cmip6_validation_q11", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q11", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability > Analysis and results > 3. Plot and describe results > 3.3. Plot model maps", "title": "Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability", "chunk_index": 16, "token_count": 234, "text_raw": "In this section, we invoke the `plot_models()` function to visualise the variance and the coefficient of variation of filtered daily precipitation anomalies over the period 1940–201, for every model individually. As mentioned throughout this notebook, before calculating the variance and coefficient of variation, a Lanczos band-pass filter has been applied to the daily precipitation anomalies to isolate the 10–20-day high-frequency intraseasonal oscillations (HFISOs) and the 30–60-day low-frequency intraseasonal oscillations (LFISOs). Note that the model data used in this section maintains its original grid.\n\n(climate_projections-cmip6_validation_q11:section-3.4)=", "text_with_prefix": "EQC Quality Assessment: \"Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q11 | Category: Climate_Projections\nSection: Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability > Analysis and results > 3. Plot and describe results > 3.3. Plot model maps\n---\nIn this section, we invoke the `plot_models()` function to visualise the variance and the coefficient of variation of filtered daily precipitation anomalies over the period 1940–201, for every model individually. As mentioned throughout this notebook, before calculating the variance and coefficient of variation, a Lanczos band-pass filter has been applied to the daily precipitation anomalies to isolate the 10–20-day high-frequency intraseasonal oscillations (HFISOs) and the 30–60-day low-frequency intraseasonal oscillations (LFISOs). Note that the model data used in this section maintains its original grid.\n\n(climate_projections-cmip6_validation_q11:section-3.4)="} {"chunk_id": "climate_projections-cmip6_validation_q11__2ebc283e2ac8", "report_id": "climate_projections-cmip6_validation_q11", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q11", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability > Analysis and results > 3. Plot and describe results > 3.4. Plot bias maps", "title": "Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability", "chunk_index": 17, "token_count": 243, "text_raw": "In this section, we invoke the `plot_models()` function to visualise the bias of the variance and the coefficient of variation for filtered daily precipitation anomalies over the period 1940–2014, for every model individually. As mentioned throughout this notebook, before calculating the variance and coefficient of variation, a Lanczos band-pass filter has been applied to the daily precipitation anomalies to isolate the 10–20-day high-frequency intraseasonal oscillations (HFISOs) and the 30–60-day low-frequency intraseasonal oscillations (LFISOs). Note that the model data used in this section has previously been interpolated to the ERA5 grid.\n\n(climate_projections-cmip6_validation_q11:section-3.5)=", "text_with_prefix": "EQC Quality Assessment: \"Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q11 | Category: Climate_Projections\nSection: Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability > Analysis and results > 3. Plot and describe results > 3.4. Plot bias maps\n---\nIn this section, we invoke the `plot_models()` function to visualise the bias of the variance and the coefficient of variation for filtered daily precipitation anomalies over the period 1940–2014, for every model individually. As mentioned throughout this notebook, before calculating the variance and coefficient of variation, a Lanczos band-pass filter has been applied to the daily precipitation anomalies to isolate the 10–20-day high-frequency intraseasonal oscillations (HFISOs) and the 30–60-day low-frequency intraseasonal oscillations (LFISOs). Note that the model data used in this section has previously been interpolated to the ERA5 grid.\n\n(climate_projections-cmip6_validation_q11:section-3.5)="} {"chunk_id": "climate_projections-cmip6_validation_q11__3b39f9ee89b1", "report_id": "climate_projections-cmip6_validation_q11", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q11", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability > Analysis and results > 3. Plot and describe results > 3.5. Timeseries of the coefficient of variation", "title": "Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability", "chunk_index": 18, "token_count": 540, "text_raw": "This section examines the time series of spatially averaged variance and coefficient of variation, calculated from filtered daily precipitation anomalies over the historical period using a 30-year window. The analysis compares the CMIP6 ensemble median, ERA5, and individual models, highlighting trends for both the CMIP6 ensemble median and ERA5. It also displays the ensemble spread, along with the slope values of the trends for ERA5 and the CMIP6 ensemble median.\n\nAs mentioned throughout this notebook, before calculating the variance and coefficient of variation, a Lanczos band-pass filter has been applied to the daily precipitation anomalies to isolate the 10–20-day high-frequency intraseasonal oscillations (HFISOs) and the 30–60-day low-frequency intraseasonal oscillations (LFISOs).\n\nDefine the colors\nCutout region\nGet the spatial mean for ERA5\nGet the spatial mean for CMIP6, preserving each model\nCalculate the median, ensemble mean, and standard deviation\nInitialize list to store trend information\nPlot individual model series for each variance\n\n
\n
\n

Fig 13. Time series of the spatially averaged values for the chosen variability formulations: (a) variance and (b) coefficient of variation, for filtered daily precipitation anomalies over the period 1940–2014. The variance and coefficient of variation have been calculated using a 30-year window. Each time series displays the CMIP6 ensemble median (green), ERA5 (black), and individual models (grey), with dashed lines indicating the trend for the CMIP6 ensemble median and ERA5. The shaded area represents the ensemble spread. A grey box at the bottom displays the trend slope. Before calculating the variance and coefficient of variation, a Lanczos band-pass filter has been applied to the daily precipitation anomalies to isolate the 10–20-day high-frequency intraseasonal oscillations (HFISOs) and the 30–60-day low-frequency intraseasonal oscillations (LFISOs).

\n
\n\n(climate_projections-cmip6_validation_q11:section-3.6)=", "text_with_prefix": "EQC Quality Assessment: \"Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q11 | Category: Climate_Projections\nSection: Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability > Analysis and results > 3. Plot and describe results > 3.5. Timeseries of the coefficient of variation\n---\nThis section examines the time series of spatially averaged variance and coefficient of variation, calculated from filtered daily precipitation anomalies over the historical period using a 30-year window. The analysis compares the CMIP6 ensemble median, ERA5, and individual models, highlighting trends for both the CMIP6 ensemble median and ERA5. It also displays the ensemble spread, along with the slope values of the trends for ERA5 and the CMIP6 ensemble median.\n\nAs mentioned throughout this notebook, before calculating the variance and coefficient of variation, a Lanczos band-pass filter has been applied to the daily precipitation anomalies to isolate the 10–20-day high-frequency intraseasonal oscillations (HFISOs) and the 30–60-day low-frequency intraseasonal oscillations (LFISOs).\n\nDefine the colors\nCutout region\nGet the spatial mean for ERA5\nGet the spatial mean for CMIP6, preserving each model\nCalculate the median, ensemble mean, and standard deviation\nInitialize list to store trend information\nPlot individual model series for each variance\n\n
\n
\n

Fig 13. Time series of the spatially averaged values for the chosen variability formulations: (a) variance and (b) coefficient of variation, for filtered daily precipitation anomalies over the period 1940–2014. The variance and coefficient of variation have been calculated using a 30-year window. Each time series displays the CMIP6 ensemble median (green), ERA5 (black), and individual models (grey), with dashed lines indicating the trend for the CMIP6 ensemble median and ERA5. The shaded area represents the ensemble spread. A grey box at the bottom displays the trend slope. Before calculating the variance and coefficient of variation, a Lanczos band-pass filter has been applied to the daily precipitation anomalies to isolate the 10–20-day high-frequency intraseasonal oscillations (HFISOs) and the 30–60-day low-frequency intraseasonal oscillations (LFISOs).

\n
\n\n(climate_projections-cmip6_validation_q11:section-3.6)="} {"chunk_id": "climate_projections-cmip6_validation_q11__ee9c47d52a78", "report_id": "climate_projections-cmip6_validation_q11", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q11", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability > Analysis and results > 3. Plot and describe results > 3.6. Results summary and discussion", "title": "Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability", "chunk_index": 19, "token_count": 602, "text_raw": "* An assessment evaluating the [capability of CMIP6 GCMs to represent inter-annual precipitation variability](https://ecmwf-projects.github.io/c3s2-eqc-quality-assessment/Climate_Projections/CMIP6/climate_projections-cmip6_validation_q10.html#) complements the results presented in this notebook.\n\n* Intra-seasonal variability during the Monsoon season (JJAS) shows substantial differences in spatial patterns depending on the formulation used—variance or the coefficient of variation (standard deviation divided by the mean). As detailed throughout, a Lanczos band-pass filter is applied to daily precipitation anomalies to isolate 10–20-day high-frequency (HFISOs) and 30–60-day low-frequency (LFISOs) oscillations. Anomalies are computed relative to the daily climatology to remove the seasonal cycle.\n\n* Using variance to assess inter-annual variability reveals large differences in bias sign, magnitude, and spatial distribution. Despite model differences, biases are consistently influenced by orography and high-precipitation regions. These regions show high absolute biases due to high JJAS precipitation, though not necessarily large relative to the mean. The coefficient of variation (CV) normalises this effect, offering a potentially more meaningful measure of variability—though it can yield high values where climatological means are near zero.\n\n* Differences between models are smaller when using CV compared to variance. Inland north-western regions (with near-zero or very low seasonal averages) show strong negative biases and high inter-model spread. One hypothesis to explain this may be that unusually wet spells are underestimated in comparison to ERA5.\n\n* The spatial patterns of variance and CV for HFISOs and LFISOs are largely similar.\n\n* Intra-seasonal variability patterns strongly resemble those seen in the inter-annual assessment, suggesting intra-seasonal timescales may influence inter-annual variability [[10]](https://doi.org/10.1175/JCLI-D-18-0319.1).\n\n* CMIP6 ensemble median does not accurately capture the temporal evolution of variance and coefficient of variation in filtered daily anomalies averaged over India (HFISOs and LFISOs). While ERA5 shows a decreasing trend in variance and increasing trend in coefficient of variation, CMIP6 models indicate opposite trends, highlighting discrepancies in temporal representation. The trends for the selection region are statistically significant with p-values below 0.05.\n\n(climate_projections-cmip6_validation_q11:section-3.7)=", "text_with_prefix": "EQC Quality Assessment: \"Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q11 | Category: Climate_Projections\nSection: Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability > Analysis and results > 3. Plot and describe results > 3.6. Results summary and discussion\n---\n* An assessment evaluating the [capability of CMIP6 GCMs to represent inter-annual precipitation variability](https://ecmwf-projects.github.io/c3s2-eqc-quality-assessment/Climate_Projections/CMIP6/climate_projections-cmip6_validation_q10.html#) complements the results presented in this notebook.\n\n* Intra-seasonal variability during the Monsoon season (JJAS) shows substantial differences in spatial patterns depending on the formulation used—variance or the coefficient of variation (standard deviation divided by the mean). As detailed throughout, a Lanczos band-pass filter is applied to daily precipitation anomalies to isolate 10–20-day high-frequency (HFISOs) and 30–60-day low-frequency (LFISOs) oscillations. Anomalies are computed relative to the daily climatology to remove the seasonal cycle.\n\n* Using variance to assess inter-annual variability reveals large differences in bias sign, magnitude, and spatial distribution. Despite model differences, biases are consistently influenced by orography and high-precipitation regions. These regions show high absolute biases due to high JJAS precipitation, though not necessarily large relative to the mean. The coefficient of variation (CV) normalises this effect, offering a potentially more meaningful measure of variability—though it can yield high values where climatological means are near zero.\n\n* Differences between models are smaller when using CV compared to variance. Inland north-western regions (with near-zero or very low seasonal averages) show strong negative biases and high inter-model spread. One hypothesis to explain this may be that unusually wet spells are underestimated in comparison to ERA5.\n\n* The spatial patterns of variance and CV for HFISOs and LFISOs are largely similar.\n\n* Intra-seasonal variability patterns strongly resemble those seen in the inter-annual assessment, suggesting intra-seasonal timescales may influence inter-annual variability [[10]](https://doi.org/10.1175/JCLI-D-18-0319.1).\n\n* CMIP6 ensemble median does not accurately capture the temporal evolution of variance and coefficient of variation in filtered daily anomalies averaged over India (HFISOs and LFISOs). While ERA5 shows a decreasing trend in variance and increasing trend in coefficient of variation, CMIP6 models indicate opposite trends, highlighting discrepancies in temporal representation. The trends for the selection region are statistically significant with p-values below 0.05.\n\n(climate_projections-cmip6_validation_q11:section-3.7)="} {"chunk_id": "climate_projections-cmip6_validation_q11__7ba03f516755", "report_id": "climate_projections-cmip6_validation_q11", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q11", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability > Analysis and results > 3. Plot and describe results > 3.7. Implications for the users", "title": "Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability", "chunk_index": 20, "token_count": 712, "text_raw": "* **The selection of the variability formulation matters**. The Coefficient of Variation (CV) reduces orographic influence by normalising variability by the mean, and shows smaller inter-model differences than variance. However, it can produce artefactually high values in north-western inland regions where mean precipitation is very low.\n\n* **ERA5 has biases** [[13]](https://doi.org/10.1002/qj.3616)[[14]](https://doi.org/10.3390/atmos12111462). While ERA5 is used for its temporal coverage, it has known biases. Users may consider alternative references like the [GPCP](https://cds.climate.copernicus.eu/datasets/satellite-precipitation?tab=overview) dataset, which may yield different results.\n\n* Unlike the [inter-annual assessment](https://ecmwf-projects.github.io/c3s2-eqc-quality-assessment/Climate_Projections/CMIP6/climate_projections-cmip6_validation_q10.html#), CV and variance show opposite trends over time. ERA5 indicates a decline in variance and a rise in the coefficient of variation, while CMIP6 models show the reverse.\n\n* It is **important to consider a larger ensemble and acknowledge that the time series in this assessment represent spatially averaged values**, which may lead to different results when focusing on specific sub-regions.\n\n* **Frequency component consistency**: No significant differences are seen in the spatial or temporal patterns between the low-frequency (LFISOs) and high-frequency (HFISOs) components of intra-seasonal oscillations.\n\n* Given the challenges in replicating the spatial patterns of variance and the coefficient of variation, as well as the difficulties in capturing their temporal evolution, **bias correction of CMIP6 projections may be essential for effectively using future projections of intra-seasonal precipitation variability. Such improvements could facilitate the adoption of CMIP6 future projections in the water management sector, aiding in the design of strategies to adapt to projected changes in precipitation.**\n\n* **Intra-seasonal and inter-annual links.** The spatial patterns of intra-seasonal variability closely resemble those found in the inter-annual variability assessment. Intra-seasonal variability may influence inter-annual variability [[10]](https://doi.org/10.1175/JCLI-D-18-0319.1) and, therefore, it may be important to bias correct data at the daily timescale.\n\n* **Assessments using regional climate models (RCMs) may produce different results** (not necessarily better) due to their higher spatial resolution, better representation of local processes, and more detailed treatment of orography and land-atmosphere interactions.\n\n\n
\nRESULTS NOTE:
\nIt is important to note that the results presented are specific to the 16 models chosen, and users should aim to assess as wide a range of models as possible before making a sub-selection.", "text_with_prefix": "EQC Quality Assessment: \"Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q11 | Category: Climate_Projections\nSection: Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability > Analysis and results > 3. Plot and describe results > 3.7. Implications for the users\n---\n* **The selection of the variability formulation matters**. The Coefficient of Variation (CV) reduces orographic influence by normalising variability by the mean, and shows smaller inter-model differences than variance. However, it can produce artefactually high values in north-western inland regions where mean precipitation is very low.\n\n* **ERA5 has biases** [[13]](https://doi.org/10.1002/qj.3616)[[14]](https://doi.org/10.3390/atmos12111462). While ERA5 is used for its temporal coverage, it has known biases. Users may consider alternative references like the [GPCP](https://cds.climate.copernicus.eu/datasets/satellite-precipitation?tab=overview) dataset, which may yield different results.\n\n* Unlike the [inter-annual assessment](https://ecmwf-projects.github.io/c3s2-eqc-quality-assessment/Climate_Projections/CMIP6/climate_projections-cmip6_validation_q10.html#), CV and variance show opposite trends over time. ERA5 indicates a decline in variance and a rise in the coefficient of variation, while CMIP6 models show the reverse.\n\n* It is **important to consider a larger ensemble and acknowledge that the time series in this assessment represent spatially averaged values**, which may lead to different results when focusing on specific sub-regions.\n\n* **Frequency component consistency**: No significant differences are seen in the spatial or temporal patterns between the low-frequency (LFISOs) and high-frequency (HFISOs) components of intra-seasonal oscillations.\n\n* Given the challenges in replicating the spatial patterns of variance and the coefficient of variation, as well as the difficulties in capturing their temporal evolution, **bias correction of CMIP6 projections may be essential for effectively using future projections of intra-seasonal precipitation variability. Such improvements could facilitate the adoption of CMIP6 future projections in the water management sector, aiding in the design of strategies to adapt to projected changes in precipitation.**\n\n* **Intra-seasonal and inter-annual links.** The spatial patterns of intra-seasonal variability closely resemble those found in the inter-annual variability assessment. Intra-seasonal variability may influence inter-annual variability [[10]](https://doi.org/10.1175/JCLI-D-18-0319.1) and, therefore, it may be important to bias correct data at the daily timescale.\n\n* **Assessments using regional climate models (RCMs) may produce different results** (not necessarily better) due to their higher spatial resolution, better representation of local processes, and more detailed treatment of orography and land-atmosphere interactions.\n\n\n
\nRESULTS NOTE:
\nIt is important to note that the results presented are specific to the 16 models chosen, and users should aim to assess as wide a range of models as possible before making a sub-selection."} {"chunk_id": "climate_projections-cmip6_validation_q11__fad8dd2a48fb", "report_id": "climate_projections-cmip6_validation_q11", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q11", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability > ℹ️ If you want to know more > Key resources", "title": "Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability", "chunk_index": 21, "token_count": 252, "text_raw": "Some key resources and further reading were linked throughout this assessment.\n\nThe CDS catalogue entries for the data used were:\n* CMIP6 climate projections (Daily - Precipitation): https://cds.climate.copernicus.eu/datasets/projections-cmip6?tab=overview\n* ERA5 hourly data on single levels from 1940 to present (total precipitation): https://cds.climate.copernicus.eu/datasets/reanalysis-era5-single-levels?tab=overview\n\nCode libraries used:\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)", "text_with_prefix": "EQC Quality Assessment: \"Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q11 | Category: Climate_Projections\nSection: Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability > ℹ️ If you want to know more > Key resources\n---\nSome key resources and further reading were linked throughout this assessment.\n\nThe CDS catalogue entries for the data used were:\n* CMIP6 climate projections (Daily - Precipitation): https://cds.climate.copernicus.eu/datasets/projections-cmip6?tab=overview\n* ERA5 hourly data on single levels from 1940 to present (total precipitation): https://cds.climate.copernicus.eu/datasets/reanalysis-era5-single-levels?tab=overview\n\nCode libraries used:\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)"} {"chunk_id": "climate_projections-cmip6_validation_q11__78508175f26b", "report_id": "climate_projections-cmip6_validation_q11", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q11", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability > ℹ️ If you want to know more > References", "title": "Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability", "chunk_index": 22, "token_count": 1031, "text_raw": "[[1]](https://doi.org/10.1111/gcb.12581) Thornton, P. K., Ericksen, P. J., Herrero, M., and Challinor, A. J., 2014. Climate variability and vulnerability to climate change: A review. Global Change Biol. 20, https://doi.org/10.1111/gcb.12581\n\n[[2]](https://doi.org/10.1016/j.wace.2020.100277) Masroor, M., Rehman, S., Avtar, R., Sahana, M., Ahmed, R., Sajjad, H., 2020. Exploring climate variability and its impact on drought occurrence: evidence from Godavari Middle sub-basin, India. Weather and Climate Extremes. 30, 100277. https://doi.org/10.1016/j.wace.2020.100277\n\n[[3]](https://doi.org/10.1111/j.2153-3490.1980.tb01717.x) Krishnamurti, T. N. and Ardanuy, P., 1980. 10 to 20-day westward propagating mode and “breaks in the monsoons”. Tellus, 32, 15–26. https://doi.org/10.1111/j.2153-3490.1980.tb01717.x\n\n[[4]](https://doi.org/10.1175/2007MWR1991.1) Wen, M. and Zhang, R., 2007. Role of the quasi-biweekly oscillation in the onset of convection over the Indochina Peninsula, Q. J. R. Meteor. Soc., 133, 433–444. https://doi.org/10.1002/qj.38\n\n[[5]](https://doi.org/10.1175/2011JCLI3916.1) Wen, M., Yang, S., Higgins, R. W., Zhang, R., 2011. Characteristics of the dominant modes of atmospheric quasi-biweekly oscillation over tropical-subtropical Americas, J. Climate, 24, 3956–3970. https://doi.org/10.1175/2011JCLI3916.1\n\n[[6]](https://doi.org/10.1175/1520-0469%281971%29028%3C0702:DOADOI%3E2.0.CO;2) Madden, R. A. and Julian P. R., 1971. Detection of a 40–50 day oscillation in the zonal wind in the tropical Pacific, J. Atmos. Sci., 28, 702–708. https://doi.org/10.1175/1520-0469%281971%29028%3C0702:DOADOI%3E2.0.CO;2\n\n[[7]](https://doi.org/10.1175/1520-0469%281972%29029<1109:DOGSCC>2.0.CO;2) Madden, R. A. and Julian P. R., 1972. Description of global scale circulation cells in the tropics with 40–50 day period, J. Atmos. Sci., 29, 1109–1123. https://doi.org/10.1175/1520-0469%281972%29029<1109:DOGSCC>2.0.CO;2\n\n[[8]](https://doi.org/10.1038/s41612-022-00253-7) Liu, F., Wang, B., Ouyang, Y., Wang, H., Qiao,S., Chen, G., Dong, W., 2022. Intraseasonal variability of global land monsoon precipitation and its recent trend. npj Clim Atmos Sci 5, 30. https://doi.org/10.1038/s41612-022-00253-7\n\n[[9]](https://doi.org/10.1002/jgrd.50422), Jia, X. and Yang, S., 2013. Impact of the quasi-biweekly oscillation over the western North Pacific on East Asian subtropical monsoon during early summer, J. Geophys. Res. Atmos., 118, 4421–4434, https://doi.org/10.1002/jgrd.50422", "text_with_prefix": "EQC Quality Assessment: \"Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q11 | Category: Climate_Projections\nSection: Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability > ℹ️ If you want to know more > References\n---\n[[1]](https://doi.org/10.1111/gcb.12581) Thornton, P. K., Ericksen, P. J., Herrero, M., and Challinor, A. J., 2014. Climate variability and vulnerability to climate change: A review. Global Change Biol. 20, https://doi.org/10.1111/gcb.12581\n\n[[2]](https://doi.org/10.1016/j.wace.2020.100277) Masroor, M., Rehman, S., Avtar, R., Sahana, M., Ahmed, R., Sajjad, H., 2020. Exploring climate variability and its impact on drought occurrence: evidence from Godavari Middle sub-basin, India. Weather and Climate Extremes. 30, 100277. https://doi.org/10.1016/j.wace.2020.100277\n\n[[3]](https://doi.org/10.1111/j.2153-3490.1980.tb01717.x) Krishnamurti, T. N. and Ardanuy, P., 1980. 10 to 20-day westward propagating mode and “breaks in the monsoons”. Tellus, 32, 15–26. https://doi.org/10.1111/j.2153-3490.1980.tb01717.x\n\n[[4]](https://doi.org/10.1175/2007MWR1991.1) Wen, M. and Zhang, R., 2007. Role of the quasi-biweekly oscillation in the onset of convection over the Indochina Peninsula, Q. J. R. Meteor. Soc., 133, 433–444. https://doi.org/10.1002/qj.38\n\n[[5]](https://doi.org/10.1175/2011JCLI3916.1) Wen, M., Yang, S., Higgins, R. W., Zhang, R., 2011. Characteristics of the dominant modes of atmospheric quasi-biweekly oscillation over tropical-subtropical Americas, J. Climate, 24, 3956–3970. https://doi.org/10.1175/2011JCLI3916.1\n\n[[6]](https://doi.org/10.1175/1520-0469%281971%29028%3C0702:DOADOI%3E2.0.CO;2) Madden, R. A. and Julian P. R., 1971. Detection of a 40–50 day oscillation in the zonal wind in the tropical Pacific, J. Atmos. Sci., 28, 702–708. https://doi.org/10.1175/1520-0469%281971%29028%3C0702:DOADOI%3E2.0.CO;2\n\n[[7]](https://doi.org/10.1175/1520-0469%281972%29029<1109:DOGSCC>2.0.CO;2) Madden, R. A. and Julian P. R., 1972. Description of global scale circulation cells in the tropics with 40–50 day period, J. Atmos. Sci., 29, 1109–1123. https://doi.org/10.1175/1520-0469%281972%29029<1109:DOGSCC>2.0.CO;2\n\n[[8]](https://doi.org/10.1038/s41612-022-00253-7) Liu, F., Wang, B., Ouyang, Y., Wang, H., Qiao,S., Chen, G., Dong, W., 2022. Intraseasonal variability of global land monsoon precipitation and its recent trend. npj Clim Atmos Sci 5, 30. https://doi.org/10.1038/s41612-022-00253-7\n\n[[9]](https://doi.org/10.1002/jgrd.50422), Jia, X. and Yang, S., 2013. Impact of the quasi-biweekly oscillation over the western North Pacific on East Asian subtropical monsoon during early summer, J. Geophys. Res. Atmos., 118, 4421–4434, https://doi.org/10.1002/jgrd.50422"} {"chunk_id": "climate_projections-cmip6_validation_q11__98b5831c9995", "report_id": "climate_projections-cmip6_validation_q11", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q11", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability > ℹ️ If you want to know more > References", "title": "Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability", "chunk_index": 23, "token_count": 618, "text_raw": "North Pacific on East Asian subtropical monsoon during early summer, J. Geophys. Res. Atmos., 118, 4421–4434, https://doi.org/10.1002/jgrd.50422\n\n[[10]](https://doi.org/10.1175/JCLI-D-18-0319.1) Jiang, X. and Ting, M., 2019. Intraseasonal Variability of Rainfall and Its Effect on Interannual Variability across the Indian Subcontinent and the Tibetan Plateau. J. Climate, 32, 2227–2245, https://doi.org/10.1175/JCLI-D-18-0319.1\n\n[[11]](https://doi.org/10.1175/JCLI-D-16-0360.1) Hatsuzuka, D., and Fujinami, H., 2017. Effects of the South Asian Monsoon Intraseasonal Modes on Genesis of Low Pressure Systems over Bangladesh. J. Climate, 30, 2481–2499, https://doi.org/10.1175/JCLI-D-16-0360.1\n\n[[12]](https://doi.org/10.1007/s00382-019-04946-3) Yan, X., Yang, S., Wang, T., Maloney, E.D., Dong, S., Wei, W., He, S., 2019. Quasi-biweekly oscillation of the Asian monsoon rainfall in late summer and autumn: different types of structure and propagation. Clim Dyn 53, 6611–6628. https://doi.org/10.1007/s00382-019-04946-3\n\n[[13]](https://doi.org/10.1002/qj.3616) Ramon, J., Lledó L., Torralba, V., Soret, A., Doblas-Reyes, F.J., 2019. What global reanalysis best represents near-surface winds?. Q J R Meteorol Soc. 2019; 145: 3236–3251. https://doi.org/10.1002/qj.3616\n\n[[14]](https://doi.org/10.3390/atmos12111462) Hassler, B., and Lauer, A., 2021. Comparison of Reanalysis and Observational Precipitation Datasets Including ERA5 and WFDE5. Atmosphere, 12, 1462. https://doi.org/10.3390/atmos12111462", "text_with_prefix": "EQC Quality Assessment: \"Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q11 | Category: Climate_Projections\nSection: Testing the capability of CMIP6 GCMs to represent precipitation intra-seasonal variability > ℹ️ If you want to know more > References\n---\nNorth Pacific on East Asian subtropical monsoon during early summer, J. Geophys. Res. Atmos., 118, 4421–4434, https://doi.org/10.1002/jgrd.50422\n\n[[10]](https://doi.org/10.1175/JCLI-D-18-0319.1) Jiang, X. and Ting, M., 2019. Intraseasonal Variability of Rainfall and Its Effect on Interannual Variability across the Indian Subcontinent and the Tibetan Plateau. J. Climate, 32, 2227–2245, https://doi.org/10.1175/JCLI-D-18-0319.1\n\n[[11]](https://doi.org/10.1175/JCLI-D-16-0360.1) Hatsuzuka, D., and Fujinami, H., 2017. Effects of the South Asian Monsoon Intraseasonal Modes on Genesis of Low Pressure Systems over Bangladesh. J. Climate, 30, 2481–2499, https://doi.org/10.1175/JCLI-D-16-0360.1\n\n[[12]](https://doi.org/10.1007/s00382-019-04946-3) Yan, X., Yang, S., Wang, T., Maloney, E.D., Dong, S., Wei, W., He, S., 2019. Quasi-biweekly oscillation of the Asian monsoon rainfall in late summer and autumn: different types of structure and propagation. Clim Dyn 53, 6611–6628. https://doi.org/10.1007/s00382-019-04946-3\n\n[[13]](https://doi.org/10.1002/qj.3616) Ramon, J., Lledó L., Torralba, V., Soret, A., Doblas-Reyes, F.J., 2019. What global reanalysis best represents near-surface winds?. Q J R Meteorol Soc. 2019; 145: 3236–3251. https://doi.org/10.1002/qj.3616\n\n[[14]](https://doi.org/10.3390/atmos12111462) Hassler, B., and Lauer, A., 2021. Comparison of Reanalysis and Observational Precipitation Datasets Including ERA5 and WFDE5. Atmosphere, 12, 1462. https://doi.org/10.3390/atmos12111462"} {"chunk_id": "climate_projections-cmip6_validation_q12__fed05005680f", "report_id": "climate_projections-cmip6_validation_q12", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q12", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands", "title": "Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands", "chunk_index": 0, "token_count": 112, "text_raw": "Production date: 27-06-2025\n\nProduced by: CMCC foundation - Euro-Mediterranean Center on Climate Change. Albert Martinez Boti and Lorenzo Sangelantoni.", "text_with_prefix": "EQC Quality Assessment: \"Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q12 | Category: Climate_Projections\nSection: Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands\n---\nProduction date: 27-06-2025\n\nProduced by: CMCC foundation - Euro-Mediterranean Center on Climate Change. Albert Martinez Boti and Lorenzo Sangelantoni."} {"chunk_id": "climate_projections-cmip6_validation_q12__b2de2c36fdd1", "report_id": "climate_projections-cmip6_validation_q12", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q12", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands > Quality assessment question", "title": "Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands", "chunk_index": 1, "token_count": 636, "text_raw": "* **How is warming propagating across pressure levels and different latitudinal bands?**\n* **Is there signal consistence between CMIP6 ensemble results and ERA5, and among CMIP6 ensemble GCMs?​**\n\nThe vertical profile of temperature trends is widely recognised as an important fingerprint of climate change (e.g., [[1]](https://doi.org/10.1002/joc.1756),[[2]](https://doi.org/10.1126/science.274.5290.1170)),[[3]](https://doi.org/10.1073/pnas.1305332110). One of the most studied features of this vertical structure is the amplified warming in the tropical upper troposphere compared to the surface, commonly referred to as tropical tropospheric amplification [[4]](https://doi.org/10.1126/science.1114867). This behaviour is consistent with the basic theory of moist adiabatic processes, which predicts stronger warming in the tropical free troposphere than near the surface due to latent heat release during deep convection [[5]](https://doi.org/10.1175/1520-0469%281979%29036<0415:ALRRAT>2.0.CO;2).\n\nWhile tropical amplification has been widely studied, other regions and their representation in CMIP6 models remain less explored. To develop a more comprehensive evaluation framework, it is essential to assess how warming propagates vertically across pressure levels and latitudinal bands. Understanding the vertical structure of temperature trends across these bands provides critical insights into model performance and the physical processes driving climate change. This vertical perspective is key for model developers to identify and diagnose persistent biases, such as errors in atmospheric circulation, lapse rates, and vertical energy transport, and thereby improve model simulations. Since vertical temperature profiles are considered important fingerprints of climate change, improving their representation also strengthens confidence in these signals, benefiting researchers, impact assessors, and decision-makers.\n\nThis assessment evaluates temperature mean and trend biases across multiple pressure levels using the full ensemble of [CMIP6](https://cds.climate.copernicus.eu/datasets/projections-cmip6?tab=overview) models available through the Copernicus Climate Data Store (CDS). Model outputs are compared against the [ERA5](https://cds.climate.copernicus.eu/datasets/reanalysis-era5-complete?tab=doc) reanalysis, which is used as the reference product. The analysis is performed over latitudinal bands—including Global, Northern and Southern hemispheres, Tropics, Extratropics, and Polar regions—to characterise spatial variability in biases and explore their dependence on latitude and atmospheric dynamics.", "text_with_prefix": "EQC Quality Assessment: \"Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q12 | Category: Climate_Projections\nSection: Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands > Quality assessment question\n---\n* **How is warming propagating across pressure levels and different latitudinal bands?**\n* **Is there signal consistence between CMIP6 ensemble results and ERA5, and among CMIP6 ensemble GCMs?​**\n\nThe vertical profile of temperature trends is widely recognised as an important fingerprint of climate change (e.g., [[1]](https://doi.org/10.1002/joc.1756),[[2]](https://doi.org/10.1126/science.274.5290.1170)),[[3]](https://doi.org/10.1073/pnas.1305332110). One of the most studied features of this vertical structure is the amplified warming in the tropical upper troposphere compared to the surface, commonly referred to as tropical tropospheric amplification [[4]](https://doi.org/10.1126/science.1114867). This behaviour is consistent with the basic theory of moist adiabatic processes, which predicts stronger warming in the tropical free troposphere than near the surface due to latent heat release during deep convection [[5]](https://doi.org/10.1175/1520-0469%281979%29036<0415:ALRRAT>2.0.CO;2).\n\nWhile tropical amplification has been widely studied, other regions and their representation in CMIP6 models remain less explored. To develop a more comprehensive evaluation framework, it is essential to assess how warming propagates vertically across pressure levels and latitudinal bands. Understanding the vertical structure of temperature trends across these bands provides critical insights into model performance and the physical processes driving climate change. This vertical perspective is key for model developers to identify and diagnose persistent biases, such as errors in atmospheric circulation, lapse rates, and vertical energy transport, and thereby improve model simulations. Since vertical temperature profiles are considered important fingerprints of climate change, improving their representation also strengthens confidence in these signals, benefiting researchers, impact assessors, and decision-makers.\n\nThis assessment evaluates temperature mean and trend biases across multiple pressure levels using the full ensemble of [CMIP6](https://cds.climate.copernicus.eu/datasets/projections-cmip6?tab=overview) models available through the Copernicus Climate Data Store (CDS). Model outputs are compared against the [ERA5](https://cds.climate.copernicus.eu/datasets/reanalysis-era5-complete?tab=doc) reanalysis, which is used as the reference product. The analysis is performed over latitudinal bands—including Global, Northern and Southern hemispheres, Tropics, Extratropics, and Polar regions—to characterise spatial variability in biases and explore their dependence on latitude and atmospheric dynamics."} {"chunk_id": "climate_projections-cmip6_validation_q12__66cd230a5bd7", "report_id": "climate_projections-cmip6_validation_q12", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q12", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands > Quality assessment statement", "title": "Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands", "chunk_index": 2, "token_count": 622, "text_raw": "These are the key outcomes of this assessment\n\n* Model spread in mean temperature climatology increases with altitude, especially above the tropopause, highlighting substantial uncertainty in simulating its height and structure—critical for model evaluation and development.\n\n* The median of CMIP6 climatological temperatures shows predominantly cold biases across all latitudinal bands, largest near 250–200 hPa. Near-surface and lower-stratosphere biases are smaller (though the spread is high), except in polar regions, where both the bias and the spread increase.\n\n* Temperature trends show robust tropospheric warming and stratospheric cooling across all latitudinal bands in both ERA5 and CMIP6. CMIP6 generally exhibits stronger warming trends than ERA5.\n\n* Tropical amplification of trends is evident, with CMIP6 exhibiting both stronger surface and upper-troposphere warming than ERA5, resulting also in a more pronounced amplification. Amplification is also high when considering the southern-hemisphere for both ERA5 and CMIP6.\n\n* Trend magnitudes and structures vary with dataset processing, leading to moderate to low confidence in tropospheric warming rates (IPCC AR5 [[6]](https://www.ipcc.ch/site/assets/uploads/2017/09/WG1AR5_Chapter02_FINAL.pdf)).\n\n```\n\nattachment:b8bbb2de-eba2-453c-83d4-2d349379cdda.png\n---\nwidth: 900px\nalt: Visual abstract \n---\nVertical profiles of the annual mean temperature trend (1979-2014). Each subplot corresponds to a latitudinal band (Global, Northern Hemisphere, Southern Hemisphere, Tropics, Extratropics, and Polar). ERA5 is shown as a green line, the CMIP6 ensemble median as a solid black line, individual CMIP6 models as grey dashed lines, and the ensemble spread as a grey shaded area. ERA5 points with non-significant trends (p-value > 0.05) are marked with green crosses. For the CMIP6 ensemble median, trend significance is assessed using agreement categories based on the approach described in the [IPCC AR6 Atlas](https://www.ipcc.ch/report/ar6/wg1/downloads/report/IPCC_AR6_WGI_Atlas.pdf) (pp. 1945–1950). Black crosses indicate pressure levels where fewer than 66% of models show statistically significant trends (p-value < 0.05). If over 66% of the models are significant but fewer than 80% agree on the sign of the trend, a black dot is placed on the CMIP6 ensemble median to indicate low directional agreement.\n```", "text_with_prefix": "EQC Quality Assessment: \"Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q12 | Category: Climate_Projections\nSection: Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands > Quality assessment statement\n---\nThese are the key outcomes of this assessment\n\n* Model spread in mean temperature climatology increases with altitude, especially above the tropopause, highlighting substantial uncertainty in simulating its height and structure—critical for model evaluation and development.\n\n* The median of CMIP6 climatological temperatures shows predominantly cold biases across all latitudinal bands, largest near 250–200 hPa. Near-surface and lower-stratosphere biases are smaller (though the spread is high), except in polar regions, where both the bias and the spread increase.\n\n* Temperature trends show robust tropospheric warming and stratospheric cooling across all latitudinal bands in both ERA5 and CMIP6. CMIP6 generally exhibits stronger warming trends than ERA5.\n\n* Tropical amplification of trends is evident, with CMIP6 exhibiting both stronger surface and upper-troposphere warming than ERA5, resulting also in a more pronounced amplification. Amplification is also high when considering the southern-hemisphere for both ERA5 and CMIP6.\n\n* Trend magnitudes and structures vary with dataset processing, leading to moderate to low confidence in tropospheric warming rates (IPCC AR5 [[6]](https://www.ipcc.ch/site/assets/uploads/2017/09/WG1AR5_Chapter02_FINAL.pdf)).\n\n```\n\nattachment:b8bbb2de-eba2-453c-83d4-2d349379cdda.png\n---\nwidth: 900px\nalt: Visual abstract \n---\nVertical profiles of the annual mean temperature trend (1979-2014). Each subplot corresponds to a latitudinal band (Global, Northern Hemisphere, Southern Hemisphere, Tropics, Extratropics, and Polar). ERA5 is shown as a green line, the CMIP6 ensemble median as a solid black line, individual CMIP6 models as grey dashed lines, and the ensemble spread as a grey shaded area. ERA5 points with non-significant trends (p-value > 0.05) are marked with green crosses. For the CMIP6 ensemble median, trend significance is assessed using agreement categories based on the approach described in the [IPCC AR6 Atlas](https://www.ipcc.ch/report/ar6/wg1/downloads/report/IPCC_AR6_WGI_Atlas.pdf) (pp. 1945–1950). Black crosses indicate pressure levels where fewer than 66% of models show statistically significant trends (p-value < 0.05). If over 66% of the models are significant but fewer than 80% agree on the sign of the trend, a black dot is placed on the CMIP6 ensemble median to indicate low directional agreement.\n```"} {"chunk_id": "climate_projections-cmip6_validation_q12__1324138ecf04", "report_id": "climate_projections-cmip6_validation_q12", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q12", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands > Methodology", "title": "Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands", "chunk_index": 3, "token_count": 893, "text_raw": "A subset of 33 models from [CMIP6](https://cds.climate.copernicus.eu/datasets/projections-cmip6?tab=overview) are used to assess temperature trends and climatology at different pressure levels, comparing them with those derived from\nthe [ERA5](https://cds.climate.copernicus.eu/datasets/reanalysis-era5-complete?tab=doc) reference product. While alternative radiosonde datasets, such as [RHARM](https://cds.climate.copernicus.eu/datasets/insitu-observations-igra-baseline-network?tab=overview) available through the CDS, exist, they are not used here because, despite extensive harmonisation efforts [[7]](https://doi.org/10.1029/2021JD035220), residual inconsistencies—such as limited station coverage, short time series, and sensitivity to data selection—could compromise global-scale analyses of vertical temperature profiles.\n\nThe analysis is performed across the following latitudinal bands:\n\n- Global (90°S to 90°N)\n\n- Northern-hemisphere (0°S to 90°N)\n\n- Southern-hemisphere (90°S to 0°N)\n\n- Tropics (23.5°S to 23.5°N)\n\n- Extratropics (23.5° to 66.5° in both hemispheres)\n\n- Polar (66.5° to 90° in both hemispheres)\n\nFor each latitudinal band, temperature climatology and trends are calculated at the following pressure levels: 925, 850, 700, 600, 500, 400, 300, 250, 200, 150, 100, 70, and 50 hPa. Results are presented using subplots—one per band—showing the vertical profiles of climatology and trends for each CMIP6 model alongside ERA5, following an approach similar to [[8]](https://doi.org/10.1175/JCLI-D-19-0998.1), who analysed ground- and satellite-based temperature trends over 1979–2018. Biases in both climatology and trends are also shown. The analysis period is 1979–2014.\n\nThe analysis and results follow the next outline:\n\n**[](climate_projections-cmip6_validation_q12:section-1)**\n * [](climate_projections-cmip6_validation_q12:section-1.1)\n * [](climate_projections-cmip6_validation_q12:section-1.2)\n * [](climate_projections-cmip6_validation_q12:section-1.3)\n * [](climate_projections-cmip6_validation_q12:section-1.4)\n * [](climate_projections-cmip6_validation_q12:section-1.5)\n * [](climate_projections-cmip6_validation_q12:section-1.6)\n\n**[](climate_projections-cmip6_validation_q12:section-2)**\n * [](climate_projections-cmip6_validation_q12:section-2.1)\n * [](climate_projections-cmip6_validation_q12:section-2.2)\n\n**[](climate_projections-cmip6_validation_q12:section-3)**\n * [](climate_projections-cmip6_validation_q12:section-3.1)\n * [](climate_projections-cmip6_validation_q12:section-3.2)\n * [](climate_projections-cmip6_validation_q12:section-3.3)\n * [](climate_projections-cmip6_validation_q12:section-3.4)\n * [](climate_projections-cmip6_validation_q12:section-3.5)\n * [](climate_projections-cmip6_validation_q12:section-3.6)\n * [](climate_projections-cmip6_validation_q12:section-3.7)", "text_with_prefix": "EQC Quality Assessment: \"Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q12 | Category: Climate_Projections\nSection: Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands > Methodology\n---\nA subset of 33 models from [CMIP6](https://cds.climate.copernicus.eu/datasets/projections-cmip6?tab=overview) are used to assess temperature trends and climatology at different pressure levels, comparing them with those derived from\nthe [ERA5](https://cds.climate.copernicus.eu/datasets/reanalysis-era5-complete?tab=doc) reference product. While alternative radiosonde datasets, such as [RHARM](https://cds.climate.copernicus.eu/datasets/insitu-observations-igra-baseline-network?tab=overview) available through the CDS, exist, they are not used here because, despite extensive harmonisation efforts [[7]](https://doi.org/10.1029/2021JD035220), residual inconsistencies—such as limited station coverage, short time series, and sensitivity to data selection—could compromise global-scale analyses of vertical temperature profiles.\n\nThe analysis is performed across the following latitudinal bands:\n\n- Global (90°S to 90°N)\n\n- Northern-hemisphere (0°S to 90°N)\n\n- Southern-hemisphere (90°S to 0°N)\n\n- Tropics (23.5°S to 23.5°N)\n\n- Extratropics (23.5° to 66.5° in both hemispheres)\n\n- Polar (66.5° to 90° in both hemispheres)\n\nFor each latitudinal band, temperature climatology and trends are calculated at the following pressure levels: 925, 850, 700, 600, 500, 400, 300, 250, 200, 150, 100, 70, and 50 hPa. Results are presented using subplots—one per band—showing the vertical profiles of climatology and trends for each CMIP6 model alongside ERA5, following an approach similar to [[8]](https://doi.org/10.1175/JCLI-D-19-0998.1), who analysed ground- and satellite-based temperature trends over 1979–2018. Biases in both climatology and trends are also shown. The analysis period is 1979–2014.\n\nThe analysis and results follow the next outline:\n\n**[](climate_projections-cmip6_validation_q12:section-1)**\n * [](climate_projections-cmip6_validation_q12:section-1.1)\n * [](climate_projections-cmip6_validation_q12:section-1.2)\n * [](climate_projections-cmip6_validation_q12:section-1.3)\n * [](climate_projections-cmip6_validation_q12:section-1.4)\n * [](climate_projections-cmip6_validation_q12:section-1.5)\n * [](climate_projections-cmip6_validation_q12:section-1.6)\n\n**[](climate_projections-cmip6_validation_q12:section-2)**\n * [](climate_projections-cmip6_validation_q12:section-2.1)\n * [](climate_projections-cmip6_validation_q12:section-2.2)\n\n**[](climate_projections-cmip6_validation_q12:section-3)**\n * [](climate_projections-cmip6_validation_q12:section-3.1)\n * [](climate_projections-cmip6_validation_q12:section-3.2)\n * [](climate_projections-cmip6_validation_q12:section-3.3)\n * [](climate_projections-cmip6_validation_q12:section-3.4)\n * [](climate_projections-cmip6_validation_q12:section-3.5)\n * [](climate_projections-cmip6_validation_q12:section-3.6)\n * [](climate_projections-cmip6_validation_q12:section-3.7)"} {"chunk_id": "climate_projections-cmip6_validation_q12__bafaa569f2e2", "report_id": "climate_projections-cmip6_validation_q12", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q12", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands > Analysis and results > 1. Parameters, requests and functions definition > 1.1. Import packages", "title": "Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands", "chunk_index": 4, "token_count": 128, "text_raw": "Set the custom temporary directory\nSet the default temporary directory to the custom directory\n\n(climate_projections-cmip6_validation_q12:section-1.2)=", "text_with_prefix": "EQC Quality Assessment: \"Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q12 | Category: Climate_Projections\nSection: Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands > Analysis and results > 1. Parameters, requests and functions definition > 1.1. Import packages\n---\nSet the custom temporary directory\nSet the default temporary directory to the custom directory\n\n(climate_projections-cmip6_validation_q12:section-1.2)="} {"chunk_id": "climate_projections-cmip6_validation_q12__7c9b1a515c84", "report_id": "climate_projections-cmip6_validation_q12", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q12", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands > Analysis and results > 1. Parameters, requests and functions definition > 1.2. Define Parameters", "title": "Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands", "chunk_index": 5, "token_count": 300, "text_raw": "In the \"Define Parameters\" section, various customisable options for the notebook are specified:\n\n- The start and end years of the historical period can be adjusted using the parameters `year_start` and `year_stop`. The default period (1979–2014) is selected to ensure consistency with the satellite era in ERA5, which began assimilating satellite observations in 1979.\n- The `chunk` selection allows the user to define if dividing into chunks when downloading the data on their local machine. Although it does not significantly affect the analysis, it is recommended to keep the default value for optimal performance.\n- `variable` and `collection_id` are not customisable for this assessment and are set to 'temperature' and 'CMIP6'. Expert users can use this notebook as a guide to create similar analyses for other variables or model sets (such as CORDEX).\n\nInterpolation method\nChunks for download\nCollection id\n\n(climate_projections-cmip6_validation_q12:section-1.3)=", "text_with_prefix": "EQC Quality Assessment: \"Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q12 | Category: Climate_Projections\nSection: Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands > Analysis and results > 1. Parameters, requests and functions definition > 1.2. Define Parameters\n---\nIn the \"Define Parameters\" section, various customisable options for the notebook are specified:\n\n- The start and end years of the historical period can be adjusted using the parameters `year_start` and `year_stop`. The default period (1979–2014) is selected to ensure consistency with the satellite era in ERA5, which began assimilating satellite observations in 1979.\n- The `chunk` selection allows the user to define if dividing into chunks when downloading the data on their local machine. Although it does not significantly affect the analysis, it is recommended to keep the default value for optimal performance.\n- `variable` and `collection_id` are not customisable for this assessment and are set to 'temperature' and 'CMIP6'. Expert users can use this notebook as a guide to create similar analyses for other variables or model sets (such as CORDEX).\n\nInterpolation method\nChunks for download\nCollection id\n\n(climate_projections-cmip6_validation_q12:section-1.3)="} {"chunk_id": "climate_projections-cmip6_validation_q12__c634d9978917", "report_id": "climate_projections-cmip6_validation_q12", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q12", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands > Analysis and results > 1. Parameters, requests and functions definition > 1.3. Define models", "title": "Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands", "chunk_index": 6, "token_count": 172, "text_raw": "The following climate analyses are performed considering the CMIP6 models that have available both the historical and SSP5-8.5 experiments within the Climate Data Store ([CDS](https://cds.climate.copernicus.eu/datasets/projections-cmip6?tab=overview)).\n\n(climate_projections-cmip6_validation_q12:section-1.4)=", "text_with_prefix": "EQC Quality Assessment: \"Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q12 | Category: Climate_Projections\nSection: Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands > Analysis and results > 1. Parameters, requests and functions definition > 1.3. Define models\n---\nThe following climate analyses are performed considering the CMIP6 models that have available both the historical and SSP5-8.5 experiments within the Climate Data Store ([CDS](https://cds.climate.copernicus.eu/datasets/projections-cmip6?tab=overview)).\n\n(climate_projections-cmip6_validation_q12:section-1.4)="} {"chunk_id": "climate_projections-cmip6_validation_q12__47f0f6c40ab9", "report_id": "climate_projections-cmip6_validation_q12", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q12", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands > Analysis and results > 1. Parameters, requests and functions definition > 1.4. Define ERA5 request", "title": "Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands", "chunk_index": 7, "token_count": 171, "text_raw": "Within this notebook, ERA5 serves as the reference product. In this section, we set the required parameters for the cds-api data-request of ERA5.\n\n`request_era_ps_mask` is used to retrieve surface pressure data used to identify and mask out pressure levels below the surface.\n\n(climate_projections-cmip6_validation_q12:section-1.5)=", "text_with_prefix": "EQC Quality Assessment: \"Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q12 | Category: Climate_Projections\nSection: Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands > Analysis and results > 1. Parameters, requests and functions definition > 1.4. Define ERA5 request\n---\nWithin this notebook, ERA5 serves as the reference product. In this section, we set the required parameters for the cds-api data-request of ERA5.\n\n`request_era_ps_mask` is used to retrieve surface pressure data used to identify and mask out pressure levels below the surface.\n\n(climate_projections-cmip6_validation_q12:section-1.5)="} {"chunk_id": "climate_projections-cmip6_validation_q12__b0c913d6e775", "report_id": "climate_projections-cmip6_validation_q12", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q12", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands > Analysis and results > 1. Parameters, requests and functions definition > 1.5. Define model requests", "title": "Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands", "chunk_index": 8, "token_count": 134, "text_raw": "In this section we set the required parameters for the cds-api data-request.\n\n\"format\": \"zip\",\n\n(climate_projections-cmip6_validation_q12:section-1.6)=", "text_with_prefix": "EQC Quality Assessment: \"Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q12 | Category: Climate_Projections\nSection: Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands > Analysis and results > 1. Parameters, requests and functions definition > 1.5. Define model requests\n---\nIn this section we set the required parameters for the cds-api data-request.\n\n\"format\": \"zip\",\n\n(climate_projections-cmip6_validation_q12:section-1.6)="} {"chunk_id": "climate_projections-cmip6_validation_q12__5b692dd3578e", "report_id": "climate_projections-cmip6_validation_q12", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q12", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands > Analysis and results > 1. Parameters, requests and functions definition > 1.6. Functions to cache", "title": "Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands", "chunk_index": 9, "token_count": 318, "text_raw": "In this section, functions that will be executed in the caching phase are defined. Caching is the process of storing copies of files in a temporary storage location, so that they can be accessed more quickly. This process also checks if the user has already downloaded a file, avoiding redundant downloads.\n\nDescription of the main functions:\n\n- **`compute_band_trends_clim`**: Computes vertical profiles of temperature climatology and trends. It first calculates annual means, then calls `get_latband_mean`, which averages over longitudes for each latitude within the band, and subsequently across those latitudes. The resulting time series is passed to `compute_trends_clim` to estimate the climatology and linear trend.\n\n- **`compute_trends_clim`**: Computes the climatology and linear trend of a time series using ordinary least squares regression.\n\nEnsure correct slicing order for descending or ascending latitude\nharmonize pressure coordinate name\nConcatenate along pressure_level\nCompute means for each band\nAdd a new coordinate for each band and concatenate\n\n(climate_projections-cmip6_validation_q12:section-2)=", "text_with_prefix": "EQC Quality Assessment: \"Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q12 | Category: Climate_Projections\nSection: Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands > Analysis and results > 1. Parameters, requests and functions definition > 1.6. Functions to cache\n---\nIn this section, functions that will be executed in the caching phase are defined. Caching is the process of storing copies of files in a temporary storage location, so that they can be accessed more quickly. This process also checks if the user has already downloaded a file, avoiding redundant downloads.\n\nDescription of the main functions:\n\n- **`compute_band_trends_clim`**: Computes vertical profiles of temperature climatology and trends. It first calculates annual means, then calls `get_latband_mean`, which averages over longitudes for each latitude within the band, and subsequently across those latitudes. The resulting time series is passed to `compute_trends_clim` to estimate the climatology and linear trend.\n\n- **`compute_trends_clim`**: Computes the climatology and linear trend of a time series using ordinary least squares regression.\n\nEnsure correct slicing order for descending or ascending latitude\nharmonize pressure coordinate name\nConcatenate along pressure_level\nCompute means for each band\nAdd a new coordinate for each band and concatenate\n\n(climate_projections-cmip6_validation_q12:section-2)="} {"chunk_id": "climate_projections-cmip6_validation_q12__09334a521550", "report_id": "climate_projections-cmip6_validation_q12", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q12", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands > Analysis and results > 2. Downloading and processing > 2.1. Download and transform ERA5", "title": "Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands", "chunk_index": 10, "token_count": 190, "text_raw": "In this section, the `download.download_and_transform` function from the 'c3s_eqc_automatic_quality_control' package is used to retrieve ERA5 monthly reference data, compute annual means, calculate latitudinal band averages, and derive the climatology and linear trend at each pressure level for each latitudinal band. Results are cached to avoid redundant downloads and repeated processing.\n\n(climate_projections-cmip6_validation_q12:section-2.2)=", "text_with_prefix": "EQC Quality Assessment: \"Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q12 | Category: Climate_Projections\nSection: Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands > Analysis and results > 2. Downloading and processing > 2.1. Download and transform ERA5\n---\nIn this section, the `download.download_and_transform` function from the 'c3s_eqc_automatic_quality_control' package is used to retrieve ERA5 monthly reference data, compute annual means, calculate latitudinal band averages, and derive the climatology and linear trend at each pressure level for each latitudinal band. Results are cached to avoid redundant downloads and repeated processing.\n\n(climate_projections-cmip6_validation_q12:section-2.2)="} {"chunk_id": "climate_projections-cmip6_validation_q12__a83d511f81c1", "report_id": "climate_projections-cmip6_validation_q12", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q12", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands > Analysis and results > 2. Downloading and processing > 2.2. Download and transform models", "title": "Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands", "chunk_index": 11, "token_count": 483, "text_raw": "In this section, the `download.download_and_transform` function from the 'c3s_eqc_automatic_quality_control' package is employed to download monthly data from the CMIP6 models, compute annual means, calculate latitudinal band averages, and derive the climatology and linear trend at each pressure level for each latitudinal band. Results are cached to avoid redundant downloads and repeated processing.\n\n1000hPa are excluded (too close to surface)\n\n```text\nmodel='access_cm2'\nmodel='awi_cm_1_1_mr'\nmodel='canesm5'\nmodel='canesm5_canoe'\nmodel='ciesm'\nmodel='cmcc_cm2_sr5'\nmodel='cesm2'\nmodel='cmcc_esm2'\nmodel='cnrm_cm6_1'\nmodel='cnrm_cm6_1_hr'\nmodel='cnrm_esm2_1'\nmodel='ec_earth3_cc'\nmodel='ec_earth3_veg_lr'\nmodel='fgoals_g3'\nmodel='fgoals_f3_l'\nmodel='fio_esm_2_0'\nmodel='gfdl_esm4'\nmodel='hadgem3_gc31_ll'\nmodel='hadgem3_gc31_mm'\nmodel='iitm_esm'\nmodel='inm_cm4_8'\nmodel='inm_cm5_0'\nmodel='ipsl_cm6a_lr'\nmodel='kiost_esm'\nmodel='miroc6'\nmodel='miroc_es2l'\nmodel='mcm_ua_1_0'\nmodel='mpi_esm1_2_lr'\nmodel='mri_esm2_0'\nmodel='nesm3'\nmodel='noresm2_mm'\nmodel='taiesm1'\nmodel='ukesm1_0_ll'\n```\n\n(climate_projections-cmip6_validation_q12:section-3)=", "text_with_prefix": "EQC Quality Assessment: \"Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q12 | Category: Climate_Projections\nSection: Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands > Analysis and results > 2. Downloading and processing > 2.2. Download and transform models\n---\nIn this section, the `download.download_and_transform` function from the 'c3s_eqc_automatic_quality_control' package is employed to download monthly data from the CMIP6 models, compute annual means, calculate latitudinal band averages, and derive the climatology and linear trend at each pressure level for each latitudinal band. Results are cached to avoid redundant downloads and repeated processing.\n\n1000hPa are excluded (too close to surface)\n\n```text\nmodel='access_cm2'\nmodel='awi_cm_1_1_mr'\nmodel='canesm5'\nmodel='canesm5_canoe'\nmodel='ciesm'\nmodel='cmcc_cm2_sr5'\nmodel='cesm2'\nmodel='cmcc_esm2'\nmodel='cnrm_cm6_1'\nmodel='cnrm_cm6_1_hr'\nmodel='cnrm_esm2_1'\nmodel='ec_earth3_cc'\nmodel='ec_earth3_veg_lr'\nmodel='fgoals_g3'\nmodel='fgoals_f3_l'\nmodel='fio_esm_2_0'\nmodel='gfdl_esm4'\nmodel='hadgem3_gc31_ll'\nmodel='hadgem3_gc31_mm'\nmodel='iitm_esm'\nmodel='inm_cm4_8'\nmodel='inm_cm5_0'\nmodel='ipsl_cm6a_lr'\nmodel='kiost_esm'\nmodel='miroc6'\nmodel='miroc_es2l'\nmodel='mcm_ua_1_0'\nmodel='mpi_esm1_2_lr'\nmodel='mri_esm2_0'\nmodel='nesm3'\nmodel='noresm2_mm'\nmodel='taiesm1'\nmodel='ukesm1_0_ll'\n```\n\n(climate_projections-cmip6_validation_q12:section-3)="} {"chunk_id": "climate_projections-cmip6_validation_q12__2d959cc8d1df", "report_id": "climate_projections-cmip6_validation_q12", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q12", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands > Analysis and results > 3. Plot and describe results", "title": "Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands", "chunk_index": 12, "token_count": 236, "text_raw": "This section will display the following results:\n\n- **Vertical profiles of the climatology** for each latitudinal band, including ERA5, the median of the CMIP6 models, and the model spread.\n\n- **Vertical profiles of the mean bias** for each latitudinal band, showing the median of the CMIP6 model biases and their spread.\n\n- **Vertical profiles of the trend** for each latitudinal band, including ERA5, the median of the CMIP6 models, and the spread.\n\n- **Vertical profiles of the trend bias** for each latitudinal band, showing the median of the CMIP6 model biases and their spread.\n\n(climate_projections-cmip6_validation_q12:section-3.1)=", "text_with_prefix": "EQC Quality Assessment: \"Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q12 | Category: Climate_Projections\nSection: Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands > Analysis and results > 3. Plot and describe results\n---\nThis section will display the following results:\n\n- **Vertical profiles of the climatology** for each latitudinal band, including ERA5, the median of the CMIP6 models, and the model spread.\n\n- **Vertical profiles of the mean bias** for each latitudinal band, showing the median of the CMIP6 model biases and their spread.\n\n- **Vertical profiles of the trend** for each latitudinal band, including ERA5, the median of the CMIP6 models, and the spread.\n\n- **Vertical profiles of the trend bias** for each latitudinal band, showing the median of the CMIP6 model biases and their spread.\n\n(climate_projections-cmip6_validation_q12:section-3.1)="} {"chunk_id": "climate_projections-cmip6_validation_q12__16e9b0c462a8", "report_id": "climate_projections-cmip6_validation_q12", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q12", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands > Analysis and results > 3. Plot and describe results > 3.1. Define plotting functions", "title": "Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands", "chunk_index": 13, "token_count": 374, "text_raw": "The function **`plot_profile`** is used to display vertical profiles of the climatology and trend for each latitudinal band. When the `bias` option is set to False, ERA5 is shown as a green line, the CMIP6 ensemble median as a solid black line, individual CMIP6 models as grey dashed lines, and the model spread as a grey shaded area. When `bias=True`, the function instead plots the corresponding biases: the median CMIP6 bias is shown in blue, individual model biases in grey dashed lines, and their spread as a blue shaded area.\n\nConvert pressure to hPa if needed\nSort pressure descending\nSetup according to plot type\nExtract data\nCompute model biases\nZero line always for bias\nPlot individual model biases\nPlot bias median and spread\nZero line for trends only\nPlot individual CMIP6 models\nCMIP6 median + spread\n=== Significance markers only here ===\nFor pressures where at least 66% of models are significant:\nCount models with positive and negative trend signs (ALL models, regardless of significance)\nCalculate sign ratio as max count divided by total number of models\nFind pressures where sign agreement < 0.80\nX-axis limits and labels\nY-axis label only for first column\nCollect handles for shared legend\n\n(climate_projections-cmip6_validation_q12:section-3.2)=", "text_with_prefix": "EQC Quality Assessment: \"Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q12 | Category: Climate_Projections\nSection: Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands > Analysis and results > 3. Plot and describe results > 3.1. Define plotting functions\n---\nThe function **`plot_profile`** is used to display vertical profiles of the climatology and trend for each latitudinal band. When the `bias` option is set to False, ERA5 is shown as a green line, the CMIP6 ensemble median as a solid black line, individual CMIP6 models as grey dashed lines, and the model spread as a grey shaded area. When `bias=True`, the function instead plots the corresponding biases: the median CMIP6 bias is shown in blue, individual model biases in grey dashed lines, and their spread as a blue shaded area.\n\nConvert pressure to hPa if needed\nSort pressure descending\nSetup according to plot type\nExtract data\nCompute model biases\nZero line always for bias\nPlot individual model biases\nPlot bias median and spread\nZero line for trends only\nPlot individual CMIP6 models\nCMIP6 median + spread\n=== Significance markers only here ===\nFor pressures where at least 66% of models are significant:\nCount models with positive and negative trend signs (ALL models, regardless of significance)\nCalculate sign ratio as max count divided by total number of models\nFind pressures where sign agreement < 0.80\nX-axis limits and labels\nY-axis label only for first column\nCollect handles for shared legend\n\n(climate_projections-cmip6_validation_q12:section-3.2)="} {"chunk_id": "climate_projections-cmip6_validation_q12__e50322a3af81", "report_id": "climate_projections-cmip6_validation_q12", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q12", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands > Analysis and results > 3. Plot and describe results > 3.2. Plot climatology maps", "title": "Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands", "chunk_index": 14, "token_count": 272, "text_raw": "In this section, we invoke the `plot_profile()` function to display the vertical profiles of the climatology for each latitudinal band, including ERA5, the median of the CMIP6 models, and the model spread.\n\n
\n
\n

Fig 1. \n Vertical profiles of the annual mean temperature climatology (1979-2014). Each subplot shows a latitudinal band (Global, Northern and Southern hemispheres, Tropics, Extratropics, and Polar). ERA5 is shown as a green line, the CMIP6 median as a solid black line, individual CMIP6 models as grey dashed lines, and the spread of models as a grey shaded area.\n\n

\n\n(climate_projections-cmip6_validation_q12:section-3.3)=", "text_with_prefix": "EQC Quality Assessment: \"Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q12 | Category: Climate_Projections\nSection: Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands > Analysis and results > 3. Plot and describe results > 3.2. Plot climatology maps\n---\nIn this section, we invoke the `plot_profile()` function to display the vertical profiles of the climatology for each latitudinal band, including ERA5, the median of the CMIP6 models, and the model spread.\n\n
\n
\n

Fig 1. \n Vertical profiles of the annual mean temperature climatology (1979-2014). Each subplot shows a latitudinal band (Global, Northern and Southern hemispheres, Tropics, Extratropics, and Polar). ERA5 is shown as a green line, the CMIP6 median as a solid black line, individual CMIP6 models as grey dashed lines, and the spread of models as a grey shaded area.\n\n

\n\n(climate_projections-cmip6_validation_q12:section-3.3)="} {"chunk_id": "climate_projections-cmip6_validation_q12__998d1eab0906", "report_id": "climate_projections-cmip6_validation_q12", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q12", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands > Analysis and results > 3. Plot and describe results > 3.3. Plot mean bias maps", "title": "Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands", "chunk_index": 15, "token_count": 263, "text_raw": "In this section, we invoke the `plot_profile()` function to display the vertical profiles of the mean bias for each latitudinal band, showing the median of the CMIP6 model biases and their spread.\n\n
\n
\n

Fig 2. \n Vertical profiles of the mean bias of the annual mean temperature (1979-2014). Each subplot shows a latitudinal band (Global, Northern and Southern hemispheres, Tropics, Extratropics, and Polar). The CMIP6 median is shown as a solid blue line, individual CMIP6 models as grey dashed lines, and the spread of models as a blue shaded area.\n\n

\n\n(climate_projections-cmip6_validation_q12:section-3.4)=", "text_with_prefix": "EQC Quality Assessment: \"Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q12 | Category: Climate_Projections\nSection: Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands > Analysis and results > 3. Plot and describe results > 3.3. Plot mean bias maps\n---\nIn this section, we invoke the `plot_profile()` function to display the vertical profiles of the mean bias for each latitudinal band, showing the median of the CMIP6 model biases and their spread.\n\n
\n
\n

Fig 2. \n Vertical profiles of the mean bias of the annual mean temperature (1979-2014). Each subplot shows a latitudinal band (Global, Northern and Southern hemispheres, Tropics, Extratropics, and Polar). The CMIP6 median is shown as a solid blue line, individual CMIP6 models as grey dashed lines, and the spread of models as a blue shaded area.\n\n

\n\n(climate_projections-cmip6_validation_q12:section-3.4)="} {"chunk_id": "climate_projections-cmip6_validation_q12__a0cb4d997ef5", "report_id": "climate_projections-cmip6_validation_q12", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q12", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands > Analysis and results > 3. Plot and describe results > 3.4. Plot trend maps", "title": "Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands", "chunk_index": 16, "token_count": 439, "text_raw": "In this section, we invoke the `plot_profile()` function to display the vertical profiles of the trend for each latitudinal band, including ERA5, the median of the CMIP6 models, and the spread\n\n
\n
\n

Fig 3. \n Vertical profiles of the annual mean temperature trend (1979-2014). Each subplot corresponds to a latitudinal band (Global, Northern Hemisphere, Southern Hemisphere, Tropics, Extratropics, and Polar). ERA5 is shown as a green line, the CMIP6 ensemble median as a solid black line, individual CMIP6 models as grey dashed lines, and the ensemble spread as a grey shaded area. ERA5 points with non-significant trends (p-value > 0.05) are marked with green crosses. For the CMIP6 ensemble median, trend significance is assessed using agreement categories based on the approach described in the IPCC AR6 Atlas (pp. 1945–1950). Black crosses indicate pressure levels where fewer than 66% of models show statistically significant trends (p-value < 0.05). If over 66% of the models are significant but fewer than 80% agree on the sign of the trend, a black dot is placed on the CMIP6 ensemble median to indicate low directional agreement.\n

\n
\n
\n\n(climate_projections-cmip6_validation_q12:section-3.5)=", "text_with_prefix": "EQC Quality Assessment: \"Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q12 | Category: Climate_Projections\nSection: Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands > Analysis and results > 3. Plot and describe results > 3.4. Plot trend maps\n---\nIn this section, we invoke the `plot_profile()` function to display the vertical profiles of the trend for each latitudinal band, including ERA5, the median of the CMIP6 models, and the spread\n\n
\n
\n

Fig 3. \n Vertical profiles of the annual mean temperature trend (1979-2014). Each subplot corresponds to a latitudinal band (Global, Northern Hemisphere, Southern Hemisphere, Tropics, Extratropics, and Polar). ERA5 is shown as a green line, the CMIP6 ensemble median as a solid black line, individual CMIP6 models as grey dashed lines, and the ensemble spread as a grey shaded area. ERA5 points with non-significant trends (p-value > 0.05) are marked with green crosses. For the CMIP6 ensemble median, trend significance is assessed using agreement categories based on the approach described in the IPCC AR6 Atlas (pp. 1945–1950). Black crosses indicate pressure levels where fewer than 66% of models show statistically significant trends (p-value < 0.05). If over 66% of the models are significant but fewer than 80% agree on the sign of the trend, a black dot is placed on the CMIP6 ensemble median to indicate low directional agreement.\n

\n
\n
\n\n(climate_projections-cmip6_validation_q12:section-3.5)="} {"chunk_id": "climate_projections-cmip6_validation_q12__8fed67deb754", "report_id": "climate_projections-cmip6_validation_q12", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q12", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands > Analysis and results > 3. Plot and describe results > 3.5. Plot trend bias maps", "title": "Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands", "chunk_index": 17, "token_count": 263, "text_raw": "In this section, we invoke the `plot_profile()` function to display the vertical profiles of the trend bias for each latitudinal band, showing the median of the CMIP6 model biases and their spread.\n\n
\n
\n

Fig 4. \n Vertical profiles of the trend bias of the annual mean temperature (1979-2014). Each subplot shows a latitudinal band (Global, Northern and Southern hemispheres, Tropics, Extratropics, and Polar). The CMIP6 median is shown as a solid blue line, individual CMIP6 models as grey dashed lines, and the spread of models as a blue shaded area.\n\n

\n\n(climate_projections-cmip6_validation_q12:section-3.6)=", "text_with_prefix": "EQC Quality Assessment: \"Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q12 | Category: Climate_Projections\nSection: Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands > Analysis and results > 3. Plot and describe results > 3.5. Plot trend bias maps\n---\nIn this section, we invoke the `plot_profile()` function to display the vertical profiles of the trend bias for each latitudinal band, showing the median of the CMIP6 model biases and their spread.\n\n
\n
\n

Fig 4. \n Vertical profiles of the trend bias of the annual mean temperature (1979-2014). Each subplot shows a latitudinal band (Global, Northern and Southern hemispheres, Tropics, Extratropics, and Polar). The CMIP6 median is shown as a solid blue line, individual CMIP6 models as grey dashed lines, and the spread of models as a blue shaded area.\n\n

\n\n(climate_projections-cmip6_validation_q12:section-3.6)="} {"chunk_id": "climate_projections-cmip6_validation_q12__dc99c77125b3", "report_id": "climate_projections-cmip6_validation_q12", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q12", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands > Analysis and results > 3. Plot and describe results > 3.6. Results summary and discussion - Climatologies", "title": "Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands", "chunk_index": 18, "token_count": 281, "text_raw": "- Vertical temperature profiles are broadly consistent across latitudinal bands, with the tropopause—defined as the level where the temperature inversion begins—occurring at lower pressures in the tropics and higher pressures in colder regions (e.g. polar and extratropical).\n\n- Biases are predominantly cold, with the median of the model biases showing the largest discrepancies around 250–200 hPa, where CMIP6 models exhibit a sharper temperature transition than ERA5. Near-surface and stratospheric median biases are generally smaller, except in polar regions.\n\n- Model spread increases with altitude, particularly in the stratosphere, highlighting reduced inter-model agreement at higher levels.\n\n- The large spread near the tropopause underscores the substantial uncertainty in simulating its height and structure, which may have important implications for model evaluation and development.\n\n(climate_projections-cmip6_validation_q12:section-3.7)=", "text_with_prefix": "EQC Quality Assessment: \"Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q12 | Category: Climate_Projections\nSection: Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands > Analysis and results > 3. Plot and describe results > 3.6. Results summary and discussion - Climatologies\n---\n- Vertical temperature profiles are broadly consistent across latitudinal bands, with the tropopause—defined as the level where the temperature inversion begins—occurring at lower pressures in the tropics and higher pressures in colder regions (e.g. polar and extratropical).\n\n- Biases are predominantly cold, with the median of the model biases showing the largest discrepancies around 250–200 hPa, where CMIP6 models exhibit a sharper temperature transition than ERA5. Near-surface and stratospheric median biases are generally smaller, except in polar regions.\n\n- Model spread increases with altitude, particularly in the stratosphere, highlighting reduced inter-model agreement at higher levels.\n\n- The large spread near the tropopause underscores the substantial uncertainty in simulating its height and structure, which may have important implications for model evaluation and development.\n\n(climate_projections-cmip6_validation_q12:section-3.7)="} {"chunk_id": "climate_projections-cmip6_validation_q12__1aaa6226e370", "report_id": "climate_projections-cmip6_validation_q12", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q12", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands > Analysis and results > 3. Plot and describe results > 3.7. Results summary and discussion - Trends", "title": "Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands", "chunk_index": 19, "token_count": 541, "text_raw": "* Tropospheric warming and stratospheric cooling are robust fingerprints of greenhouse-gas-induced climate change [[3]](https://doi.org/10.1073/pnas.1305332110). ERA5 and CMIP6 both reproduce this vertical structure across all latitudinal bands, although the magnitude and detailed profile are sensitive to dataset processing and homogenisation, leading to medium to low confidence in long-term trend estimates [[6]](https://www.ipcc.ch/site/assets/uploads/2017/09/WG1AR5_Chapter02_FINAL.pdf).\n\n* Since vertical temperature trend profiles are important fingerprints of climate change, it is important to improve their representation, as this would enhance confidence in these signals and benefit researchers, impact assessors, and decision-makers.\n\n* Tropical amplification—enhanced warming in the upper tropical troposphere relative to the surface—is evident in both ERA5 and CMIP6, although CMIP6 tends to show a stronger signal.\n\n* Amplification varies across latitudinal bands. CMIP6 models display amplification in all regions except the polar zones, whereas ERA5 shows amplification everywhere except in the polar and Northern Hemisphere bands, where warming is stronger near the surface. However, in the global and extratropical bands, both CMIP6 and ERA5 exhibit minimal amplification.\n\n* Differences between CMIP6 and ERA5 can be linked to biases in surface warming [[9]](https://doi.org/10.1002/grl.50465). Studies using models with prescribed SSTs (e.g. [[10]](https://doi.org/10.1088/1748-9326/ab9af7)) report reduced biases. In addition, upper-troposphere trend estimates can vary depending on the analysis period, due to factors such as ozone changes.\n\n* Given the uncertainties associated with individual datasets, using multiple reference products—including those external to the CDS—would be advisable for this type of assessment. For example, [[8]](https://doi.org/10.1175/JCLI-D-19-0998.1) compared tropospheric and stratospheric trends using both ground- and satellite-based observations.", "text_with_prefix": "EQC Quality Assessment: \"Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q12 | Category: Climate_Projections\nSection: Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands > Analysis and results > 3. Plot and describe results > 3.7. Results summary and discussion - Trends\n---\n* Tropospheric warming and stratospheric cooling are robust fingerprints of greenhouse-gas-induced climate change [[3]](https://doi.org/10.1073/pnas.1305332110). ERA5 and CMIP6 both reproduce this vertical structure across all latitudinal bands, although the magnitude and detailed profile are sensitive to dataset processing and homogenisation, leading to medium to low confidence in long-term trend estimates [[6]](https://www.ipcc.ch/site/assets/uploads/2017/09/WG1AR5_Chapter02_FINAL.pdf).\n\n* Since vertical temperature trend profiles are important fingerprints of climate change, it is important to improve their representation, as this would enhance confidence in these signals and benefit researchers, impact assessors, and decision-makers.\n\n* Tropical amplification—enhanced warming in the upper tropical troposphere relative to the surface—is evident in both ERA5 and CMIP6, although CMIP6 tends to show a stronger signal.\n\n* Amplification varies across latitudinal bands. CMIP6 models display amplification in all regions except the polar zones, whereas ERA5 shows amplification everywhere except in the polar and Northern Hemisphere bands, where warming is stronger near the surface. However, in the global and extratropical bands, both CMIP6 and ERA5 exhibit minimal amplification.\n\n* Differences between CMIP6 and ERA5 can be linked to biases in surface warming [[9]](https://doi.org/10.1002/grl.50465). Studies using models with prescribed SSTs (e.g. [[10]](https://doi.org/10.1088/1748-9326/ab9af7)) report reduced biases. In addition, upper-troposphere trend estimates can vary depending on the analysis period, due to factors such as ozone changes.\n\n* Given the uncertainties associated with individual datasets, using multiple reference products—including those external to the CDS—would be advisable for this type of assessment. For example, [[8]](https://doi.org/10.1175/JCLI-D-19-0998.1) compared tropospheric and stratospheric trends using both ground- and satellite-based observations."} {"chunk_id": "climate_projections-cmip6_validation_q12__d0612c18dbdf", "report_id": "climate_projections-cmip6_validation_q12", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q12", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands > ℹ️ If you want to know more > Key resources", "title": "Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands", "chunk_index": 20, "token_count": 305, "text_raw": "Some key resources and further reading were linked throughout this assessment.\n\nThe CDS catalogue entries for the data used were:\n* CMIP6 climate projections (Monthly - Air temperature): https://cds.climate.copernicus.eu/datasets/projections-cmip6?tab=overview\n* ERA5 monthly averaged data on pressure levels from 1940 to present (Temperature): https://cds.climate.copernicus.eu/datasets/reanalysis-era5-pressure-levels-monthly-means?tab=overview\n* ERA5 monthly averaged data on single levels from 1940 to present (Surface pressure): https://cds.climate.copernicus.eu/datasets/reanalysis-era5-single-levels-monthly-means?tab=overview\n\nCode libraries used:\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)", "text_with_prefix": "EQC Quality Assessment: \"Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q12 | Category: Climate_Projections\nSection: Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands > ℹ️ If you want to know more > Key resources\n---\nSome key resources and further reading were linked throughout this assessment.\n\nThe CDS catalogue entries for the data used were:\n* CMIP6 climate projections (Monthly - Air temperature): https://cds.climate.copernicus.eu/datasets/projections-cmip6?tab=overview\n* ERA5 monthly averaged data on pressure levels from 1940 to present (Temperature): https://cds.climate.copernicus.eu/datasets/reanalysis-era5-pressure-levels-monthly-means?tab=overview\n* ERA5 monthly averaged data on single levels from 1940 to present (Surface pressure): https://cds.climate.copernicus.eu/datasets/reanalysis-era5-single-levels-monthly-means?tab=overview\n\nCode libraries used:\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)"} {"chunk_id": "climate_projections-cmip6_validation_q12__ac249ba2961d", "report_id": "climate_projections-cmip6_validation_q12", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q12", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands > ℹ️ If you want to know more > References", "title": "Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands", "chunk_index": 21, "token_count": 1069, "text_raw": "[[1]](https://doi.org/10.1002/joc.1756) Santer, B.D., Thorne, P.W., Haimberger, L., Taylor, K.E., Wigley, T.M.L., Lanzante, J.R., Solomon, S., Free, M., Gleckler, P.J., Jones, P.D., Karl, T.R., Klein, S.A., Mears, C., Nychka, D., Schmidt, G.A., Sherwood, S.C. and Wentz, F.J., 2008. Consistency of modelled and observed temperature trends in the tropical troposphere. Int. J. Climatol., 28: 1703-1722. https://doi.org/10.1002/joc.1756\n\n[[2]](https://doi.org/10.1126/science.274.5290.1170) Simon F. B. Tett et al., 1996. Human Influence on the Atmospheric Vertical Temperature Structure: Detection and Observations.Science274,1170-1173. https://doi.org/10.1126/science.274.5290.1170\n\n[[3]](https://doi.org/10.1073/pnas.1305332110) Santer, B.D., Painter, J.F., Bonfils, C., Mears, C.A., Solomon, S., Wigley, T.M.L., Gleckler,P.J., Schmidt, G.A., Doutriaux,C., Gillett, N.P., Taylor,K.E., Thorne, P.W., and Wentz, F.J., 2013. Human and natural influences on the changing thermal structure of the atmosphere, Proc. Natl. Acad. Sci. U.S.A. 110 (43) 17235-17240. https://doi.org/10.1073/pnas.1305332110\n\n[[4]](https://doi.org/10.1126/science.1114867) B. D. Santer et al., 2005. Amplification of Surface Temperature Trends and Variability in the Tropical Atmosphere.Science309,1551-1556. https://doi.org/10.1126/science.1114867\n\n[[5]](https://doi.org/10.1175/1520-0469%281979%29036<0415:ALRRAT>2.0.CO;2) Stone, P. H., and Carlson, J. H., 1979. Atmospheric lapse rate regimes and their parameterization. J. Atmos. Sci., 36, 415–423. https://doi.org/10.1175/1520-0469%281979%29036<0415:ALRRAT>2.0.CO;2 .\n\n[[6]](https://www.ipcc.ch/site/assets/uploads/2017/09/WG1AR5_Chapter02_FINAL.pdf) Hartmann, D. L., et al., 2013. Observations: Atmosphere and surface. Climate Change 2013: The Physical Science Basis, T. F. Stocker et al., Eds., Cambridge University Press, 159–254.\n\n[[7]](https://doi.org/10.1029/2021JD035220) Madonna, F., Tramutola, E., SY, S., Serva, F., Proto, M., Rosoldi, M., et al., 2022. The new Radiosounding HARMonization (RHARM) data set of homogenized radiosounding temperature, humidity, and wind profiles with uncertainties. Journal of Geophysical Research: Atmospheres, 127, e2021JD035220. https://doi.org/10.1029/2021JD035220\n\n[[8]](https://doi.org/10.1175/JCLI-D-19-0998.1) Steiner, A. K. et al., 2020. Observed Temperature Changes in the Troposphere and Stratosphere from 1979 to 2018. J. Climate, 33, 8165–8194, https://doi.org/10.1175/JCLI-D-19-0998.1\n\n[[9]](https://doi.org/10.1002/grl.50465) Mitchell, D. M., Thorne, P. W., Stott, P. A., and Gray, L. J., 2013. Revisiting the controversial issue of tropical tropospheric temperature trends, Geophys. Res. Lett., 40, 2801–2806, https://doi.org/10.1002/grl.50465", "text_with_prefix": "EQC Quality Assessment: \"Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q12 | Category: Climate_Projections\nSection: Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands > ℹ️ If you want to know more > References\n---\n[[1]](https://doi.org/10.1002/joc.1756) Santer, B.D., Thorne, P.W., Haimberger, L., Taylor, K.E., Wigley, T.M.L., Lanzante, J.R., Solomon, S., Free, M., Gleckler, P.J., Jones, P.D., Karl, T.R., Klein, S.A., Mears, C., Nychka, D., Schmidt, G.A., Sherwood, S.C. and Wentz, F.J., 2008. Consistency of modelled and observed temperature trends in the tropical troposphere. Int. J. Climatol., 28: 1703-1722. https://doi.org/10.1002/joc.1756\n\n[[2]](https://doi.org/10.1126/science.274.5290.1170) Simon F. B. Tett et al., 1996. Human Influence on the Atmospheric Vertical Temperature Structure: Detection and Observations.Science274,1170-1173. https://doi.org/10.1126/science.274.5290.1170\n\n[[3]](https://doi.org/10.1073/pnas.1305332110) Santer, B.D., Painter, J.F., Bonfils, C., Mears, C.A., Solomon, S., Wigley, T.M.L., Gleckler,P.J., Schmidt, G.A., Doutriaux,C., Gillett, N.P., Taylor,K.E., Thorne, P.W., and Wentz, F.J., 2013. Human and natural influences on the changing thermal structure of the atmosphere, Proc. Natl. Acad. Sci. U.S.A. 110 (43) 17235-17240. https://doi.org/10.1073/pnas.1305332110\n\n[[4]](https://doi.org/10.1126/science.1114867) B. D. Santer et al., 2005. Amplification of Surface Temperature Trends and Variability in the Tropical Atmosphere.Science309,1551-1556. https://doi.org/10.1126/science.1114867\n\n[[5]](https://doi.org/10.1175/1520-0469%281979%29036<0415:ALRRAT>2.0.CO;2) Stone, P. H., and Carlson, J. H., 1979. Atmospheric lapse rate regimes and their parameterization. J. Atmos. Sci., 36, 415–423. https://doi.org/10.1175/1520-0469%281979%29036<0415:ALRRAT>2.0.CO;2 .\n\n[[6]](https://www.ipcc.ch/site/assets/uploads/2017/09/WG1AR5_Chapter02_FINAL.pdf) Hartmann, D. L., et al., 2013. Observations: Atmosphere and surface. Climate Change 2013: The Physical Science Basis, T. F. Stocker et al., Eds., Cambridge University Press, 159–254.\n\n[[7]](https://doi.org/10.1029/2021JD035220) Madonna, F., Tramutola, E., SY, S., Serva, F., Proto, M., Rosoldi, M., et al., 2022. The new Radiosounding HARMonization (RHARM) data set of homogenized radiosounding temperature, humidity, and wind profiles with uncertainties. Journal of Geophysical Research: Atmospheres, 127, e2021JD035220. https://doi.org/10.1029/2021JD035220\n\n[[8]](https://doi.org/10.1175/JCLI-D-19-0998.1) Steiner, A. K. et al., 2020. Observed Temperature Changes in the Troposphere and Stratosphere from 1979 to 2018. J. Climate, 33, 8165–8194, https://doi.org/10.1175/JCLI-D-19-0998.1\n\n[[9]](https://doi.org/10.1002/grl.50465) Mitchell, D. M., Thorne, P. W., Stott, P. A., and Gray, L. J., 2013. Revisiting the controversial issue of tropical tropospheric temperature trends, Geophys. Res. Lett., 40, 2801–2806, https://doi.org/10.1002/grl.50465"} {"chunk_id": "climate_projections-cmip6_validation_q12__8cba95d8d3a1", "report_id": "climate_projections-cmip6_validation_q12", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q12", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands > ℹ️ If you want to know more > References", "title": "Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands", "chunk_index": 22, "token_count": 225, "text_raw": "2013. Revisiting the controversial issue of tropical tropospheric temperature trends, Geophys. Res. Lett., 40, 2801–2806, https://doi.org/10.1002/grl.50465\n\n[[10]](https://doi.org/10.1088/1748-9326/ab9af7) Dann M. Mitchell et al., 2020. The vertical profile of recent tropical temperature trends: Persistent model biases in the context of internal variability. Environ. Res. Lett. 15 1040b4. https://doi.org/10.1088/1748-9326/ab9af7", "text_with_prefix": "EQC Quality Assessment: \"Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q12 | Category: Climate_Projections\nSection: Biases in temperature trends at different pressure levels using CMIP6 across latitudinal bands > ℹ️ If you want to know more > References\n---\n2013. Revisiting the controversial issue of tropical tropospheric temperature trends, Geophys. Res. Lett., 40, 2801–2806, https://doi.org/10.1002/grl.50465\n\n[[10]](https://doi.org/10.1088/1748-9326/ab9af7) Dann M. Mitchell et al., 2020. The vertical profile of recent tropical temperature trends: Persistent model biases in the context of internal variability. Environ. Res. Lett. 15 1040b4. https://doi.org/10.1088/1748-9326/ab9af7"} {"chunk_id": "climate_projections-cmip6_validation_q13__1030c8f978d3", "report_id": "climate_projections-cmip6_validation_q13", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q13", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CMIP6 biases in the SPEI6 drought index over the Mediterranean region", "title": "CMIP6 biases in the SPEI6 drought index over the Mediterranean region", "chunk_index": 0, "token_count": 106, "text_raw": "Production date: 20-09-2025\n\nProduced by: CMCC foundation - Euro-Mediterranean Center on Climate Change. Albert Martinez Boti and Lorenzo Sangelantoni.", "text_with_prefix": "EQC Quality Assessment: \"CMIP6 biases in the SPEI6 drought index over the Mediterranean region\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q13 | Category: Climate_Projections\nSection: CMIP6 biases in the SPEI6 drought index over the Mediterranean region\n---\nProduction date: 20-09-2025\n\nProduced by: CMCC foundation - Euro-Mediterranean Center on Climate Change. Albert Martinez Boti and Lorenzo Sangelantoni."} {"chunk_id": "climate_projections-cmip6_validation_q13__66d35b175c1d", "report_id": "climate_projections-cmip6_validation_q13", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q13", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CMIP6 biases in the SPEI6 drought index over the Mediterranean region > Quality assessment question", "title": "CMIP6 biases in the SPEI6 drought index over the Mediterranean region", "chunk_index": 1, "token_count": 563, "text_raw": "* **How well do CMIP6 models simulate drought compared to signals derived from ERA5?**\n\nThe Standardized Precipitation-Evapotranspiration Index (SPEI) [[1]](https://doi.org/10.1175/2009JCLI2909.1) is a widely used drought indicator that integrates precipitation and potential evapotranspiration, providing a more complete measure of water deficits than precipitation alone. Its multiscalar formulation allows the evaluation of different types of drought [[2]](https://digitalcommons.unl.edu/droughtfacpub/69/), from short-term agricultural or soil moisture deficits to long-term hydrological or ecological events. Being standardized, SPEI is dimensionless, facilitating comparisons across regions and climates. By capturing variations in frequency, duration, and severity, it effectively reflects the complex nature of drought, making it a robust and physically meaningful proxy for drought assessment in the context of climate change [[3]](https://doi.org/10.1175/2012EI000434.1).\n\nUnderstanding SPEI is particularly relevant for the water management sector. By capturing multiple types of drought SPEI serves as a versatile indicator for various users. Climate projections using SPEI enable planners to anticipate future water deficits, guide allocation, and design effective strategies for reservoir management, irrigation, and drought preparedness, helping to reduce socio-economic and ecological impacts [[4]](https://doi.org/10.1016/j.scitotenv.2019.135245)[[5]](https://doi.org/10.3390/w17101507).\n\nThis assessment evaluates SPEI6 (six-month accumulation) across the full ensemble of [CMIP6](https://cds.climate.copernicus.eu/datasets/projections-cmip6?tab=overview) models available through the Copernicus Climate Data Store (CDS). Model outputs are compared against SPEI6 derived from [ERA5](https://cds.climate.copernicus.eu/datasets/reanalysis-era5-complete?tab=doc) reanalysis, which serves as the reference product. The standard reference period is 1971–2010. We specifically evaluate whether changes in the 1991–2020 period relative to the 1961–1990 baseline are accurately represented by CMIP6 models and the trend acroos the whole 1961-2020 period for the Mediterranean domain.", "text_with_prefix": "EQC Quality Assessment: \"CMIP6 biases in the SPEI6 drought index over the Mediterranean region\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q13 | Category: Climate_Projections\nSection: CMIP6 biases in the SPEI6 drought index over the Mediterranean region > Quality assessment question\n---\n* **How well do CMIP6 models simulate drought compared to signals derived from ERA5?**\n\nThe Standardized Precipitation-Evapotranspiration Index (SPEI) [[1]](https://doi.org/10.1175/2009JCLI2909.1) is a widely used drought indicator that integrates precipitation and potential evapotranspiration, providing a more complete measure of water deficits than precipitation alone. Its multiscalar formulation allows the evaluation of different types of drought [[2]](https://digitalcommons.unl.edu/droughtfacpub/69/), from short-term agricultural or soil moisture deficits to long-term hydrological or ecological events. Being standardized, SPEI is dimensionless, facilitating comparisons across regions and climates. By capturing variations in frequency, duration, and severity, it effectively reflects the complex nature of drought, making it a robust and physically meaningful proxy for drought assessment in the context of climate change [[3]](https://doi.org/10.1175/2012EI000434.1).\n\nUnderstanding SPEI is particularly relevant for the water management sector. By capturing multiple types of drought SPEI serves as a versatile indicator for various users. Climate projections using SPEI enable planners to anticipate future water deficits, guide allocation, and design effective strategies for reservoir management, irrigation, and drought preparedness, helping to reduce socio-economic and ecological impacts [[4]](https://doi.org/10.1016/j.scitotenv.2019.135245)[[5]](https://doi.org/10.3390/w17101507).\n\nThis assessment evaluates SPEI6 (six-month accumulation) across the full ensemble of [CMIP6](https://cds.climate.copernicus.eu/datasets/projections-cmip6?tab=overview) models available through the Copernicus Climate Data Store (CDS). Model outputs are compared against SPEI6 derived from [ERA5](https://cds.climate.copernicus.eu/datasets/reanalysis-era5-complete?tab=doc) reanalysis, which serves as the reference product. The standard reference period is 1971–2010. We specifically evaluate whether changes in the 1991–2020 period relative to the 1961–1990 baseline are accurately represented by CMIP6 models and the trend acroos the whole 1961-2020 period for the Mediterranean domain."} {"chunk_id": "climate_projections-cmip6_validation_q13__fd0a7f60ee89", "report_id": "climate_projections-cmip6_validation_q13", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q13", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CMIP6 biases in the SPEI6 drought index over the Mediterranean region > Quality assessment statement", "title": "CMIP6 biases in the SPEI6 drought index over the Mediterranean region", "chunk_index": 2, "token_count": 431, "text_raw": "These are the key outcomes of this assessment\n\n* CMIP6 ensemble median captures the general intensification of drought over the Mediterranean seen in ERA5, though the magnitude of the change is underestimated, particularly over eastern Spain, southern France, and much of Turkey.\n\n* Individual models do not reproduce the same spatial pattern, highlighting the importance of using large ensembles rather than relying on a single model for planning purposes.\n\n* Despite inherent biases, the considered subset of CMIP6 models provides a foundation for understanding the evolution of past drought conditions. These results support the use of SPEI to anticipate water deficits and guide drought preparedness, increasing confidence (though not ensuring accuracy) when assessing future trends.\n\n* Regions with high inter-model spread, such as North Africa, Turkey, and the eastern Mediterranean, may be associated with future higher uncertainty, suggesting that adaptation strategies in these areas should consider a wider range of possible futures.\n\n```\n\nattachment:7b5c6a55-25e9-4938-99e9-a977bad344e6.png\n---\nwidth: 1500px\nalt: Visual abstract \n---\nSpatial patterns of SPEI6 change over the Mediterranean region. Panels show (a) ERA5 reference, (b) CMIP6 ensemble median (median of the selected subset of models at each grid cell), (c) ensemble median bias relative to ERA5, and (d) ensemble spread (standard deviation across the selected models). SPEI6 is standardised over the reference period 1971–2010. SPEI6 change is calculated at each grid point as the difference between the monthly climatologies of the target period (1991–2020) and the baseline period (1961–1990), and then averaged to obtain the annual mean.\n```", "text_with_prefix": "EQC Quality Assessment: \"CMIP6 biases in the SPEI6 drought index over the Mediterranean region\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q13 | Category: Climate_Projections\nSection: CMIP6 biases in the SPEI6 drought index over the Mediterranean region > Quality assessment statement\n---\nThese are the key outcomes of this assessment\n\n* CMIP6 ensemble median captures the general intensification of drought over the Mediterranean seen in ERA5, though the magnitude of the change is underestimated, particularly over eastern Spain, southern France, and much of Turkey.\n\n* Individual models do not reproduce the same spatial pattern, highlighting the importance of using large ensembles rather than relying on a single model for planning purposes.\n\n* Despite inherent biases, the considered subset of CMIP6 models provides a foundation for understanding the evolution of past drought conditions. These results support the use of SPEI to anticipate water deficits and guide drought preparedness, increasing confidence (though not ensuring accuracy) when assessing future trends.\n\n* Regions with high inter-model spread, such as North Africa, Turkey, and the eastern Mediterranean, may be associated with future higher uncertainty, suggesting that adaptation strategies in these areas should consider a wider range of possible futures.\n\n```\n\nattachment:7b5c6a55-25e9-4938-99e9-a977bad344e6.png\n---\nwidth: 1500px\nalt: Visual abstract \n---\nSpatial patterns of SPEI6 change over the Mediterranean region. Panels show (a) ERA5 reference, (b) CMIP6 ensemble median (median of the selected subset of models at each grid cell), (c) ensemble median bias relative to ERA5, and (d) ensemble spread (standard deviation across the selected models). SPEI6 is standardised over the reference period 1971–2010. SPEI6 change is calculated at each grid point as the difference between the monthly climatologies of the target period (1991–2020) and the baseline period (1961–1990), and then averaged to obtain the annual mean.\n```"} {"chunk_id": "climate_projections-cmip6_validation_q13__8a26acdcc47c", "report_id": "climate_projections-cmip6_validation_q13", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q13", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CMIP6 biases in the SPEI6 drought index over the Mediterranean region > Methodology", "title": "CMIP6 biases in the SPEI6 drought index over the Mediterranean region", "chunk_index": 3, "token_count": 1057, "text_raw": "The Standardized Precipitation-Evapotranspiration Index accumulated to six months (SPEI6) has been calculated for this assessment — through the use of [xclim](https://xclim.readthedocs.io/en/stable/) — as a proxy to evaluate drought conditions for a subset of [CMIP6](https://cds.climate.copernicus.eu/datasets/projections-cmip6?tab=overview) models, which are evaluated against SPEI6 derived from [ERA5](https://cds.climate.copernicus.eu/datasets/reanalysis-era5-complete?tab=doc) reanalysis. The study covers the Mediterranean domain, following IPCC-AR6 regional definitions [[6]](https://doi.org/10.5194/essd-12-2959-2020), and results include spatial maps of SPEI6 changes for ERA5, CMIP6, and the biases of these changes. Additionally, time series showing the temporal evolution of drought across the domain are calculated, comparing trends from the models with ERA5 over the historical period considered within this assessment (1961–2020).\n\nPotential evapotranspiration (PET) is calculated with the Hargreaves method [[7]](https://doi.org/10.13031/2013.26773), which requires only maximum and minimum temperatures (and optionally mean temperature), providing a computationally efficient yet robust estimate (e.g., [[8]](https://doi.org/10.1002/joc.3887),). The accumulated series of precipitation minus PET is standardized for the reference period defined by the [C3S Atlas](https://confluence.ecmwf.int/display/CKB/Gridded+data+underpinning+the+Copernicus+Interactive+Climate+Atlas%3A+Description+of+the+datasets+and+variables#GriddeddataunderpinningtheCopernicusInteractiveClimateAtlas:Descriptionofthedatasetsandvariables) (1971–2010).\n\nFive key concepts are defined:\n\n- **Reference period:** The period used to standardize SPEI values, here 1971–2010, following the [C3S Atlas](https://confluence.ecmwf.int/display/CKB/Gridded+data+underpinning+the+Copernicus+Interactive+Climate+Atlas%3A+Description+of+the+datasets+and+variables#GriddeddataunderpinningtheCopernicusInteractiveClimateAtlas:Descriptionofthedatasetsandvariables). \n- **Baseline period:** The period against which changes are calculated, here 1961–1990. \n- **Target period:** The period for which changes are evaluated, here 1991–2020. \n- **Historical period:** The entire period considered in this assessment, here 1961–2020. \n- **Change assessment:** Differences in SPEI6 between the target period and the baseline, representing anomalies in drought conditions.\n\nThe analysis and results follow the next outline:\n\n**[](climate_projections-cmip6_validation_q13:section-1)**\n * [](climate_projections-cmip6_validation_q13:section-1.1)\n * [](climate_projections-cmip6_validation_q13:section-1.2)\n * [](climate_projections-cmip6_validation_q13:section-1.3)\n * [](climate_projections-cmip6_validation_q13:section-1.4)\n * [](climate_projections-cmip6_validation_q13:section-1.5)\n * [](climate_projections-cmip6_validation_q13:section-1.6)\n\n**[](climate_projections-cmip6_validation_q13:section-2)**\n * [](climate_projections-cmip6_validation_q13:section-2.1)\n * [](climate_projections-cmip6_validation_q13:section-2.2)\n * [](climate_projections-cmip6_validation_q13:section-2.3)\n\n**[](climate_projections-cmip6_validation_q13:section-3)**\n * [](climate_projections-cmip6_validation_q13:section-3.1)\n * [](climate_projections-cmip6_validation_q13:section-3.2)\n * [](climate_projections-cmip6_validation_q13:section-3.3)\n * [](climate_projections-cmip6_validation_q13:section-3.4)\n * [](climate_projections-cmip6_validation_q13:section-3.5)\n * [](climate_projections-cmip6_validation_q13:section-3.6)\n * [](climate_projections-cmip6_validation_q13:section-3.7)", "text_with_prefix": "EQC Quality Assessment: \"CMIP6 biases in the SPEI6 drought index over the Mediterranean region\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q13 | Category: Climate_Projections\nSection: CMIP6 biases in the SPEI6 drought index over the Mediterranean region > Methodology\n---\nThe Standardized Precipitation-Evapotranspiration Index accumulated to six months (SPEI6) has been calculated for this assessment — through the use of [xclim](https://xclim.readthedocs.io/en/stable/) — as a proxy to evaluate drought conditions for a subset of [CMIP6](https://cds.climate.copernicus.eu/datasets/projections-cmip6?tab=overview) models, which are evaluated against SPEI6 derived from [ERA5](https://cds.climate.copernicus.eu/datasets/reanalysis-era5-complete?tab=doc) reanalysis. The study covers the Mediterranean domain, following IPCC-AR6 regional definitions [[6]](https://doi.org/10.5194/essd-12-2959-2020), and results include spatial maps of SPEI6 changes for ERA5, CMIP6, and the biases of these changes. Additionally, time series showing the temporal evolution of drought across the domain are calculated, comparing trends from the models with ERA5 over the historical period considered within this assessment (1961–2020).\n\nPotential evapotranspiration (PET) is calculated with the Hargreaves method [[7]](https://doi.org/10.13031/2013.26773), which requires only maximum and minimum temperatures (and optionally mean temperature), providing a computationally efficient yet robust estimate (e.g., [[8]](https://doi.org/10.1002/joc.3887),). The accumulated series of precipitation minus PET is standardized for the reference period defined by the [C3S Atlas](https://confluence.ecmwf.int/display/CKB/Gridded+data+underpinning+the+Copernicus+Interactive+Climate+Atlas%3A+Description+of+the+datasets+and+variables#GriddeddataunderpinningtheCopernicusInteractiveClimateAtlas:Descriptionofthedatasetsandvariables) (1971–2010).\n\nFive key concepts are defined:\n\n- **Reference period:** The period used to standardize SPEI values, here 1971–2010, following the [C3S Atlas](https://confluence.ecmwf.int/display/CKB/Gridded+data+underpinning+the+Copernicus+Interactive+Climate+Atlas%3A+Description+of+the+datasets+and+variables#GriddeddataunderpinningtheCopernicusInteractiveClimateAtlas:Descriptionofthedatasetsandvariables). \n- **Baseline period:** The period against which changes are calculated, here 1961–1990. \n- **Target period:** The period for which changes are evaluated, here 1991–2020. \n- **Historical period:** The entire period considered in this assessment, here 1961–2020. \n- **Change assessment:** Differences in SPEI6 between the target period and the baseline, representing anomalies in drought conditions.\n\nThe analysis and results follow the next outline:\n\n**[](climate_projections-cmip6_validation_q13:section-1)**\n * [](climate_projections-cmip6_validation_q13:section-1.1)\n * [](climate_projections-cmip6_validation_q13:section-1.2)\n * [](climate_projections-cmip6_validation_q13:section-1.3)\n * [](climate_projections-cmip6_validation_q13:section-1.4)\n * [](climate_projections-cmip6_validation_q13:section-1.5)\n * [](climate_projections-cmip6_validation_q13:section-1.6)\n\n**[](climate_projections-cmip6_validation_q13:section-2)**\n * [](climate_projections-cmip6_validation_q13:section-2.1)\n * [](climate_projections-cmip6_validation_q13:section-2.2)\n * [](climate_projections-cmip6_validation_q13:section-2.3)\n\n**[](climate_projections-cmip6_validation_q13:section-3)**\n * [](climate_projections-cmip6_validation_q13:section-3.1)\n * [](climate_projections-cmip6_validation_q13:section-3.2)\n * [](climate_projections-cmip6_validation_q13:section-3.3)\n * [](climate_projections-cmip6_validation_q13:section-3.4)\n * [](climate_projections-cmip6_validation_q13:section-3.5)\n * [](climate_projections-cmip6_validation_q13:section-3.6)\n * [](climate_projections-cmip6_validation_q13:section-3.7)"} {"chunk_id": "climate_projections-cmip6_validation_q13__3852651395a2", "report_id": "climate_projections-cmip6_validation_q13", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q13", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CMIP6 biases in the SPEI6 drought index over the Mediterranean region > Analysis and results > 1. Parameters, requests and functions definition > 1.2. Define Parameters", "title": "CMIP6 biases in the SPEI6 drought index over the Mediterranean region", "chunk_index": 4, "token_count": 607, "text_raw": "In the \"Define Parameters\" section, various customisable options for the notebook are specified:\n\n- The historical period can be adjusted by modifying `historical_slice`; the default range is 1961-2020.\n- The reference period can be adjusted by modifying `reference_slice`; the default range is 1971–2010, following the [C3S Atlas](https://confluence.ecmwf.int/display/CKB/Gridded+data+underpinning+the+Copernicus+Interactive+Climate+Atlas%3A+Description+of+the+datasets+and+variables#GriddeddataunderpinningtheCopernicusInteractiveClimateAtlas:Descriptionofthedatasetsandvariables).\n- The baseline period can be adjusted by modifying `baseline_slice`; the default range is 1961-1990.\n- The target period can be adjusted by modifying `target_slice`; the default range is 1991-2020.\n- The `interpolation_method` parameter allows selecting the interpolation method when regridding is performed.\n- `collection_id` is not customisable for this assessment and is set to 'CMIP6'. Expert users can use this notebook as a guide to create similar analyses for other model sets (such as CORDEX).\n- The `area` parameter specifies the geographical domain of interest. By default, the Mediterranean domain is used, following IPCC-AR6 regional definitions [[6]](https://doi.org/10.5194/essd-12-2959-2020)\n- The `time_agg` allows selecting the time aggregation for the analysis. Options include: annual, DJF, MAM, JJA, SON, Jan, Feb, Mar, Apr, May, Jun, Jul, Aug, Sep, Oct, Nov, Dec.\n- SPEI accumulation window (`window`): Defines the accumulation period in months for the SPEI calculation. Common options include 3, 6, or 12 months, depending on whether short-term or long-term droughts are of interest. In this assessment, a six-month accumulation is used.\n- The `chunks` selection allows the user to define if dividing into chunks when downloading the data on their local machine. Although it does not significantly affect the analysis, it is recommended to keep the default value for optimal performance.\n\nPeriods definition\nInterpolation method\nCollection id\nDefine region for analysis\ntime aggregation\nSPEI accumulation\nChunks for download\n\n(climate_projections-cmip6_validation_q13:section-1.3)=", "text_with_prefix": "EQC Quality Assessment: \"CMIP6 biases in the SPEI6 drought index over the Mediterranean region\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q13 | Category: Climate_Projections\nSection: CMIP6 biases in the SPEI6 drought index over the Mediterranean region > Analysis and results > 1. Parameters, requests and functions definition > 1.2. Define Parameters\n---\nIn the \"Define Parameters\" section, various customisable options for the notebook are specified:\n\n- The historical period can be adjusted by modifying `historical_slice`; the default range is 1961-2020.\n- The reference period can be adjusted by modifying `reference_slice`; the default range is 1971–2010, following the [C3S Atlas](https://confluence.ecmwf.int/display/CKB/Gridded+data+underpinning+the+Copernicus+Interactive+Climate+Atlas%3A+Description+of+the+datasets+and+variables#GriddeddataunderpinningtheCopernicusInteractiveClimateAtlas:Descriptionofthedatasetsandvariables).\n- The baseline period can be adjusted by modifying `baseline_slice`; the default range is 1961-1990.\n- The target period can be adjusted by modifying `target_slice`; the default range is 1991-2020.\n- The `interpolation_method` parameter allows selecting the interpolation method when regridding is performed.\n- `collection_id` is not customisable for this assessment and is set to 'CMIP6'. Expert users can use this notebook as a guide to create similar analyses for other model sets (such as CORDEX).\n- The `area` parameter specifies the geographical domain of interest. By default, the Mediterranean domain is used, following IPCC-AR6 regional definitions [[6]](https://doi.org/10.5194/essd-12-2959-2020)\n- The `time_agg` allows selecting the time aggregation for the analysis. Options include: annual, DJF, MAM, JJA, SON, Jan, Feb, Mar, Apr, May, Jun, Jul, Aug, Sep, Oct, Nov, Dec.\n- SPEI accumulation window (`window`): Defines the accumulation period in months for the SPEI calculation. Common options include 3, 6, or 12 months, depending on whether short-term or long-term droughts are of interest. In this assessment, a six-month accumulation is used.\n- The `chunks` selection allows the user to define if dividing into chunks when downloading the data on their local machine. Although it does not significantly affect the analysis, it is recommended to keep the default value for optimal performance.\n\nPeriods definition\nInterpolation method\nCollection id\nDefine region for analysis\ntime aggregation\nSPEI accumulation\nChunks for download\n\n(climate_projections-cmip6_validation_q13:section-1.3)="} {"chunk_id": "climate_projections-cmip6_validation_q13__efcd2fd0b233", "report_id": "climate_projections-cmip6_validation_q13", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q13", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CMIP6 biases in the SPEI6 drought index over the Mediterranean region > Analysis and results > 1. Parameters, requests and functions definition > 1.3. Define models", "title": "CMIP6 biases in the SPEI6 drought index over the Mediterranean region", "chunk_index": 5, "token_count": 177, "text_raw": "The following climate analyses are performed using CMIP6 models that provide both the historical and SSP5-8.5 experiments, as well as maximum, minimum and mean temperatures and precipitation, within the Climate Data Store ([CDS](https://cds.climate.copernicus.eu/datasets/projections-cmip6?tab=overview)).\n\n(climate_projections-cmip6_validation_q13:section-1.4)=", "text_with_prefix": "EQC Quality Assessment: \"CMIP6 biases in the SPEI6 drought index over the Mediterranean region\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q13 | Category: Climate_Projections\nSection: CMIP6 biases in the SPEI6 drought index over the Mediterranean region > Analysis and results > 1. Parameters, requests and functions definition > 1.3. Define models\n---\nThe following climate analyses are performed using CMIP6 models that provide both the historical and SSP5-8.5 experiments, as well as maximum, minimum and mean temperatures and precipitation, within the Climate Data Store ([CDS](https://cds.climate.copernicus.eu/datasets/projections-cmip6?tab=overview)).\n\n(climate_projections-cmip6_validation_q13:section-1.4)="} {"chunk_id": "climate_projections-cmip6_validation_q13__ca4fe77c6c82", "report_id": "climate_projections-cmip6_validation_q13", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q13", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CMIP6 biases in the SPEI6 drought index over the Mediterranean region > Analysis and results > 1. Parameters, requests and functions definition > 1.4. Define ERA5 request", "title": "CMIP6 biases in the SPEI6 drought index over the Mediterranean region", "chunk_index": 6, "token_count": 139, "text_raw": "Within this notebook, ERA5 serves as the reference product. In this section, we set the required parameters for the cds-api data-request of ERA5.\n\n(climate_projections-cmip6_validation_q13:section-1.5)=", "text_with_prefix": "EQC Quality Assessment: \"CMIP6 biases in the SPEI6 drought index over the Mediterranean region\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q13 | Category: Climate_Projections\nSection: CMIP6 biases in the SPEI6 drought index over the Mediterranean region > Analysis and results > 1. Parameters, requests and functions definition > 1.4. Define ERA5 request\n---\nWithin this notebook, ERA5 serves as the reference product. In this section, we set the required parameters for the cds-api data-request of ERA5.\n\n(climate_projections-cmip6_validation_q13:section-1.5)="} {"chunk_id": "climate_projections-cmip6_validation_q13__7246f121d314", "report_id": "climate_projections-cmip6_validation_q13", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q13", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CMIP6 biases in the SPEI6 drought index over the Mediterranean region > Analysis and results > 1. Parameters, requests and functions definition > 1.5. Define model requests", "title": "CMIP6 biases in the SPEI6 drought index over the Mediterranean region", "chunk_index": 7, "token_count": 143, "text_raw": "In this section we set the required parameters for the cds-api data-request.\n\nOnly historical data\nHistorical until 2014\nFuture from 2015 to historical_slice.stop\n\n(climate_projections-cmip6_validation_q13:section-1.6)=", "text_with_prefix": "EQC Quality Assessment: \"CMIP6 biases in the SPEI6 drought index over the Mediterranean region\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q13 | Category: Climate_Projections\nSection: CMIP6 biases in the SPEI6 drought index over the Mediterranean region > Analysis and results > 1. Parameters, requests and functions definition > 1.5. Define model requests\n---\nIn this section we set the required parameters for the cds-api data-request.\n\nOnly historical data\nHistorical until 2014\nFuture from 2015 to historical_slice.stop\n\n(climate_projections-cmip6_validation_q13:section-1.6)="} {"chunk_id": "climate_projections-cmip6_validation_q13__c58f4cf12af0", "report_id": "climate_projections-cmip6_validation_q13", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q13", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CMIP6 biases in the SPEI6 drought index over the Mediterranean region > Analysis and results > 1. Parameters, requests and functions definition > 1.6. Functions to cache", "title": "CMIP6 biases in the SPEI6 drought index over the Mediterranean region", "chunk_index": 8, "token_count": 316, "text_raw": "In this section, functions that will be executed in the caching phase are defined. Caching is the process of storing copies of files in a temporary storage location, so that they can be accessed more quickly. This process also checks if the user has already downloaded a file, avoiding redundant downloads.\n\nDescription of the main functions:\n\n- **`compute_spei`**: Uses the xclim package to calculate the Standardized Precipitation-Evapotranspiration Index (SPEI). The `reference_slice` argument specifies the reference period used for standardization, while the `window` argument defines the number of months over which precipitation and evapotranspiration are accumulated (6 months in this assessment).\n\n- **`get_mask_below_03`**: Generates a mask identifying grid points where precipitation is below 0.3 mm/day, allowing filtering of extremely dry regions.\n\nDownload the other vars if we are dealing with CMIP6\nOriginal bounds for conservative interpolation\nApply correct daily reductions\nWrap into dataset\nSelect pr and\nApply correct daily reductions\n\n(climate_projections-cmip6_validation_q13:section-2)=", "text_with_prefix": "EQC Quality Assessment: \"CMIP6 biases in the SPEI6 drought index over the Mediterranean region\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q13 | Category: Climate_Projections\nSection: CMIP6 biases in the SPEI6 drought index over the Mediterranean region > Analysis and results > 1. Parameters, requests and functions definition > 1.6. Functions to cache\n---\nIn this section, functions that will be executed in the caching phase are defined. Caching is the process of storing copies of files in a temporary storage location, so that they can be accessed more quickly. This process also checks if the user has already downloaded a file, avoiding redundant downloads.\n\nDescription of the main functions:\n\n- **`compute_spei`**: Uses the xclim package to calculate the Standardized Precipitation-Evapotranspiration Index (SPEI). The `reference_slice` argument specifies the reference period used for standardization, while the `window` argument defines the number of months over which precipitation and evapotranspiration are accumulated (6 months in this assessment).\n\n- **`get_mask_below_03`**: Generates a mask identifying grid points where precipitation is below 0.3 mm/day, allowing filtering of extremely dry regions.\n\nDownload the other vars if we are dealing with CMIP6\nOriginal bounds for conservative interpolation\nApply correct daily reductions\nWrap into dataset\nSelect pr and\nApply correct daily reductions\n\n(climate_projections-cmip6_validation_q13:section-2)="} {"chunk_id": "climate_projections-cmip6_validation_q13__9d348eda12d1", "report_id": "climate_projections-cmip6_validation_q13", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q13", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CMIP6 biases in the SPEI6 drought index over the Mediterranean region > Analysis and results > 2. Downloading and processing > 2.1. Download and transform ERA5", "title": "CMIP6 biases in the SPEI6 drought index over the Mediterranean region", "chunk_index": 9, "token_count": 189, "text_raw": "In this section, the `download.download_and_transform` function from the 'c3s_eqc_automatic_quality_control' package is used to retrieve ERA5 hourly reference data, resample it to daily frequency, compute daily potential evapotranspiration, calculate the daily water balance, and derive the SPEI (SPEI6 for this assessment). Results are cached to avoid redundant downloads and repeated processing.\n\n(climate_projections-cmip6_validation_q13:section-2.2)=", "text_with_prefix": "EQC Quality Assessment: \"CMIP6 biases in the SPEI6 drought index over the Mediterranean region\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q13 | Category: Climate_Projections\nSection: CMIP6 biases in the SPEI6 drought index over the Mediterranean region > Analysis and results > 2. Downloading and processing > 2.1. Download and transform ERA5\n---\nIn this section, the `download.download_and_transform` function from the 'c3s_eqc_automatic_quality_control' package is used to retrieve ERA5 hourly reference data, resample it to daily frequency, compute daily potential evapotranspiration, calculate the daily water balance, and derive the SPEI (SPEI6 for this assessment). Results are cached to avoid redundant downloads and repeated processing.\n\n(climate_projections-cmip6_validation_q13:section-2.2)="} {"chunk_id": "climate_projections-cmip6_validation_q13__2c13072a5fb6", "report_id": "climate_projections-cmip6_validation_q13", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q13", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CMIP6 biases in the SPEI6 drought index over the Mediterranean region > Analysis and results > 2. Downloading and processing > 2.2. Download and transform models", "title": "CMIP6 biases in the SPEI6 drought index over the Mediterranean region", "chunk_index": 10, "token_count": 386, "text_raw": "In this section, the `download.download_and_transform` function from the 'c3s_eqc_automatic_quality_control' package is used to download daily data from CMIP6 models, compute daily potential evapotranspiration, calculate the daily water balance, and derive the SPEI (SPEI6 for this assessment). SPEI6 is computed on each model's native grid and also on the ERA5 grid to enable bias assessment. When regridding is required, it is performed for each essential climate variable before the index calculation to preserve standardisation. Results are cached to avoid redundant downloads and repeated processing.\n\nOriginal model\nInterpolated model\n\n```text\nmodel='access_cm2'\nmodel='canesm5'\nmodel='cmcc_esm2'\nmodel='cnrm_cm6_1'\nmodel='cnrm_cm6_1_hr'\nmodel='cnrm_esm2_1'\nmodel='ec_earth3_cc'\nmodel='gfdl_esm4'\nmodel='inm_cm4_8'\nmodel='inm_cm5_0'\nmodel='kace_1_0_g'\nmodel='miroc6'\nmodel='miroc_es2l'\nmodel='mpi_esm1_2_lr'\nmodel='mri_esm2_0'\nmodel='nesm3'\nmodel='noresm2_mm'\n```\n\n(climate_projections-cmip6_validation_q13:section-2.3)=", "text_with_prefix": "EQC Quality Assessment: \"CMIP6 biases in the SPEI6 drought index over the Mediterranean region\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q13 | Category: Climate_Projections\nSection: CMIP6 biases in the SPEI6 drought index over the Mediterranean region > Analysis and results > 2. Downloading and processing > 2.2. Download and transform models\n---\nIn this section, the `download.download_and_transform` function from the 'c3s_eqc_automatic_quality_control' package is used to download daily data from CMIP6 models, compute daily potential evapotranspiration, calculate the daily water balance, and derive the SPEI (SPEI6 for this assessment). SPEI6 is computed on each model's native grid and also on the ERA5 grid to enable bias assessment. When regridding is required, it is performed for each essential climate variable before the index calculation to preserve standardisation. Results are cached to avoid redundant downloads and repeated processing.\n\nOriginal model\nInterpolated model\n\n```text\nmodel='access_cm2'\nmodel='canesm5'\nmodel='cmcc_esm2'\nmodel='cnrm_cm6_1'\nmodel='cnrm_cm6_1_hr'\nmodel='cnrm_esm2_1'\nmodel='ec_earth3_cc'\nmodel='gfdl_esm4'\nmodel='inm_cm4_8'\nmodel='inm_cm5_0'\nmodel='kace_1_0_g'\nmodel='miroc6'\nmodel='miroc_es2l'\nmodel='mpi_esm1_2_lr'\nmodel='mri_esm2_0'\nmodel='nesm3'\nmodel='noresm2_mm'\n```\n\n(climate_projections-cmip6_validation_q13:section-2.3)="} {"chunk_id": "climate_projections-cmip6_validation_q13__35f3c2971a3a", "report_id": "climate_projections-cmip6_validation_q13", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q13", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CMIP6 biases in the SPEI6 drought index over the Mediterranean region > Analysis and results > 2. Downloading and processing > 2.3. Apply land-sea mask and change some attributes", "title": "CMIP6 biases in the SPEI6 drought index over the Mediterranean region", "chunk_index": 11, "token_count": 432, "text_raw": "This section changes some attributes and applies a land–sea mask to all models, as well as to ERA5. A bare-earth (barrean) mask is also applied to ensure meaningful results in arid regions, as in the [C3S atlas](https://atlas.climate.copernicus.eu/atlas). In particular, we follow the methodology available in their [Github repository](https://github.com/ecmwf-projects/c3s-atlas/blob/main/auxiliar/barrean_mask.ipynb). While their implementation of the mask also considers factors such as snow depth and vegetation cover, here we only apply the filter based on annual mean precipitation for 1991–2020 being less than 0.3 mm day⁻¹. This simplification is justified because, in the region under consideration, the bare-earth mask obtained with the full set of filters closely matches the one derived from the precipitation threshold alone.\n\n**Note:** `ds_interpolated` contains data from the models regridded to the ERA5 grid. `model_datasets` contain the same data on the original grid of each model. Regridding is performed for every essential climate variable prior to the index calculation in order to avoid compromising standardisation.\n\nDownload and prepare lsm\nEnsure CF-compliant attributes for model datasets\nRegrid each model dataset\nApply precipitation mask to ERA5 and interpolated\nApply PR mask to each model\n\n```text\n100%|██████████| 1/1 [00:00<00:00, 27.58it/s]\n```\n\n(climate_projections-cmip6_validation_q13:section-3)=", "text_with_prefix": "EQC Quality Assessment: \"CMIP6 biases in the SPEI6 drought index over the Mediterranean region\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q13 | Category: Climate_Projections\nSection: CMIP6 biases in the SPEI6 drought index over the Mediterranean region > Analysis and results > 2. Downloading and processing > 2.3. Apply land-sea mask and change some attributes\n---\nThis section changes some attributes and applies a land–sea mask to all models, as well as to ERA5. A bare-earth (barrean) mask is also applied to ensure meaningful results in arid regions, as in the [C3S atlas](https://atlas.climate.copernicus.eu/atlas). In particular, we follow the methodology available in their [Github repository](https://github.com/ecmwf-projects/c3s-atlas/blob/main/auxiliar/barrean_mask.ipynb). While their implementation of the mask also considers factors such as snow depth and vegetation cover, here we only apply the filter based on annual mean precipitation for 1991–2020 being less than 0.3 mm day⁻¹. This simplification is justified because, in the region under consideration, the bare-earth mask obtained with the full set of filters closely matches the one derived from the precipitation threshold alone.\n\n**Note:** `ds_interpolated` contains data from the models regridded to the ERA5 grid. `model_datasets` contain the same data on the original grid of each model. Regridding is performed for every essential climate variable prior to the index calculation in order to avoid compromising standardisation.\n\nDownload and prepare lsm\nEnsure CF-compliant attributes for model datasets\nRegrid each model dataset\nApply precipitation mask to ERA5 and interpolated\nApply PR mask to each model\n\n```text\n100%|██████████| 1/1 [00:00<00:00, 27.58it/s]\n```\n\n(climate_projections-cmip6_validation_q13:section-3)="} {"chunk_id": "climate_projections-cmip6_validation_q13__8407bc1bb18a", "report_id": "climate_projections-cmip6_validation_q13", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q13", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CMIP6 biases in the SPEI6 drought index over the Mediterranean region > Analysis and results > 3. Plot and describe results", "title": "CMIP6 biases in the SPEI6 drought index over the Mediterranean region", "chunk_index": 12, "token_count": 259, "text_raw": "This section will display the following results:\n\n- SPEI6 change maps comparing ERA5 and the ensemble median (defined as the median of the change values for the selected subset of models at each grid cell). The layout includes ERA5, the ensemble median, the ensemble median bias, and the ensemble spread (calculated as the standard deviation of the change across the selected subset of models). \n- SPEI6 change maps for each individual model. \n- Bias maps of the SPEI6 change for each model. \n- Time series of SPEI6 change (averaged over the Mediterranean region).\n\n**Note:** SPEI6 change is calculated at each grid point as the arithmetic difference between climatologies in the target period (1991–2020) and the baseline period (1961–1990).\n\n(climate_projections-cmip6_validation_q13:section-3.1)=", "text_with_prefix": "EQC Quality Assessment: \"CMIP6 biases in the SPEI6 drought index over the Mediterranean region\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q13 | Category: Climate_Projections\nSection: CMIP6 biases in the SPEI6 drought index over the Mediterranean region > Analysis and results > 3. Plot and describe results\n---\nThis section will display the following results:\n\n- SPEI6 change maps comparing ERA5 and the ensemble median (defined as the median of the change values for the selected subset of models at each grid cell). The layout includes ERA5, the ensemble median, the ensemble median bias, and the ensemble spread (calculated as the standard deviation of the change across the selected subset of models). \n- SPEI6 change maps for each individual model. \n- Bias maps of the SPEI6 change for each model. \n- Time series of SPEI6 change (averaged over the Mediterranean region).\n\n**Note:** SPEI6 change is calculated at each grid point as the arithmetic difference between climatologies in the target period (1991–2020) and the baseline period (1961–1990).\n\n(climate_projections-cmip6_validation_q13:section-3.1)="} {"chunk_id": "climate_projections-cmip6_validation_q13__e43b4ad11d1a", "report_id": "climate_projections-cmip6_validation_q13", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q13", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CMIP6 biases in the SPEI6 drought index over the Mediterranean region > Analysis and results > 3. Plot and describe results > 3.1. Define plotting functions", "title": "CMIP6 biases in the SPEI6 drought index over the Mediterranean region", "chunk_index": 13, "token_count": 507, "text_raw": "The functions presented here are used to calculate SPEI6 changes and generate corresponding layout plots. Three types of layout can be displayed, depending on the plotting function:\n\n1. Reference and ensemble summary: Includes the ERA5 reference product, the ensemble median, the bias of the ensemble median, and the ensemble spread. This is generated using `plot_ensemble()`.\n\n2. Individual models: Displays all models individually using `plot_models()`.\n\n3. Model biases: Displays the bias of each model relative to the reference, also using `plot_models()`.\n\n**Calculation of SPEI6 change:**\n\n- `compute_spei_change()` calculates SPEI6 change at each grid point as the arithmetic difference between monthly climatologies of the target period (1991–2020) and the baseline period (1961–1990). It then aggregates the change according to the user-specified `time_agg` (annual, seasonal, or monthly).\n\n- `compute_change_4timeseries()` computes SPEI6 anomalies relative to the baseline climatology for each month of the timeseries and aggregates them according to the `time_agg` parameter. This provides monthly, seasonal, or annual-level information depending on the user’s selection.\n\nMapping months to names\nMapping seasons to months\nSelect baseline and target periods\nMonthly climatologies\nAggregate according to user choice\nMapping months to names\nMapping seasons to months\nSelect baseline period\nMonthly climatology for the baseline\nCompute monthly anomalies relative to baseline climatology\nPreserve attributes\nAggregate anomalies\nGroup by year\nSelect months for the season, then group by year\nSelect the month, keep time as is\n--- Determine colour parameters ---\nUse user-defined cmap_params if provided; otherwise compute automatically\n--- Plot each model ---\n--- Hide empty axes ---\n--- Shared extent and colourbar ---\nDefault behaviour\nCreate figure and axes\n--- ERA5 plot ---\n--- Ensemble Median ---\n--- Ensemble Bias (Median - ERA5) ---\n--- Ensemble Standard Deviation ---\n\n(climate_projections-cmip6_validation_q13:section-3.2)=", "text_with_prefix": "EQC Quality Assessment: \"CMIP6 biases in the SPEI6 drought index over the Mediterranean region\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q13 | Category: Climate_Projections\nSection: CMIP6 biases in the SPEI6 drought index over the Mediterranean region > Analysis and results > 3. Plot and describe results > 3.1. Define plotting functions\n---\nThe functions presented here are used to calculate SPEI6 changes and generate corresponding layout plots. Three types of layout can be displayed, depending on the plotting function:\n\n1. Reference and ensemble summary: Includes the ERA5 reference product, the ensemble median, the bias of the ensemble median, and the ensemble spread. This is generated using `plot_ensemble()`.\n\n2. Individual models: Displays all models individually using `plot_models()`.\n\n3. Model biases: Displays the bias of each model relative to the reference, also using `plot_models()`.\n\n**Calculation of SPEI6 change:**\n\n- `compute_spei_change()` calculates SPEI6 change at each grid point as the arithmetic difference between monthly climatologies of the target period (1991–2020) and the baseline period (1961–1990). It then aggregates the change according to the user-specified `time_agg` (annual, seasonal, or monthly).\n\n- `compute_change_4timeseries()` computes SPEI6 anomalies relative to the baseline climatology for each month of the timeseries and aggregates them according to the `time_agg` parameter. This provides monthly, seasonal, or annual-level information depending on the user’s selection.\n\nMapping months to names\nMapping seasons to months\nSelect baseline and target periods\nMonthly climatologies\nAggregate according to user choice\nMapping months to names\nMapping seasons to months\nSelect baseline period\nMonthly climatology for the baseline\nCompute monthly anomalies relative to baseline climatology\nPreserve attributes\nAggregate anomalies\nGroup by year\nSelect months for the season, then group by year\nSelect the month, keep time as is\n--- Determine colour parameters ---\nUse user-defined cmap_params if provided; otherwise compute automatically\n--- Plot each model ---\n--- Hide empty axes ---\n--- Shared extent and colourbar ---\nDefault behaviour\nCreate figure and axes\n--- ERA5 plot ---\n--- Ensemble Median ---\n--- Ensemble Bias (Median - ERA5) ---\n--- Ensemble Standard Deviation ---\n\n(climate_projections-cmip6_validation_q13:section-3.2)="} {"chunk_id": "climate_projections-cmip6_validation_q13__6c6a6e8c3af1", "report_id": "climate_projections-cmip6_validation_q13", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q13", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CMIP6 biases in the SPEI6 drought index over the Mediterranean region > Analysis and results > 3. Plot and describe results > 3.2. Plot ensemble maps", "title": "CMIP6 biases in the SPEI6 drought index over the Mediterranean region", "chunk_index": 14, "token_count": 462, "text_raw": "In this section, we invoke the `plot_ensemble()` function to visualise the SPEI6 change for the following: (a) the reference ERA5 product, (b) the ensemble median (defined as the median of the change values for the selected subset of models at each grid cell), (c) the bias of the ensemble median, and (d) the ensemble spread (calculated as the standard deviation of the change across the selected subset of models).\n\nChange default colorbars\nShared colorbar for ERA5 and ensemble median\nGet spei change\nShow the plot\n\n
\n
\n

Fig 1. \n Spatial patterns of SPEI6 change over the Mediterranean region. Panels show (a) ERA5 reference, (b) CMIP6 ensemble median (median of the selected subset of models at each grid cell), (c) ensemble median bias relative to ERA5, and (d) ensemble spread (standard deviation across the selected models). SPEI6 is standardised over the reference period 1971–2010. SPEI6 change is calculated at each grid point as the difference between the monthly climatologies of the target period (1991–2020) and the baseline period (1961–1990), and then averaged to obtain the annual mean.\n\n

\n\n\n
\nNOTE:
\nThe regridding artefacts visible in panel (d) result from the use of a conservative regridding method, which is commonly recommended when dealing with precipitation. They typically arise from large values in specific models and differences among the models’ native grids.\n\n(climate_projections-cmip6_validation_q13:section-3.3)=", "text_with_prefix": "EQC Quality Assessment: \"CMIP6 biases in the SPEI6 drought index over the Mediterranean region\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q13 | Category: Climate_Projections\nSection: CMIP6 biases in the SPEI6 drought index over the Mediterranean region > Analysis and results > 3. Plot and describe results > 3.2. Plot ensemble maps\n---\nIn this section, we invoke the `plot_ensemble()` function to visualise the SPEI6 change for the following: (a) the reference ERA5 product, (b) the ensemble median (defined as the median of the change values for the selected subset of models at each grid cell), (c) the bias of the ensemble median, and (d) the ensemble spread (calculated as the standard deviation of the change across the selected subset of models).\n\nChange default colorbars\nShared colorbar for ERA5 and ensemble median\nGet spei change\nShow the plot\n\n
\n
\n

Fig 1. \n Spatial patterns of SPEI6 change over the Mediterranean region. Panels show (a) ERA5 reference, (b) CMIP6 ensemble median (median of the selected subset of models at each grid cell), (c) ensemble median bias relative to ERA5, and (d) ensemble spread (standard deviation across the selected models). SPEI6 is standardised over the reference period 1971–2010. SPEI6 change is calculated at each grid point as the difference between the monthly climatologies of the target period (1991–2020) and the baseline period (1961–1990), and then averaged to obtain the annual mean.\n\n

\n\n\n
\nNOTE:
\nThe regridding artefacts visible in panel (d) result from the use of a conservative regridding method, which is commonly recommended when dealing with precipitation. They typically arise from large values in specific models and differences among the models’ native grids.\n\n(climate_projections-cmip6_validation_q13:section-3.3)="} {"chunk_id": "climate_projections-cmip6_validation_q13__19f34cc0b8ca", "report_id": "climate_projections-cmip6_validation_q13", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q13", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CMIP6 biases in the SPEI6 drought index over the Mediterranean region > Analysis and results > 3. Plot and describe results > 3.3. Plot model maps", "title": "CMIP6 biases in the SPEI6 drought index over the Mediterranean region", "chunk_index": 15, "token_count": 261, "text_raw": "In this section, we invoke the `plot_models()` function to visualise the SPEI6 change of every model individually. Note that the model data used in this section maintains its original grid.\n\n
\n
\n

Fig 2.\n Spatial patterns of SPEI6 change for each individual CMIP6 model over the Mediterranean region. SPEI6 is standardised over the reference period 1971–2010. For each model, SPEI6 change is calculated at each grid point as the difference between the monthly climatologies of the target period (1991–2020) and the baseline period (1961–1990), and then averaged to obtain the annual mean. \n

\n\n(climate_projections-cmip6_validation_q13:section-3.4)=", "text_with_prefix": "EQC Quality Assessment: \"CMIP6 biases in the SPEI6 drought index over the Mediterranean region\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q13 | Category: Climate_Projections\nSection: CMIP6 biases in the SPEI6 drought index over the Mediterranean region > Analysis and results > 3. Plot and describe results > 3.3. Plot model maps\n---\nIn this section, we invoke the `plot_models()` function to visualise the SPEI6 change of every model individually. Note that the model data used in this section maintains its original grid.\n\n
\n
\n

Fig 2.\n Spatial patterns of SPEI6 change for each individual CMIP6 model over the Mediterranean region. SPEI6 is standardised over the reference period 1971–2010. For each model, SPEI6 change is calculated at each grid point as the difference between the monthly climatologies of the target period (1991–2020) and the baseline period (1961–1990), and then averaged to obtain the annual mean. \n

\n\n(climate_projections-cmip6_validation_q13:section-3.4)="} {"chunk_id": "climate_projections-cmip6_validation_q13__d2a8e25ab0a3", "report_id": "climate_projections-cmip6_validation_q13", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q13", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CMIP6 biases in the SPEI6 drought index over the Mediterranean region > Analysis and results > 3. Plot and describe results > 3.4. Plot bias maps", "title": "CMIP6 biases in the SPEI6 drought index over the Mediterranean region", "chunk_index": 16, "token_count": 303, "text_raw": "In this section, we invoke the `plot_models()` function to visualise the bias of the vSPEI6 change for every model individually. Note that the model data used in this section has previously been interpolated to the ERA5 grid. Regridding is performed for each essential climate variable before the index calculation to preserve standardisation.\n\nPlot using the shared bias parameters\n\n
\n
\n

Fig 3.\n Bias of SPEI6 change for each individual CMIP6 model relative to ERA5 over the Mediterranean region. SPEI6 is standardised over the reference period 1971–2010. For each model and ERA5, SPEI6 change is calculated at each grid point as the difference between the monthly climatologies of the target period (1991–2020) and the baseline period (1961–1990), and then averaged to obtain the annual mean. \n

\n\n(climate_projections-cmip6_validation_q13:section-3.5)=", "text_with_prefix": "EQC Quality Assessment: \"CMIP6 biases in the SPEI6 drought index over the Mediterranean region\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q13 | Category: Climate_Projections\nSection: CMIP6 biases in the SPEI6 drought index over the Mediterranean region > Analysis and results > 3. Plot and describe results > 3.4. Plot bias maps\n---\nIn this section, we invoke the `plot_models()` function to visualise the bias of the vSPEI6 change for every model individually. Note that the model data used in this section has previously been interpolated to the ERA5 grid. Regridding is performed for each essential climate variable before the index calculation to preserve standardisation.\n\nPlot using the shared bias parameters\n\n
\n
\n

Fig 3.\n Bias of SPEI6 change for each individual CMIP6 model relative to ERA5 over the Mediterranean region. SPEI6 is standardised over the reference period 1971–2010. For each model and ERA5, SPEI6 change is calculated at each grid point as the difference between the monthly climatologies of the target period (1991–2020) and the baseline period (1961–1990), and then averaged to obtain the annual mean. \n

\n\n(climate_projections-cmip6_validation_q13:section-3.5)="} {"chunk_id": "climate_projections-cmip6_validation_q13__f02a664cb008", "report_id": "climate_projections-cmip6_validation_q13", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q13", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CMIP6 biases in the SPEI6 drought index over the Mediterranean region > Analysis and results > 3. Plot and describe results > 3.5. Timeseries", "title": "CMIP6 biases in the SPEI6 drought index over the Mediterranean region", "chunk_index": 17, "token_count": 482, "text_raw": "This section examines the time series of SPEI6 anomalies relative to the baseline climatology. We first use `compute_change_4timeseries()`, which calculates SPEI6 anomalies for each month of the timeseries (relative to the baseline climatology of each month) and aggregates them according to the `time_agg parameter`. Spatially averaged values are then obtained. Light blue shading highlights the baseline period, while light orange shading marks the target period. The analysis compares the CMIP6 ensemble median, ERA5, and individual models, displaying trends for both ERA5 and the ensemble median. Additionally, it shows the ensemble spread and the slope values of the trends for ERA5 and the CMIP6 ensemble median.\n\nDefine colors\nCutout region\nCompute the change/anomaly for ERA5\nRegionalise and compute spatially weighted mean\nCompute change/anomaly for each model and spatially aggregate\nEnsemble statistics\nInitialize plot\nShade baseline period (light blue)\nShade target period (light orange)\n--- Create custom legend entries with full opacity ---\nPlot individual model time series\nAdd legend entry for individual models\nPlot CMIP6 median\nFill ±1 std (spread)\nCompute trends\nLabels and grid\nFormat x-axis as years\nAnnotate trends\n\n
\n
\n

Fig 4.\n Time series of SPEI6 anomalies over the Mediterranean region, calculated relative to the baseline climatology. Monthly anomalies are aggregated according to the specified timescale and then spatially averaged over the region. Light blue shading indicates the baseline period, and light orange shading indicates the target period. The figure compares ERA5, the CMIP6 ensemble median, and individual models, showing trends for ERA5 and the ensemble median, as well as the ensemble spread and slope values of the trends.\n

\n\n(climate_projections-cmip6_validation_q13:section-3.6)=", "text_with_prefix": "EQC Quality Assessment: \"CMIP6 biases in the SPEI6 drought index over the Mediterranean region\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q13 | Category: Climate_Projections\nSection: CMIP6 biases in the SPEI6 drought index over the Mediterranean region > Analysis and results > 3. Plot and describe results > 3.5. Timeseries\n---\nThis section examines the time series of SPEI6 anomalies relative to the baseline climatology. We first use `compute_change_4timeseries()`, which calculates SPEI6 anomalies for each month of the timeseries (relative to the baseline climatology of each month) and aggregates them according to the `time_agg parameter`. Spatially averaged values are then obtained. Light blue shading highlights the baseline period, while light orange shading marks the target period. The analysis compares the CMIP6 ensemble median, ERA5, and individual models, displaying trends for both ERA5 and the ensemble median. Additionally, it shows the ensemble spread and the slope values of the trends for ERA5 and the CMIP6 ensemble median.\n\nDefine colors\nCutout region\nCompute the change/anomaly for ERA5\nRegionalise and compute spatially weighted mean\nCompute change/anomaly for each model and spatially aggregate\nEnsemble statistics\nInitialize plot\nShade baseline period (light blue)\nShade target period (light orange)\n--- Create custom legend entries with full opacity ---\nPlot individual model time series\nAdd legend entry for individual models\nPlot CMIP6 median\nFill ±1 std (spread)\nCompute trends\nLabels and grid\nFormat x-axis as years\nAnnotate trends\n\n
\n
\n

Fig 4.\n Time series of SPEI6 anomalies over the Mediterranean region, calculated relative to the baseline climatology. Monthly anomalies are aggregated according to the specified timescale and then spatially averaged over the region. Light blue shading indicates the baseline period, and light orange shading indicates the target period. The figure compares ERA5, the CMIP6 ensemble median, and individual models, showing trends for ERA5 and the ensemble median, as well as the ensemble spread and slope values of the trends.\n

\n\n(climate_projections-cmip6_validation_q13:section-3.6)="} {"chunk_id": "climate_projections-cmip6_validation_q13__99f7c4676fae", "report_id": "climate_projections-cmip6_validation_q13", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q13", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CMIP6 biases in the SPEI6 drought index over the Mediterranean region > Analysis and results > 3. Plot and describe results > 3.6. Results summary", "title": "CMIP6 biases in the SPEI6 drought index over the Mediterranean region", "chunk_index": 18, "token_count": 513, "text_raw": "- ERA5 indicates a general intensification of drought conditions over the Mediterranean region when comparing the baseline (1961–1990) and target (1991–2020) periods.\n\n- The CMIP6 ensemble median captures this SPEI6 change signal, but the magnitude of the increase in drought conditions is underestimated, particularly over eastern Spain, southern France, and much of Turkey.\n\n- With respect to inter-model spread, models do not reproduce the same spatial pattern. The largest spread appears over North African countries bordering the Mediterranean, Turkey, Lebanon, and Syria, and to a lesser extent over the Balkans, Portugal, and central and western Spain.\n\n- Time series of spatial means over the Mediterranean confirm this increase in drought conditions and the underestimation by CMIP6 models, with trends of –0.15 for ERA5 and –0.08 (about half) for the CMIP6 median.\n\n- Although users can modify the time aggregation in this notebook to analyse other seasons or specific months, the [C3S Atlas](https://confluence.ecmwf.int/display/CKB/Gridded+data+underpinning+the+Copernicus+Interactive+Climate+Atlas%3A+Description+of+the+datasets+and+variables#GriddeddataunderpinningtheCopernicusInteractiveClimateAtlas:Descriptionofthedatasetsandvariables) also provides complementary information on CMIP6 and ERA5 SPEI changes by season.\n\n- According to ERA5, SPEI6 decreased most strongly in autumn and summer, although increases in some localised regions during autumn moderate the Mediterranean’s spatial mean. Spring shows weaker but more spatially consistent decreases, producing mean values similar to autumn, while winter (DJF) exhibits the smallest reductions. CMIP6 broadly reproduces this seasonal pattern, with the strongest decreases in autumn and summer and weaker changes in winter and spring; in spring, some northern parts of the Euro-Mediterranean region even show increases, contrary to ERA5.\n\n(climate_projections-cmip6_validation_q13:section-3.7)=", "text_with_prefix": "EQC Quality Assessment: \"CMIP6 biases in the SPEI6 drought index over the Mediterranean region\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q13 | Category: Climate_Projections\nSection: CMIP6 biases in the SPEI6 drought index over the Mediterranean region > Analysis and results > 3. Plot and describe results > 3.6. Results summary\n---\n- ERA5 indicates a general intensification of drought conditions over the Mediterranean region when comparing the baseline (1961–1990) and target (1991–2020) periods.\n\n- The CMIP6 ensemble median captures this SPEI6 change signal, but the magnitude of the increase in drought conditions is underestimated, particularly over eastern Spain, southern France, and much of Turkey.\n\n- With respect to inter-model spread, models do not reproduce the same spatial pattern. The largest spread appears over North African countries bordering the Mediterranean, Turkey, Lebanon, and Syria, and to a lesser extent over the Balkans, Portugal, and central and western Spain.\n\n- Time series of spatial means over the Mediterranean confirm this increase in drought conditions and the underestimation by CMIP6 models, with trends of –0.15 for ERA5 and –0.08 (about half) for the CMIP6 median.\n\n- Although users can modify the time aggregation in this notebook to analyse other seasons or specific months, the [C3S Atlas](https://confluence.ecmwf.int/display/CKB/Gridded+data+underpinning+the+Copernicus+Interactive+Climate+Atlas%3A+Description+of+the+datasets+and+variables#GriddeddataunderpinningtheCopernicusInteractiveClimateAtlas:Descriptionofthedatasetsandvariables) also provides complementary information on CMIP6 and ERA5 SPEI changes by season.\n\n- According to ERA5, SPEI6 decreased most strongly in autumn and summer, although increases in some localised regions during autumn moderate the Mediterranean’s spatial mean. Spring shows weaker but more spatially consistent decreases, producing mean values similar to autumn, while winter (DJF) exhibits the smallest reductions. CMIP6 broadly reproduces this seasonal pattern, with the strongest decreases in autumn and summer and weaker changes in winter and spring; in spring, some northern parts of the Euro-Mediterranean region even show increases, contrary to ERA5.\n\n(climate_projections-cmip6_validation_q13:section-3.7)="} {"chunk_id": "climate_projections-cmip6_validation_q13__52224fbdb516", "report_id": "climate_projections-cmip6_validation_q13", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q13", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CMIP6 biases in the SPEI6 drought index over the Mediterranean region > Analysis and results > 3. Plot and describe results > 3.7. Implications for the users", "title": "CMIP6 biases in the SPEI6 drought index over the Mediterranean region", "chunk_index": 19, "token_count": 379, "text_raw": "- Models do not reproduce the same exact spatial pattern, but the ensemble median captures the SPEI6 change. This underlines the importance of considering large ensembles rather than relying on a single or a few models.\n\n- Although the CMIP6 ensemble median captures the intensification of drought conditions, caution is needed when looking into the future. Precipitation projections generally involve more uncertainty than temperature, and model biases may evolve under changing climate conditions [[9]](https://doi.org/10.5194/hess-25-273-2021).\n\n- This assessment shows consistent results with those obtained in the [C3S Atlas](https://confluence.ecmwf.int/display/CKB/Gridded+data+underpinning+the+Copernicus+Interactive+Climate+Atlas%3A+Description+of+the+datasets+and+variables#GriddeddataunderpinningtheCopernicusInteractiveClimateAtlas:Descriptionofthedatasetsandvariables). In addition, it provides users with insight into the rationale behind SPEI and how models reproduce historical conditions over the Mediterranean, reinforcing its value as a tool for water management and drought preparedness.\n\n- The pronounced inter-model spread over North Africa, Turkey, and the eastern Mediterranean implies that users in these regions may face higher future uncertainty. This highlights the need to stress-test adaptation strategies against a wide range of plausible futures rather than relying on a single scenario.", "text_with_prefix": "EQC Quality Assessment: \"CMIP6 biases in the SPEI6 drought index over the Mediterranean region\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q13 | Category: Climate_Projections\nSection: CMIP6 biases in the SPEI6 drought index over the Mediterranean region > Analysis and results > 3. Plot and describe results > 3.7. Implications for the users\n---\n- Models do not reproduce the same exact spatial pattern, but the ensemble median captures the SPEI6 change. This underlines the importance of considering large ensembles rather than relying on a single or a few models.\n\n- Although the CMIP6 ensemble median captures the intensification of drought conditions, caution is needed when looking into the future. Precipitation projections generally involve more uncertainty than temperature, and model biases may evolve under changing climate conditions [[9]](https://doi.org/10.5194/hess-25-273-2021).\n\n- This assessment shows consistent results with those obtained in the [C3S Atlas](https://confluence.ecmwf.int/display/CKB/Gridded+data+underpinning+the+Copernicus+Interactive+Climate+Atlas%3A+Description+of+the+datasets+and+variables#GriddeddataunderpinningtheCopernicusInteractiveClimateAtlas:Descriptionofthedatasetsandvariables). In addition, it provides users with insight into the rationale behind SPEI and how models reproduce historical conditions over the Mediterranean, reinforcing its value as a tool for water management and drought preparedness.\n\n- The pronounced inter-model spread over North Africa, Turkey, and the eastern Mediterranean implies that users in these regions may face higher future uncertainty. This highlights the need to stress-test adaptation strategies against a wide range of plausible futures rather than relying on a single scenario."} {"chunk_id": "climate_projections-cmip6_validation_q13__53f6db47d463", "report_id": "climate_projections-cmip6_validation_q13", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q13", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CMIP6 biases in the SPEI6 drought index over the Mediterranean region > ℹ️ If you want to know more > Key resources", "title": "CMIP6 biases in the SPEI6 drought index over the Mediterranean region", "chunk_index": 20, "token_count": 270, "text_raw": "Some key resources and further reading were linked throughout this assessment.\n\nThe CDS catalogue entries for the data used were:\n* CMIP6 climate projections (daily - daily maximum near-surface air temperature, daily minimum near-surface air temperature and precipitation): https://cds.climate.copernicus.eu/datasets/projections-cmip6?tab=overview\n* ERA5 hourly data on single levels from 1940 to present (2m temperature and total precipitation): https://cds.climate.copernicus.eu/datasets/reanalysis-era5-pressure-levels-monthly-means?tab=overview\n\nCode libraries used:\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)", "text_with_prefix": "EQC Quality Assessment: \"CMIP6 biases in the SPEI6 drought index over the Mediterranean region\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q13 | Category: Climate_Projections\nSection: CMIP6 biases in the SPEI6 drought index over the Mediterranean region > ℹ️ If you want to know more > Key resources\n---\nSome key resources and further reading were linked throughout this assessment.\n\nThe CDS catalogue entries for the data used were:\n* CMIP6 climate projections (daily - daily maximum near-surface air temperature, daily minimum near-surface air temperature and precipitation): https://cds.climate.copernicus.eu/datasets/projections-cmip6?tab=overview\n* ERA5 hourly data on single levels from 1940 to present (2m temperature and total precipitation): https://cds.climate.copernicus.eu/datasets/reanalysis-era5-pressure-levels-monthly-means?tab=overview\n\nCode libraries used:\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)"} {"chunk_id": "climate_projections-cmip6_validation_q13__40ba92ab12be", "report_id": "climate_projections-cmip6_validation_q13", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q13", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CMIP6 biases in the SPEI6 drought index over the Mediterranean region > ℹ️ If you want to know more > References", "title": "CMIP6 biases in the SPEI6 drought index over the Mediterranean region", "chunk_index": 21, "token_count": 1062, "text_raw": "[[1]](https://doi.org/10.1175/2009JCLI2909.1) Vicente-Serrano, S. M., Beguerı́a, S., and López-Moreno, J. I., 2010. A multiscalar drought index sensitive to global warming: The standardized precipitation evapotranspiration index, Journal of Climate, 23, 1696–1718. https://doi.org/10.1175/2009JCLI2909.1\n\n[[2]](https://digitalcommons.unl.edu/droughtfacpub/69/) Wilhite, D. A., 2000. Droughts as a natural hazard: concepts and definitions. In: DROUGHT, A Global Assessment, vol. I and II, Routledge Hazards and Disasters Series, Routledge.\n\n[[3]](https://doi.org/10.1175/2012EI000434.1) Vicente-Serrano, S. M., Beguerı́a, S., Lorenzo-Lacruz, J., Camarero, J. J., López-Moreno, J. I., Azorin-Molina, C., Revuelto, J., Morán-Tejeda, E., and Sanchez-Lorenzo, A., 2012. Performance of drought indices for ecological, agricultural, and hydrological applications, Earth Interactions, 16, https://doi.org/10.1175/2012EI000434.1\n\n[[4]](https://doi.org/10.1016/j.scitotenv.2019.135245) Yao, N., Li, L., Feng, P., Feng, H., Liu, D. L., Liu, Y., Jiang, K., Hu, X., and Li, Y., 2020. Projections of drought characteristics in China based on a standardized precipitation and evapotranspiration index and multiple GCMs. Science of The Total Environment, 704, 135245. https://doi.org/10.1016/j.scitotenv.2019.135245\n\n[[5]](https://doi.org/10.3390/w17101507) Tuong, V. Q., Kiet, B. A., and Pham, T. T., 2025. Assessment of Future Drought Characteristics Using Various Temporal Scales and Multiple Drought Indices over Mekong Basin Under Climate Changes. Water, 17(10), 1507. https://doi.org/10.3390/w17101507\n\n[[6]](https://doi.org/10.5194/essd-12-2959-2020) Iturbide, M., Gutiérrez, J. M., Alves, L. M., Bedia, J., Cerezo-Mota, R., Cimadevilla, E., Cofiño, A. S., Di Luca, A., Faria, S. H., Gorodetskaya, I. V., Hauser, M., Herrera, S., Hennessy, K., Hewitt, H. T., Jones, R. G., Krakovska, S., Manzanas, R., Martínez-Castro, D., Narisma, G. T., Nurhati, I. S., Pinto, I., Seneviratne, S. I., van den Hurk, B., and Vera, C. S., 2020. An update of IPCC climate reference regions for subcontinental analysis of climate model data: definition and aggregated datasets, Earth Syst. Sci. Data, 12, 2959–2970, https://doi.org/10.5194/essd-12-2959-2020\n\n[[7]](https://doi.org/10.13031/2013.26773) Hargreaves, G. H., and Samani, Z. A., 1985. Reference crop evapotranspiration from temperature, Applied engineering in agriculture, pp. 96–99. https://doi.org/10.13031/2013.26773\n\n[[8]](https://doi.org/10.1002/joc.3887) Beguerı́a, S., Vicente-Serrano, S. M., Reig, F., and Latorre, B., 2014. Standardized precipitation evapotranspiration index (spei) revisited: Parameter fitting, evapotranspiration models, tools, datasets and drought monitoring, International Journal of Climatology, 34, 3001–3023. https://doi.org/10.1002/joc.3887", "text_with_prefix": "EQC Quality Assessment: \"CMIP6 biases in the SPEI6 drought index over the Mediterranean region\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q13 | Category: Climate_Projections\nSection: CMIP6 biases in the SPEI6 drought index over the Mediterranean region > ℹ️ If you want to know more > References\n---\n[[1]](https://doi.org/10.1175/2009JCLI2909.1) Vicente-Serrano, S. M., Beguerı́a, S., and López-Moreno, J. I., 2010. A multiscalar drought index sensitive to global warming: The standardized precipitation evapotranspiration index, Journal of Climate, 23, 1696–1718. https://doi.org/10.1175/2009JCLI2909.1\n\n[[2]](https://digitalcommons.unl.edu/droughtfacpub/69/) Wilhite, D. A., 2000. Droughts as a natural hazard: concepts and definitions. In: DROUGHT, A Global Assessment, vol. I and II, Routledge Hazards and Disasters Series, Routledge.\n\n[[3]](https://doi.org/10.1175/2012EI000434.1) Vicente-Serrano, S. M., Beguerı́a, S., Lorenzo-Lacruz, J., Camarero, J. J., López-Moreno, J. I., Azorin-Molina, C., Revuelto, J., Morán-Tejeda, E., and Sanchez-Lorenzo, A., 2012. Performance of drought indices for ecological, agricultural, and hydrological applications, Earth Interactions, 16, https://doi.org/10.1175/2012EI000434.1\n\n[[4]](https://doi.org/10.1016/j.scitotenv.2019.135245) Yao, N., Li, L., Feng, P., Feng, H., Liu, D. L., Liu, Y., Jiang, K., Hu, X., and Li, Y., 2020. Projections of drought characteristics in China based on a standardized precipitation and evapotranspiration index and multiple GCMs. Science of The Total Environment, 704, 135245. https://doi.org/10.1016/j.scitotenv.2019.135245\n\n[[5]](https://doi.org/10.3390/w17101507) Tuong, V. Q., Kiet, B. A., and Pham, T. T., 2025. Assessment of Future Drought Characteristics Using Various Temporal Scales and Multiple Drought Indices over Mekong Basin Under Climate Changes. Water, 17(10), 1507. https://doi.org/10.3390/w17101507\n\n[[6]](https://doi.org/10.5194/essd-12-2959-2020) Iturbide, M., Gutiérrez, J. M., Alves, L. M., Bedia, J., Cerezo-Mota, R., Cimadevilla, E., Cofiño, A. S., Di Luca, A., Faria, S. H., Gorodetskaya, I. V., Hauser, M., Herrera, S., Hennessy, K., Hewitt, H. T., Jones, R. G., Krakovska, S., Manzanas, R., Martínez-Castro, D., Narisma, G. T., Nurhati, I. S., Pinto, I., Seneviratne, S. I., van den Hurk, B., and Vera, C. S., 2020. An update of IPCC climate reference regions for subcontinental analysis of climate model data: definition and aggregated datasets, Earth Syst. Sci. Data, 12, 2959–2970, https://doi.org/10.5194/essd-12-2959-2020\n\n[[7]](https://doi.org/10.13031/2013.26773) Hargreaves, G. H., and Samani, Z. A., 1985. Reference crop evapotranspiration from temperature, Applied engineering in agriculture, pp. 96–99. https://doi.org/10.13031/2013.26773\n\n[[8]](https://doi.org/10.1002/joc.3887) Beguerı́a, S., Vicente-Serrano, S. M., Reig, F., and Latorre, B., 2014. Standardized precipitation evapotranspiration index (spei) revisited: Parameter fitting, evapotranspiration models, tools, datasets and drought monitoring, International Journal of Climatology, 34, 3001–3023. https://doi.org/10.1002/joc.3887"} {"chunk_id": "climate_projections-cmip6_validation_q13__5080104fdab0", "report_id": "climate_projections-cmip6_validation_q13", "dataset_id": "projections-cmip6", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q13", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CMIP6 biases in the SPEI6 drought index over the Mediterranean region > ℹ️ If you want to know more > References", "title": "CMIP6 biases in the SPEI6 drought index over the Mediterranean region", "chunk_index": 22, "token_count": 263, "text_raw": "revisited: Parameter fitting, evapotranspiration models, tools, datasets and drought monitoring, International Journal of Climatology, 34, 3001–3023. https://doi.org/10.1002/joc.3887\n\n[[9]](https://doi.org/10.5194/hess-25-273-2021) Schmith, T., Thejll, P., Berg, P., Boberg, F., Christensen, O. B., Christiansen, B., Christensen, J. H., Madsen, M. S., and Steger, C., 2021. Identifying robust bias adjustment methods for European extreme precipitation in a multi-model pseudo-reality setting, Hydrol. Earth Syst. Sci., 25, 273–290, https://doi.org/10.5194/hess-25-273-2021", "text_with_prefix": "EQC Quality Assessment: \"CMIP6 biases in the SPEI6 drought index over the Mediterranean region\"\nDataset: projections-cmip6 [CDS]\nAspect: validation_q13 | Category: Climate_Projections\nSection: CMIP6 biases in the SPEI6 drought index over the Mediterranean region > ℹ️ If you want to know more > References\n---\nrevisited: Parameter fitting, evapotranspiration models, tools, datasets and drought monitoring, International Journal of Climatology, 34, 3001–3023. https://doi.org/10.1002/joc.3887\n\n[[9]](https://doi.org/10.5194/hess-25-273-2021) Schmith, T., Thejll, P., Berg, P., Boberg, F., Christensen, O. B., Christiansen, B., Christensen, J. H., Madsen, M. S., and Steger, C., 2021. Identifying robust bias adjustment methods for European extreme precipitation in a multi-model pseudo-reality setting, Hydrol. Earth Syst. Sci., 25, 273–290, https://doi.org/10.5194/hess-25-273-2021"} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q01__040aef108a65", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q01", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q01", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Bias in extreme temperature indices for the reinsurance sector", "title": "Bias in extreme temperature indices for the reinsurance sector", "chunk_index": 0, "token_count": 100, "text_raw": "Production date: 29-05-2024\n\nProduced by: CMCC foundation - Euro-Mediterranean Center on Climate Change. Albert Martinez Boti.", "text_with_prefix": "EQC Quality Assessment: \"Bias in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q01 | Category: Climate_Projections\nSection: Bias in extreme temperature indices for the reinsurance sector\n---\nProduction date: 29-05-2024\n\nProduced by: CMCC foundation - Euro-Mediterranean Center on Climate Change. Albert Martinez Boti."} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q01__91d8e28c299e", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q01", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q01", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Bias in extreme temperature indices for the reinsurance sector > Quality assessment question", "title": "Bias in extreme temperature indices for the reinsurance sector", "chunk_index": 1, "token_count": 636, "text_raw": "* **How well do CORDEX projections represent climatology and trends of air temperature extremes in Europe?**\n* **Do I get different results than if I use CMIP6 projections?**\n\nClimate change has a major impact on the reinsurance market [[1]](https://doi.org/10.3390/atmos11020146)[[2]](https://doi.org/10.5194/nhess-22-659-2022). In the third assessment report of the IPCC, hot temperature extremes were already presented as relevant to insurance and related services [[3]](https://www.ipcc.ch/site/assets/uploads/2018/03/wg2TARchap8.pdf). Consequently, the need for reliable regional and global climate projections has become paramount, offering valuable insights for optimising reinsurance strategies in the face of a changing climate landscape. Nonetheless, despite their pivotal role, uncertainties inherent in these projections can potentially lead to misuse [[4]](https://doi.org/10.1002/wcc.71)[[5]](https://doi.org/10.1002/wcc.579). This underscores the importance of accurately calculating and accounting for uncertainties to ensure their appropriate consideration. This notebook utilises data from a subset of models from **[CORDEX](https://cds.climate.copernicus.eu/datasets/projections-cordex-domains-single-levels?tab=overview)** Regional Climate Models (RCMs) and compares them with [ERA5](https://cds.climate.copernicus.eu/datasets/reanalysis-era5-complete?tab=doc) reanalysis, serving as the reference product. Two maximum-temperature-based indices from [ECA&D](https://www.ecad.eu/indicesextremes/) indices (one of physical nature and the other of statistical nature) are computed using the [icclim](https://icclim.readthedocs.io/en/stable/) Python package. The first index, identified by the ETCCDI short name 'SU', quantifies the occurrence of summer days (i.e., with daily maximum temperatures exceeding 25°C) within a year or a season (JJA in this notebook). The second index, labeled 'TX90p', describes the number of days with daily maximum temperatures exceeding the daily 90th percentile of maximum temperature for a 5-day moving window. Within this notebook, these calculations are performed over the historical period spanning from 1971 to 2000. It is important to note that the results presented here pertain to a specific subset of the CORDEX ensemble and may not be generalisable to the entire dataset. Also note that a separate assessment examines the representation of trends of these indices for the same models during a fixed future period (2015-2099).", "text_with_prefix": "EQC Quality Assessment: \"Bias in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q01 | Category: Climate_Projections\nSection: Bias in extreme temperature indices for the reinsurance sector > Quality assessment question\n---\n* **How well do CORDEX projections represent climatology and trends of air temperature extremes in Europe?**\n* **Do I get different results than if I use CMIP6 projections?**\n\nClimate change has a major impact on the reinsurance market [[1]](https://doi.org/10.3390/atmos11020146)[[2]](https://doi.org/10.5194/nhess-22-659-2022). In the third assessment report of the IPCC, hot temperature extremes were already presented as relevant to insurance and related services [[3]](https://www.ipcc.ch/site/assets/uploads/2018/03/wg2TARchap8.pdf). Consequently, the need for reliable regional and global climate projections has become paramount, offering valuable insights for optimising reinsurance strategies in the face of a changing climate landscape. Nonetheless, despite their pivotal role, uncertainties inherent in these projections can potentially lead to misuse [[4]](https://doi.org/10.1002/wcc.71)[[5]](https://doi.org/10.1002/wcc.579). This underscores the importance of accurately calculating and accounting for uncertainties to ensure their appropriate consideration. This notebook utilises data from a subset of models from **[CORDEX](https://cds.climate.copernicus.eu/datasets/projections-cordex-domains-single-levels?tab=overview)** Regional Climate Models (RCMs) and compares them with [ERA5](https://cds.climate.copernicus.eu/datasets/reanalysis-era5-complete?tab=doc) reanalysis, serving as the reference product. Two maximum-temperature-based indices from [ECA&D](https://www.ecad.eu/indicesextremes/) indices (one of physical nature and the other of statistical nature) are computed using the [icclim](https://icclim.readthedocs.io/en/stable/) Python package. The first index, identified by the ETCCDI short name 'SU', quantifies the occurrence of summer days (i.e., with daily maximum temperatures exceeding 25°C) within a year or a season (JJA in this notebook). The second index, labeled 'TX90p', describes the number of days with daily maximum temperatures exceeding the daily 90th percentile of maximum temperature for a 5-day moving window. Within this notebook, these calculations are performed over the historical period spanning from 1971 to 2000. It is important to note that the results presented here pertain to a specific subset of the CORDEX ensemble and may not be generalisable to the entire dataset. Also note that a separate assessment examines the representation of trends of these indices for the same models during a fixed future period (2015-2099)."} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q01__d6668e652b3a", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q01", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q01", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Bias in extreme temperature indices for the reinsurance sector > Quality assessment statement", "title": "Bias in extreme temperature indices for the reinsurance sector", "chunk_index": 2, "token_count": 581, "text_raw": "These are the key outcomes of this assessment\n\n* During the JJA period and for the subset of regional models considered, CORDEX projections offer valuable insights into reproducing the climatology and trend of air temperature extremes across Europe for the historical period spanning from 1971 to 2000. These insights aid in the optimisation of reinsurance protections. Nevertheless, they are accompanied by uncertainties that necessitate careful consideration for informed decision-making. Users need to understand that biases affect the climatology and trend of the indices differently and that the choice of bias correction method should be tailored accordingly [[6]](https://doi.org/10.1002/asl.1072)[[7]](https://ibicus.readthedocs.io/en/latest/).\n\n* The subset of Regional Climate Models (RCMs) consistently show a tendency to underestimate trends for the 'SU' and 'TX90p' indices compared to ERA5 during the JJA period. This underestimation of the trend magnitude is even greater than observed for another CMIP6 assessment that is performed for the same period and seasonal aggregation (\"CMIP6 Climate Projections: evaluating bias in extreme temperature indices for the reinsurance sector.\"). While RCMs improve mean values compared to Global Climate Models (GCMs), they often struggle to replicate historical trends accurately, particularly underestimating in the Mediterranean basin. These findings underscore the importance of a comprehensive GCM-RCM matrix to address uncertainties effectively.\n\n* In contrast to climatological assessments, the consistency in trend bias patterns across diverse RCMs suggests a potential dependence on the driving GCM, shared among all RCMs. A more extensive GCM-RCM matrix is imperative to accurately quantify this dependency and delineate the partitioning of uncertainty between GCMs and RCMs [[8]](https://doi.org/10.1002/wcc.8)[[9]](https://doi.org/10.1088/1748-9326/aacc77).\n\n* The outcomes of this notebook also increase confidence (though they do not ensure accuracy) when analysing future trends using these models.\n```\n\nattachment:f986e53c-d903-459a-8a4a-02caea9644da.png\n---\nalt: Mean Bias SU CORDEX\nwidth: 900px\n---\nNumber of summer days ('SU') for the temporal aggregation of 'JJA'. Mean bias for the historical period (1971 - 2000) of each individual CORDEX model.\n```", "text_with_prefix": "EQC Quality Assessment: \"Bias in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q01 | Category: Climate_Projections\nSection: Bias in extreme temperature indices for the reinsurance sector > Quality assessment statement\n---\nThese are the key outcomes of this assessment\n\n* During the JJA period and for the subset of regional models considered, CORDEX projections offer valuable insights into reproducing the climatology and trend of air temperature extremes across Europe for the historical period spanning from 1971 to 2000. These insights aid in the optimisation of reinsurance protections. Nevertheless, they are accompanied by uncertainties that necessitate careful consideration for informed decision-making. Users need to understand that biases affect the climatology and trend of the indices differently and that the choice of bias correction method should be tailored accordingly [[6]](https://doi.org/10.1002/asl.1072)[[7]](https://ibicus.readthedocs.io/en/latest/).\n\n* The subset of Regional Climate Models (RCMs) consistently show a tendency to underestimate trends for the 'SU' and 'TX90p' indices compared to ERA5 during the JJA period. This underestimation of the trend magnitude is even greater than observed for another CMIP6 assessment that is performed for the same period and seasonal aggregation (\"CMIP6 Climate Projections: evaluating bias in extreme temperature indices for the reinsurance sector.\"). While RCMs improve mean values compared to Global Climate Models (GCMs), they often struggle to replicate historical trends accurately, particularly underestimating in the Mediterranean basin. These findings underscore the importance of a comprehensive GCM-RCM matrix to address uncertainties effectively.\n\n* In contrast to climatological assessments, the consistency in trend bias patterns across diverse RCMs suggests a potential dependence on the driving GCM, shared among all RCMs. A more extensive GCM-RCM matrix is imperative to accurately quantify this dependency and delineate the partitioning of uncertainty between GCMs and RCMs [[8]](https://doi.org/10.1002/wcc.8)[[9]](https://doi.org/10.1088/1748-9326/aacc77).\n\n* The outcomes of this notebook also increase confidence (though they do not ensure accuracy) when analysing future trends using these models.\n```\n\nattachment:f986e53c-d903-459a-8a4a-02caea9644da.png\n---\nalt: Mean Bias SU CORDEX\nwidth: 900px\n---\nNumber of summer days ('SU') for the temporal aggregation of 'JJA'. Mean bias for the historical period (1971 - 2000) of each individual CORDEX model.\n```"} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q01__5aaf4e1d1a69", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q01", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q01", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Bias in extreme temperature indices for the reinsurance sector > Methodology", "title": "Bias in extreme temperature indices for the reinsurance sector", "chunk_index": 3, "token_count": 400, "text_raw": "This notebook provides an assessment of the systematic errors (trend and climate mean) in a subset of 9 models from [CORDEX](https://cds.climate.copernicus.eu/datasets/projections-cordex-domains-single-levels?tab=overview). It achieves this by comparing the model predictions with the ERA5 reanalysis for the maximum-temperature-based indices of 'SU' and 'TX90p', calculated over the temporal aggregation of JJA and for the historical period spanning from 1971 to 2000. In particular, spatial patterns of climate mean and trend, along with biases, are examined and displayed for each model and the ensemble median (calculated for each grid cell). Additionally, spatially-averaged trend values are analysed and presented using box plots to provide an overview of trend behavior across the distribution of the chosen subset of models when averaged across Europe.\n\nThe analysis and results follow the next outline:\n\n**[](section-1)**\n * [](section-1.1)\n * [](section-1.2)\n * [](section-1.3)\n * [](section-1.4)\n * [](section-1.5)\n * [](section-1.6)\n\n**[](section-2)**\n * [](section-2.1)\n * [](section-2.2)\n * [](section-2.3)\n\n**[](section-3)**\n * [](section-3.1)\n * [](section-3.2)\n * [](section-3.3)\n * [](section-3.4)\n * [](section-3.5)\n * [](section-3.6)", "text_with_prefix": "EQC Quality Assessment: \"Bias in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q01 | Category: Climate_Projections\nSection: Bias in extreme temperature indices for the reinsurance sector > Methodology\n---\nThis notebook provides an assessment of the systematic errors (trend and climate mean) in a subset of 9 models from [CORDEX](https://cds.climate.copernicus.eu/datasets/projections-cordex-domains-single-levels?tab=overview). It achieves this by comparing the model predictions with the ERA5 reanalysis for the maximum-temperature-based indices of 'SU' and 'TX90p', calculated over the temporal aggregation of JJA and for the historical period spanning from 1971 to 2000. In particular, spatial patterns of climate mean and trend, along with biases, are examined and displayed for each model and the ensemble median (calculated for each grid cell). Additionally, spatially-averaged trend values are analysed and presented using box plots to provide an overview of trend behavior across the distribution of the chosen subset of models when averaged across Europe.\n\nThe analysis and results follow the next outline:\n\n**[](section-1)**\n * [](section-1.1)\n * [](section-1.2)\n * [](section-1.3)\n * [](section-1.4)\n * [](section-1.5)\n * [](section-1.6)\n\n**[](section-2)**\n * [](section-2.1)\n * [](section-2.2)\n * [](section-2.3)\n\n**[](section-3)**\n * [](section-3.1)\n * [](section-3.2)\n * [](section-3.3)\n * [](section-3.4)\n * [](section-3.5)\n * [](section-3.6)"} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q01__bfbd6e4cbdea", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q01", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q01", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Bias in extreme temperature indices for the reinsurance sector > Analysis and results > 1. Parameters, requests and functions definition > 1.2. Define Parameters", "title": "Bias in extreme temperature indices for the reinsurance sector", "chunk_index": 4, "token_count": 340, "text_raw": "In the \"Define Parameters\" section, various customisable options for the notebook are specified:\n\n- The initial and ending year used for the historical period can be specified by changing the parameters `year_start` and `year_stop` (1971-2000 is chosen for consistency between CORDEX and CMIP6).\n- The `timeseries` set the temporal aggregation. For instance, selecting \"JJA\" implies considering only the JJA season.\n- `collection_id` provides the choice between Global Climate Models CMIP6 or Regional Climate Models CORDEX. Although the code allows choosing between CMIP6 or CORDEX, the example provided in this notebook deals with CORDEX RCMs.\n- `area` allows specifying the geographical domain of interest.\n- The `interpolation_method` parameter allows selecting the interpolation method when regridding is performed over the indices.\n- The `chunk` selection allows the user to define if dividing into chunks when downloading the data on their local machine. Although it does not significantly affect the analysis, it is recommended to keep the default value for optimal performance.\n\nChoose annual or seasonal timeseries\nChoose CORDEX or CMIP6\nInterpolation method\nArea to show\nChunks for download\n\n(section-1.3)=", "text_with_prefix": "EQC Quality Assessment: \"Bias in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q01 | Category: Climate_Projections\nSection: Bias in extreme temperature indices for the reinsurance sector > Analysis and results > 1. Parameters, requests and functions definition > 1.2. Define Parameters\n---\nIn the \"Define Parameters\" section, various customisable options for the notebook are specified:\n\n- The initial and ending year used for the historical period can be specified by changing the parameters `year_start` and `year_stop` (1971-2000 is chosen for consistency between CORDEX and CMIP6).\n- The `timeseries` set the temporal aggregation. For instance, selecting \"JJA\" implies considering only the JJA season.\n- `collection_id` provides the choice between Global Climate Models CMIP6 or Regional Climate Models CORDEX. Although the code allows choosing between CMIP6 or CORDEX, the example provided in this notebook deals with CORDEX RCMs.\n- `area` allows specifying the geographical domain of interest.\n- The `interpolation_method` parameter allows selecting the interpolation method when regridding is performed over the indices.\n- The `chunk` selection allows the user to define if dividing into chunks when downloading the data on their local machine. Although it does not significantly affect the analysis, it is recommended to keep the default value for optimal performance.\n\nChoose annual or seasonal timeseries\nChoose CORDEX or CMIP6\nInterpolation method\nArea to show\nChunks for download\n\n(section-1.3)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q01__0082196599cf", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q01", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q01", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Bias in extreme temperature indices for the reinsurance sector > Analysis and results > 1. Parameters, requests and functions definition > 1.3. Define models", "title": "Bias in extreme temperature indices for the reinsurance sector", "chunk_index": 5, "token_count": 302, "text_raw": "The following climate analyses are performed considering a subset of GCMs from CORDEX. Models names are listed in the parameters below. Some variable-dependent parameters are also selected, as the `index_names` parameter, which specifies the temperature-based indices ('SU' and 'TX90p' in our case) from the [icclim](https://icclim.readthedocs.io/en/stable/) Python package.\n\nWhen choosing Cordex models, it is crucial to consider the availability of RCMs for the selected GCM and the specified region. The listed RCMs, for instance, are accessible for the GCM “mpi_m_mpi_esm_lr” in the “europe” cordex_domain. To confirm the available combinations, refer to the [CORDEX](https://cds.climate.copernicus.eu/datasets/projections-cordex-domains-single-levels?tab=overview) CDS catalogue entry.\n\nDefine dictionaries to use in titles and caption\nDefine dictionaries to use in titles and caption\n\n(section-1.4)=", "text_with_prefix": "EQC Quality Assessment: \"Bias in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q01 | Category: Climate_Projections\nSection: Bias in extreme temperature indices for the reinsurance sector > Analysis and results > 1. Parameters, requests and functions definition > 1.3. Define models\n---\nThe following climate analyses are performed considering a subset of GCMs from CORDEX. Models names are listed in the parameters below. Some variable-dependent parameters are also selected, as the `index_names` parameter, which specifies the temperature-based indices ('SU' and 'TX90p' in our case) from the [icclim](https://icclim.readthedocs.io/en/stable/) Python package.\n\nWhen choosing Cordex models, it is crucial to consider the availability of RCMs for the selected GCM and the specified region. The listed RCMs, for instance, are accessible for the GCM “mpi_m_mpi_esm_lr” in the “europe” cordex_domain. To confirm the available combinations, refer to the [CORDEX](https://cds.climate.copernicus.eu/datasets/projections-cordex-domains-single-levels?tab=overview) CDS catalogue entry.\n\nDefine dictionaries to use in titles and caption\nDefine dictionaries to use in titles and caption\n\n(section-1.4)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q01__4382fcb98ca8", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q01", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q01", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Bias in extreme temperature indices for the reinsurance sector > Analysis and results > 1. Parameters, requests and functions definition > 1.4. Define ERA5 request", "title": "Bias in extreme temperature indices for the reinsurance sector", "chunk_index": 6, "token_count": 127, "text_raw": "Within this notebook, ERA5 serves as the reference product. In this section, we set the required parameters for the cds-api data-request of ERA5.\n\n(section-1.5)=", "text_with_prefix": "EQC Quality Assessment: \"Bias in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q01 | Category: Climate_Projections\nSection: Bias in extreme temperature indices for the reinsurance sector > Analysis and results > 1. Parameters, requests and functions definition > 1.4. Define ERA5 request\n---\nWithin this notebook, ERA5 serves as the reference product. In this section, we set the required parameters for the cds-api data-request of ERA5.\n\n(section-1.5)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q01__331dc90b5296", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q01", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q01", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Bias in extreme temperature indices for the reinsurance sector > Analysis and results > 1. Parameters, requests and functions definition > 1.5. Define model requests", "title": "Bias in extreme temperature indices for the reinsurance sector", "chunk_index": 7, "token_count": 216, "text_raw": "In this section we set the required parameters for the cds-api data-request.\n\nThe `get_cordex_years` function is employed to choose suitable data chunks for CORDEX data requests.\n\nWhen `Weights = True`, spatial weighting is applied for calculations requiring spatial data aggregation. This is particularly relevant for CMIP6 GCMs with regular lon-lat grids that do not consider varying surface extensions at different latitudes. In contrast, CORDEX RCMs, using rotated grids, inherently account for different cell surfaces based on latitude, eliminating the need for a latitude cosine multiplicative factor (`Weights = False`).\n\n(section-1.6)=", "text_with_prefix": "EQC Quality Assessment: \"Bias in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q01 | Category: Climate_Projections\nSection: Bias in extreme temperature indices for the reinsurance sector > Analysis and results > 1. Parameters, requests and functions definition > 1.5. Define model requests\n---\nIn this section we set the required parameters for the cds-api data-request.\n\nThe `get_cordex_years` function is employed to choose suitable data chunks for CORDEX data requests.\n\nWhen `Weights = True`, spatial weighting is applied for calculations requiring spatial data aggregation. This is particularly relevant for CMIP6 GCMs with regular lon-lat grids that do not consider varying surface extensions at different latitudes. In contrast, CORDEX RCMs, using rotated grids, inherently account for different cell surfaces based on latitude, eliminating the need for a latitude cosine multiplicative factor (`Weights = False`).\n\n(section-1.6)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q01__052d6703af6f", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q01", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q01", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Bias in extreme temperature indices for the reinsurance sector > Analysis and results > 1. Parameters, requests and functions definition > 1.6. Functions to cache", "title": "Bias in extreme temperature indices for the reinsurance sector", "chunk_index": 8, "token_count": 320, "text_raw": "In this section, functions that will be executed in the caching phase are defined. Caching is the process of storing copies of files in a temporary storage location, so that they can be accessed more quickly. This process also checks if the user has already downloaded a file, avoiding redundant downloads.\n\nFunctions description:\n\n- The `select_timeseries` function subsets the dataset based on the chosen `timeseries` parameter.\n\n- The `compute_indices` function utilises the icclim package to calculate the selected maximum-temperature-based indices.\n\n- The `compute_trends` function employs the Mann-Kendall test for trend calculation.\n\n- Finally, the `compute_indices_and_trends` function selects the temporal aggregation using the `select_timeseries` function. It then computes daily maximum temperature (only if we are dealing with ERA5), calculates maximum-temperature-based indices via the `compute_indices` function, determines the indices mean over the historical period (1971-2000), obtain the trends using the `compute_trends` function, and offers an option for regridding to ERA5 if required.\n\nOriginal bounds for conservative interpolation\n\n(section-2)=", "text_with_prefix": "EQC Quality Assessment: \"Bias in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q01 | Category: Climate_Projections\nSection: Bias in extreme temperature indices for the reinsurance sector > Analysis and results > 1. Parameters, requests and functions definition > 1.6. Functions to cache\n---\nIn this section, functions that will be executed in the caching phase are defined. Caching is the process of storing copies of files in a temporary storage location, so that they can be accessed more quickly. This process also checks if the user has already downloaded a file, avoiding redundant downloads.\n\nFunctions description:\n\n- The `select_timeseries` function subsets the dataset based on the chosen `timeseries` parameter.\n\n- The `compute_indices` function utilises the icclim package to calculate the selected maximum-temperature-based indices.\n\n- The `compute_trends` function employs the Mann-Kendall test for trend calculation.\n\n- Finally, the `compute_indices_and_trends` function selects the temporal aggregation using the `select_timeseries` function. It then computes daily maximum temperature (only if we are dealing with ERA5), calculates maximum-temperature-based indices via the `compute_indices` function, determines the indices mean over the historical period (1971-2000), obtain the trends using the `compute_trends` function, and offers an option for regridding to ERA5 if required.\n\nOriginal bounds for conservative interpolation\n\n(section-2)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q01__4c22d1ad6549", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q01", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q01", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Bias in extreme temperature indices for the reinsurance sector > Analysis and results > 2. Downloading and processing > 2.1. Download and transform ERA5", "title": "Bias in extreme temperature indices for the reinsurance sector", "chunk_index": 9, "token_count": 186, "text_raw": "In this section, the `download.download_and_transform` function from the 'c3s_eqc_automatic_quality_control' package is employed to download ERA5 reference data, select the temporal aggregation (\"JJA\" in our example), compute daily maximum temperature from hourly data, compute the maximum-temperature-based indices, calculate the mean and trend for the historical period (1971-2000) and cache the result (to avoid redundant downloads and processing).\n\n(section-2.2)=", "text_with_prefix": "EQC Quality Assessment: \"Bias in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q01 | Category: Climate_Projections\nSection: Bias in extreme temperature indices for the reinsurance sector > Analysis and results > 2. Downloading and processing > 2.1. Download and transform ERA5\n---\nIn this section, the `download.download_and_transform` function from the 'c3s_eqc_automatic_quality_control' package is employed to download ERA5 reference data, select the temporal aggregation (\"JJA\" in our example), compute daily maximum temperature from hourly data, compute the maximum-temperature-based indices, calculate the mean and trend for the historical period (1971-2000) and cache the result (to avoid redundant downloads and processing).\n\n(section-2.2)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q01__02bf746eaa8e", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q01", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q01", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Bias in extreme temperature indices for the reinsurance sector > Analysis and results > 2. Downloading and processing > 2.2. Download and transform models", "title": "Bias in extreme temperature indices for the reinsurance sector", "chunk_index": 10, "token_count": 314, "text_raw": "In this section, the `download.download_and_transform` function from the 'c3s_eqc_automatic_quality_control' package is employed to download daily data from the CORDEX models, compute the maximum-temperature-based indices for the selected temporal aggregation, calculate the mean and trend over the historical period (1971-2005), interpolate to ERA5's grid (only for the cases in which it is specified, in the other cases, the original model's grid is mantained), and cache the result (to avoid redundant downloads and processing).\n\nOriginal model\nInterpolated model\n\n```text\nmodel='clmcom_clm_cclm4_8_17'\nmodel='clmcom_eth_cosmo_crclim'\nmodel='cnrm_aladin63'\nmodel='dmi_hirham5'\nmodel='knmi_racmo22e'\nmodel='mohc_hadrem3_ga7_05'\nmodel='mpi_csc_remo2009'\nmodel='smhi_rca4'\nmodel='uhoh_wrf361h'\n```\n\n(section-2.3)=", "text_with_prefix": "EQC Quality Assessment: \"Bias in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q01 | Category: Climate_Projections\nSection: Bias in extreme temperature indices for the reinsurance sector > Analysis and results > 2. Downloading and processing > 2.2. Download and transform models\n---\nIn this section, the `download.download_and_transform` function from the 'c3s_eqc_automatic_quality_control' package is employed to download daily data from the CORDEX models, compute the maximum-temperature-based indices for the selected temporal aggregation, calculate the mean and trend over the historical period (1971-2005), interpolate to ERA5's grid (only for the cases in which it is specified, in the other cases, the original model's grid is mantained), and cache the result (to avoid redundant downloads and processing).\n\nOriginal model\nInterpolated model\n\n```text\nmodel='clmcom_clm_cclm4_8_17'\nmodel='clmcom_eth_cosmo_crclim'\nmodel='cnrm_aladin63'\nmodel='dmi_hirham5'\nmodel='knmi_racmo22e'\nmodel='mohc_hadrem3_ga7_05'\nmodel='mpi_csc_remo2009'\nmodel='smhi_rca4'\nmodel='uhoh_wrf361h'\n```\n\n(section-2.3)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q01__7f450810a8f3", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q01", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q01", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Bias in extreme temperature indices for the reinsurance sector > Analysis and results > 2. Downloading and processing > 2.3. Apply land-sea mask, change attributes and cut the region to show", "title": "Bias in extreme temperature indices for the reinsurance sector", "chunk_index": 11, "token_count": 264, "text_raw": "This section performs the following tasks:\n\n1. Cut the region of interest.\n2. Downloads the sea mask for ERA5.\n3. Applies the sea mask to both ERA5 data and the model data, which were previously regridded to ERA5's grid (i.e., it applies the ERA5 sea mask to `ds_interpolated`).\n4. Regrids the ERA5 land-sea mask to the model's grid and applies it to them.\n5. Change some variable attributes for plotting purposes.\n\n**Note:** `ds_interpolated` contains data from the models (mean and trend over the historical period, p-value of the trends...) regridded to ERA5. `model_datasets` contain the same data but in the original grid of each model.\n\nEdit attributes\n\n(section-3)=", "text_with_prefix": "EQC Quality Assessment: \"Bias in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q01 | Category: Climate_Projections\nSection: Bias in extreme temperature indices for the reinsurance sector > Analysis and results > 2. Downloading and processing > 2.3. Apply land-sea mask, change attributes and cut the region to show\n---\nThis section performs the following tasks:\n\n1. Cut the region of interest.\n2. Downloads the sea mask for ERA5.\n3. Applies the sea mask to both ERA5 data and the model data, which were previously regridded to ERA5's grid (i.e., it applies the ERA5 sea mask to `ds_interpolated`).\n4. Regrids the ERA5 land-sea mask to the model's grid and applies it to them.\n5. Change some variable attributes for plotting purposes.\n\n**Note:** `ds_interpolated` contains data from the models (mean and trend over the historical period, p-value of the trends...) regridded to ERA5. `model_datasets` contain the same data but in the original grid of each model.\n\nEdit attributes\n\n(section-3)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q01__4d067af1225e", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q01", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q01", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Bias in extreme temperature indices for the reinsurance sector > Analysis and results > 3. Plot and describe results", "title": "Bias in extreme temperature indices for the reinsurance sector", "chunk_index": 12, "token_count": 316, "text_raw": "This section will display the following results:\n\n- Maps representing the spatial distribution of the **historical mean values** (1971-2000) of the 'SU' index for ERA5, each model individually, the ensemble median (understood as the median of the mean values of the chosen subset of models calculated for each grid cell), and the ensemble spread (derived as the standard deviation of the distribution of the chosen subset of models).\n- Maps representing the spatial distribution of the **historical trends** (1971-2000) of the indices 'SU' and 'TX90p'. Similar to the first analysis, this includes ERA5, each model individually, the ensemble median (understood as the median of the trend values of the chosen subset of models calculated for each grid cell), and the ensemble spread (derived as the standard deviation of the distribution of the chosen subset of models).\n- **Bias maps of the historical mean values**.\n- **Trend bias maps**. \n- **Boxplots** representing statistical distributions (PDF) built on the spatially-averaged historical trends from each considered model, displayed together with ERA5.\n\n(section-3.1)=", "text_with_prefix": "EQC Quality Assessment: \"Bias in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q01 | Category: Climate_Projections\nSection: Bias in extreme temperature indices for the reinsurance sector > Analysis and results > 3. Plot and describe results\n---\nThis section will display the following results:\n\n- Maps representing the spatial distribution of the **historical mean values** (1971-2000) of the 'SU' index for ERA5, each model individually, the ensemble median (understood as the median of the mean values of the chosen subset of models calculated for each grid cell), and the ensemble spread (derived as the standard deviation of the distribution of the chosen subset of models).\n- Maps representing the spatial distribution of the **historical trends** (1971-2000) of the indices 'SU' and 'TX90p'. Similar to the first analysis, this includes ERA5, each model individually, the ensemble median (understood as the median of the trend values of the chosen subset of models calculated for each grid cell), and the ensemble spread (derived as the standard deviation of the distribution of the chosen subset of models).\n- **Bias maps of the historical mean values**.\n- **Trend bias maps**. \n- **Boxplots** representing statistical distributions (PDF) built on the spatially-averaged historical trends from each considered model, displayed together with ERA5.\n\n(section-3.1)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q01__c36147a2c919", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q01", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q01", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Bias in extreme temperature indices for the reinsurance sector > Analysis and results > 3. Plot and describe results > 3.1. Define plotting functions", "title": "Bias in extreme temperature indices for the reinsurance sector", "chunk_index": 13, "token_count": 655, "text_raw": "The functions presented here are used to plot the mean values and trends calculated over the historical period (1971-2000) for each of the indices ('SU' and 'TX90p').\n\nFor a selected index, three layout types can be displayed, depending on the chosen function:\n\n1. Layout including the reference ERA5 product, the ensemble median, the bias of the ensemble median, and the ensemble spread: `plot_ensemble()` is used.\n2. Layout including every model (for the trend and mean values): `plot_models()` is employed.\n3. Layout including the bias of every model (for the trend and mean values): `plot_models()` is used again.\n\n`trend==True` argument allows displaying trend values over the historical period, while `trend==False` will show mean values. When the `trend` argument is set to `True`, regions with no significance are hatched. For individual models and ERA5, a grid point is considered to have a statistically significant trend when the p-value is lower than 0.05 (in such cases, no hatching is shown). However, for determining trend significance for the ensemble median (understood as the median of the trend values of the chosen subset of models calculated for each grid cell), reliance is placed on agreement categories, following the advanced approach proposed in AR6 [IPCC](https://www.ipcc.ch/report/ar6/wg1/downloads/report/IPCC_AR6_WGI_Atlas.pdf) on pages 1945-1950. The `hatch_p_value_ensemble()` function is used to distinguish, for each grid point, between three possible cases:\n\n1. If more than 66% of the models are statistically significant (p-value < 0.05) and more than 80% of the models share the same sign, we consider the ensemble median trend to be statistically significant, and there is agreement on the sign. To represent this, no hatching is used.\n2. If less than 66% of the models are statistically significant, regardless of agreement on the sign of the trend, hatching is applied (indicating that the ensemble median trend is not statistically significant).\n3. If more than 66% of the models are statistically significant but less than 80% of the models share the same sign, we consider the ensemble median trend to be statistically significant, but there is no agreement on the sign of the trend. This is represented using crosses.\n\nDefine function to plot the caption of the figures (for the ensemble case)\nAdd caption to the figure\nAdd each line to the figure\nend captioning\nDefine function to plot the caption of the figures (for the individual models case)\nAdd caption to the figure\nAdd each line to the figure\n\n(section-3.2)=", "text_with_prefix": "EQC Quality Assessment: \"Bias in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q01 | Category: Climate_Projections\nSection: Bias in extreme temperature indices for the reinsurance sector > Analysis and results > 3. Plot and describe results > 3.1. Define plotting functions\n---\nThe functions presented here are used to plot the mean values and trends calculated over the historical period (1971-2000) for each of the indices ('SU' and 'TX90p').\n\nFor a selected index, three layout types can be displayed, depending on the chosen function:\n\n1. Layout including the reference ERA5 product, the ensemble median, the bias of the ensemble median, and the ensemble spread: `plot_ensemble()` is used.\n2. Layout including every model (for the trend and mean values): `plot_models()` is employed.\n3. Layout including the bias of every model (for the trend and mean values): `plot_models()` is used again.\n\n`trend==True` argument allows displaying trend values over the historical period, while `trend==False` will show mean values. When the `trend` argument is set to `True`, regions with no significance are hatched. For individual models and ERA5, a grid point is considered to have a statistically significant trend when the p-value is lower than 0.05 (in such cases, no hatching is shown). However, for determining trend significance for the ensemble median (understood as the median of the trend values of the chosen subset of models calculated for each grid cell), reliance is placed on agreement categories, following the advanced approach proposed in AR6 [IPCC](https://www.ipcc.ch/report/ar6/wg1/downloads/report/IPCC_AR6_WGI_Atlas.pdf) on pages 1945-1950. The `hatch_p_value_ensemble()` function is used to distinguish, for each grid point, between three possible cases:\n\n1. If more than 66% of the models are statistically significant (p-value < 0.05) and more than 80% of the models share the same sign, we consider the ensemble median trend to be statistically significant, and there is agreement on the sign. To represent this, no hatching is used.\n2. If less than 66% of the models are statistically significant, regardless of agreement on the sign of the trend, hatching is applied (indicating that the ensemble median trend is not statistically significant).\n3. If more than 66% of the models are statistically significant but less than 80% of the models share the same sign, we consider the ensemble median trend to be statistically significant, but there is no agreement on the sign of the trend. This is represented using crosses.\n\nDefine function to plot the caption of the figures (for the ensemble case)\nAdd caption to the figure\nAdd each line to the figure\nend captioning\nDefine function to plot the caption of the figures (for the individual models case)\nAdd caption to the figure\nAdd each line to the figure\n\n(section-3.2)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q01__8d2a541d2a6d", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q01", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q01", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Bias in extreme temperature indices for the reinsurance sector > Analysis and results > 3. Plot and describe results > 3.2. Plot ensemble maps", "title": "Bias in extreme temperature indices for the reinsurance sector", "chunk_index": 14, "token_count": 369, "text_raw": "In this section, we invoke the `plot_ensemble()` function to visualise the mean values and trends calculated over the historical period (1971-2000) for the model ensemble and ERA5 reference product across Europe. Note that the model data used in this section has previously been interpolated to the ERA5 grid.\n\nSpecifically, for each of the indices ('SU' and 'TX90p'), this section presents two layouts:\n\n1. Only for the 'SU' index: mean values of the historical period (1971-2000) for: (a) the reference ERA5 product, (b) the ensemble median (understood as the median of the mean values of the chosen subset of models calculated for each grid cell), (c) the bias of the ensemble median, and (d) the ensemble spread (derived as the standard deviation of the distribution of the chosen subset of models).\n\n2. Trend values of the historical period (1971-2000) for: (a) the reference ERA5 product, (b) the ensemble median (understood as the median of the trend values of the chosen subset of models calculated for each grid cell), (c) the bias of the ensemble median, and (d) the ensemble spread (derived as the standard deviation of the distribution of the chosen subset of models).\n\nFig number counter\nCommon title\n\n(section-3.3)=", "text_with_prefix": "EQC Quality Assessment: \"Bias in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q01 | Category: Climate_Projections\nSection: Bias in extreme temperature indices for the reinsurance sector > Analysis and results > 3. Plot and describe results > 3.2. Plot ensemble maps\n---\nIn this section, we invoke the `plot_ensemble()` function to visualise the mean values and trends calculated over the historical period (1971-2000) for the model ensemble and ERA5 reference product across Europe. Note that the model data used in this section has previously been interpolated to the ERA5 grid.\n\nSpecifically, for each of the indices ('SU' and 'TX90p'), this section presents two layouts:\n\n1. Only for the 'SU' index: mean values of the historical period (1971-2000) for: (a) the reference ERA5 product, (b) the ensemble median (understood as the median of the mean values of the chosen subset of models calculated for each grid cell), (c) the bias of the ensemble median, and (d) the ensemble spread (derived as the standard deviation of the distribution of the chosen subset of models).\n\n2. Trend values of the historical period (1971-2000) for: (a) the reference ERA5 product, (b) the ensemble median (understood as the median of the trend values of the chosen subset of models calculated for each grid cell), (c) the bias of the ensemble median, and (d) the ensemble spread (derived as the standard deviation of the distribution of the chosen subset of models).\n\nFig number counter\nCommon title\n\n(section-3.3)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q01__600ae7a1d68e", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q01", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q01", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Bias in extreme temperature indices for the reinsurance sector > Analysis and results > 3. Plot and describe results > 3.3. Plot model maps", "title": "Bias in extreme temperature indices for the reinsurance sector", "chunk_index": 15, "token_count": 215, "text_raw": "In this section, we invoke the `plot_models()` function to visualise the mean values and trends calculated over the historical period (1971-2000) for every model individually across Europe. Note that the model data used in this section maintains its original grid.\n\nSpecifically, for each of the indices ('SU' and 'TX90p'), this section presents two layouts:\n\n1. A layout including the historical mean (1971-2000) of every model (only for the 'SU' index).\n\n2. A layout including the historical trend (1971-2000) of every model.\n\n(section-3.4)=", "text_with_prefix": "EQC Quality Assessment: \"Bias in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q01 | Category: Climate_Projections\nSection: Bias in extreme temperature indices for the reinsurance sector > Analysis and results > 3. Plot and describe results > 3.3. Plot model maps\n---\nIn this section, we invoke the `plot_models()` function to visualise the mean values and trends calculated over the historical period (1971-2000) for every model individually across Europe. Note that the model data used in this section maintains its original grid.\n\nSpecifically, for each of the indices ('SU' and 'TX90p'), this section presents two layouts:\n\n1. A layout including the historical mean (1971-2000) of every model (only for the 'SU' index).\n\n2. A layout including the historical trend (1971-2000) of every model.\n\n(section-3.4)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q01__15f2db66d3bb", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q01", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q01", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Bias in extreme temperature indices for the reinsurance sector > Analysis and results > 3. Plot and describe results > 3.4. Plot bias maps", "title": "Bias in extreme temperature indices for the reinsurance sector", "chunk_index": 16, "token_count": 229, "text_raw": "In this section, we invoke the `plot_models()` function to visualise the bias for the mean values and trends calculated over the historical period (1971-2000) for every model individually across Europe. Note that the model data used in this section has previously been interpolated to the ERA5 grid.\n\nSpecifically, for each of the indices ('SU' and 'TX90p'), this section presents two layouts:\n\n1. A layout including the bias for the historical mean (1971-2000) of every model (only for the 'SU' index).\n\n2. A layout including the bias for the historical trend (1971-2000) of every model.\n\n(section-3.5)=", "text_with_prefix": "EQC Quality Assessment: \"Bias in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q01 | Category: Climate_Projections\nSection: Bias in extreme temperature indices for the reinsurance sector > Analysis and results > 3. Plot and describe results > 3.4. Plot bias maps\n---\nIn this section, we invoke the `plot_models()` function to visualise the bias for the mean values and trends calculated over the historical period (1971-2000) for every model individually across Europe. Note that the model data used in this section has previously been interpolated to the ERA5 grid.\n\nSpecifically, for each of the indices ('SU' and 'TX90p'), this section presents two layouts:\n\n1. A layout including the bias for the historical mean (1971-2000) of every model (only for the 'SU' index).\n\n2. A layout including the bias for the historical trend (1971-2000) of every model.\n\n(section-3.5)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q01__3f663a28a47b", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q01", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q01", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Bias in extreme temperature indices for the reinsurance sector > Analysis and results > 3. Plot and describe results > 3.5. Boxplots of the historical trend", "title": "Bias in extreme temperature indices for the reinsurance sector", "chunk_index": 17, "token_count": 396, "text_raw": "In this last section, we compare the trends of the climate models with the reference trend from ERA5.\n\nDots represent the spatially-averaged historical trend over the selected region (change of the number of days per decade) for each model (grey), the ensemble mean (blue), and the reference product (orange). The ensemble median is shown as a green line. Note that the spatially averaged values are calculated for each model from its original grid (i.e., no interpolated data has been used here).\n\nThe boxplot visually illustrates the distribution of trends among the climate models, with the box covering the first quartile (Q1 = 25th percentile) to the third quartile (Q3 = 75th percentile), and a green line indicating the ensemble median (Q2 = 50th percentile). Whiskers extend from the edges of the box to show the full data range.\n\nEnsemble mean\n\n
\n
\n

Fig 10. Boxplots illustrating the historical trends of the distribution of the chosen subset of models and ERA5 for: (a) the 'SU' index and (b) the 'TX90p' index. The distribution is created by considering spatially averaged trends across Europe. The ensemble mean and the ensemble median trends are both included. Outliers in the distribution are denoted by a grey circle with a black contour.

\n
\n\n(section-3.6)=", "text_with_prefix": "EQC Quality Assessment: \"Bias in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q01 | Category: Climate_Projections\nSection: Bias in extreme temperature indices for the reinsurance sector > Analysis and results > 3. Plot and describe results > 3.5. Boxplots of the historical trend\n---\nIn this last section, we compare the trends of the climate models with the reference trend from ERA5.\n\nDots represent the spatially-averaged historical trend over the selected region (change of the number of days per decade) for each model (grey), the ensemble mean (blue), and the reference product (orange). The ensemble median is shown as a green line. Note that the spatially averaged values are calculated for each model from its original grid (i.e., no interpolated data has been used here).\n\nThe boxplot visually illustrates the distribution of trends among the climate models, with the box covering the first quartile (Q1 = 25th percentile) to the third quartile (Q3 = 75th percentile), and a green line indicating the ensemble median (Q2 = 50th percentile). Whiskers extend from the edges of the box to show the full data range.\n\nEnsemble mean\n\n
\n
\n

Fig 10. Boxplots illustrating the historical trends of the distribution of the chosen subset of models and ERA5 for: (a) the 'SU' index and (b) the 'TX90p' index. The distribution is created by considering spatially averaged trends across Europe. The ensemble mean and the ensemble median trends are both included. Outliers in the distribution are denoted by a grey circle with a black contour.

\n
\n\n(section-3.6)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q01__da10958cebc1", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q01", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q01", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Bias in extreme temperature indices for the reinsurance sector > Analysis and results > 3. Plot and describe results > 3.6. Results summary and discussion", "title": "Bias in extreme temperature indices for the reinsurance sector", "chunk_index": 18, "token_count": 679, "text_raw": "- Across the examined region, the selected subset of Regional Climate Models (RCMs) consistently tend to underestimate historical trends compared to ERA5 for the JJA temporal aggregation. Specifically, for the 'SU' index, the CORDEX ensemble median, as derived from boxplots, displays a trend of approximately 0.5 days per decade, contrasting with ERA5's ~2 days per decade. Similarly, the 'TX90p' index reveals around 0.6 days per decade for the CORDEX ensemble median compared to the ~2 days per decade observed in ERA5. The interquartile range spans from approximately 0.25 to 0.75 for the 'SU' index and from 0.5 to 0.75 for the 'TX90p' index. These outcomes align with findings from the assessment done for CMIP6 Global Climate Models (GCMs) but the magnitudes of the trends are more underestimaded.\n\n- While RCMs have demonstrated capability in enhancing GCMs' mean values for the considered indices, their efficacy in replicating historical trends appears limited. Analysing the spatial patterns of trend bias, the ensemble median of the CORDEX models results closely resemble those obtained in the CMIP6 exercise. Notably, the most significant underestimation of trends is evident in the Mediterranean basin.\n\n- In contrast to climatological assessments, the consistency in trend bias patterns across diverse RCMs suggests a potential dependence on the driving GCM, shared among all RCMs. A more extensive GCM-RCM matrix is imperative to accurately quantify this dependency and delineate the partitioning of uncertainty between GCMs and RCMs.\n\n- What do the results mean for users? Are the biases relevant?\n\n- For the JJA period and the subset of regional models considered, these outcomes aid in optimising reinsurance protections. However, they are accompanied by uncertainties that require careful consideration for informed decision-making. Users should be aware that biases impact the climatology and trend of the indices differently. In this case, the climatology of the indices appears to be better represented compared to another CMIP6 assessment that covers the same period and seasonal aggregation (\"CMIP6 Climate Projections: evaluating bias in extreme temperature indices for the reinsurance sector\"), but the trend is more underestimated than in that assessment. Therefore, users should carefully tailor their choice of bias correction method [[6]](https://doi.org/10.1002/asl.1072)[[7]](https://ibicus.readthedocs.io/en/latest/).\n\n\n
\nRESULTS NOTE:
\nIt is important to note that the results presented are specific to the 9 models chosen, and users should aim to assess as wide a range of models as possible before making a sub-selection.", "text_with_prefix": "EQC Quality Assessment: \"Bias in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q01 | Category: Climate_Projections\nSection: Bias in extreme temperature indices for the reinsurance sector > Analysis and results > 3. Plot and describe results > 3.6. Results summary and discussion\n---\n- Across the examined region, the selected subset of Regional Climate Models (RCMs) consistently tend to underestimate historical trends compared to ERA5 for the JJA temporal aggregation. Specifically, for the 'SU' index, the CORDEX ensemble median, as derived from boxplots, displays a trend of approximately 0.5 days per decade, contrasting with ERA5's ~2 days per decade. Similarly, the 'TX90p' index reveals around 0.6 days per decade for the CORDEX ensemble median compared to the ~2 days per decade observed in ERA5. The interquartile range spans from approximately 0.25 to 0.75 for the 'SU' index and from 0.5 to 0.75 for the 'TX90p' index. These outcomes align with findings from the assessment done for CMIP6 Global Climate Models (GCMs) but the magnitudes of the trends are more underestimaded.\n\n- While RCMs have demonstrated capability in enhancing GCMs' mean values for the considered indices, their efficacy in replicating historical trends appears limited. Analysing the spatial patterns of trend bias, the ensemble median of the CORDEX models results closely resemble those obtained in the CMIP6 exercise. Notably, the most significant underestimation of trends is evident in the Mediterranean basin.\n\n- In contrast to climatological assessments, the consistency in trend bias patterns across diverse RCMs suggests a potential dependence on the driving GCM, shared among all RCMs. A more extensive GCM-RCM matrix is imperative to accurately quantify this dependency and delineate the partitioning of uncertainty between GCMs and RCMs.\n\n- What do the results mean for users? Are the biases relevant?\n\n- For the JJA period and the subset of regional models considered, these outcomes aid in optimising reinsurance protections. However, they are accompanied by uncertainties that require careful consideration for informed decision-making. Users should be aware that biases impact the climatology and trend of the indices differently. In this case, the climatology of the indices appears to be better represented compared to another CMIP6 assessment that covers the same period and seasonal aggregation (\"CMIP6 Climate Projections: evaluating bias in extreme temperature indices for the reinsurance sector\"), but the trend is more underestimated than in that assessment. Therefore, users should carefully tailor their choice of bias correction method [[6]](https://doi.org/10.1002/asl.1072)[[7]](https://ibicus.readthedocs.io/en/latest/).\n\n\n
\nRESULTS NOTE:
\nIt is important to note that the results presented are specific to the 9 models chosen, and users should aim to assess as wide a range of models as possible before making a sub-selection."} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q01__2c17e21d330c", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q01", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q01", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Bias in extreme temperature indices for the reinsurance sector > ℹ️ If you want to know more > Key resources", "title": "Bias in extreme temperature indices for the reinsurance sector", "chunk_index": 19, "token_count": 287, "text_raw": "Some key resources and further reading were linked throughout this assessment.\n\nThe CDS catalogue entries for the data used were:\n* CORDEX regional climate model data on single levels (Daily mean - Maximum 2m temperature in the last 24 hours): https://cds.climate.copernicus.eu/datasets/projections-cordex-domains-single-levels?tab=overview\n* ERA5 hourly data on single levels from 1940 to present (2m temperature): https://cds.climate.copernicus.eu/datasets/reanalysis-era5-single-levels?tab=overview\n\nCode libraries used:\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)\n* [icclim](https://icclim.readthedocs.io/en/stable/) Python package", "text_with_prefix": "EQC Quality Assessment: \"Bias in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q01 | Category: Climate_Projections\nSection: Bias in extreme temperature indices for the reinsurance sector > ℹ️ If you want to know more > Key resources\n---\nSome key resources and further reading were linked throughout this assessment.\n\nThe CDS catalogue entries for the data used were:\n* CORDEX regional climate model data on single levels (Daily mean - Maximum 2m temperature in the last 24 hours): https://cds.climate.copernicus.eu/datasets/projections-cordex-domains-single-levels?tab=overview\n* ERA5 hourly data on single levels from 1940 to present (2m temperature): https://cds.climate.copernicus.eu/datasets/reanalysis-era5-single-levels?tab=overview\n\nCode libraries used:\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)\n* [icclim](https://icclim.readthedocs.io/en/stable/) Python package"} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q01__1a200b322a97", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q01", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q01", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Bias in extreme temperature indices for the reinsurance sector > ℹ️ If you want to know more > References", "title": "Bias in extreme temperature indices for the reinsurance sector", "chunk_index": 20, "token_count": 940, "text_raw": "[[1]](https://doi.org/10.3390/atmos11020146) Tesselaar, M., Botzen, W.J.W., Aerts, J.C.J.H. (2020). Impacts of Climate Change and Remote Natural Catastrophes on EU Flood Insurance Markets: An Analysis of Soft and Hard Reinsurance Markets for Flood Coverage. Atmosphere 2020, 11, 146. https://doi.org/10.3390/atmos11020146\n\n[[2]](https://doi.org/10.5194/nhess-22-659-2022) Rädler, A. T. (2022). Invited perspectives: how does climate change affect the risk of natural hazards? Challenges and step changes from the reinsurance perspective. Nat. Hazards Earth Syst. Sci., 22, 659–664. https://doi.org/10.5194/nhess-22-659-2022\n\n[[3]](https://www.ipcc.ch/site/assets/uploads/2018/03/wg2TARchap8.pdf) Vellinga, P., Mills, E., Bowers, L., Berz, G.A., Huq, S., Kozak, L.M., Paultikof, J., Schanzenbacker, B., Shida, S., Soler, G., Benson, C., Bidan, P., Bruce, J.W., Huyck, P.M., Lemcke, G., Peara, A., Radevsky, R., Schoubroeck, C.V., Dlugolecki, A.F. (2001). Insurance and other financial services. In J. J. McCarthy, O. F. Canziani, N. A. Leary, D. J. Dokken, & K. S. White (Eds.), Climate change 2001: impacts, adaptation, and vulnerability. Contribution of working group 2 to the third assessment report of the intergovernmental panel on climate change. (pp. 417-450). Cambridge University Press.\n\n[[4]](https://doi.org/10.1002/wcc.71) Lemos, M.C. and Rood, R.B. (2010). Climate projections and their impact on policy and practice. WIREs Clim Chg, 1: 670-682. https://doi.org/10.1002/wcc.71\n\n[[5]](https://doi.org/10.1002/wcc.579) Nissan H, Goddard L, de Perez EC, et al. (2019). On the use and misuse of climate change projections in international development. WIREs Clim Change, 10:e579. https://doi.org/10.1002/wcc.579\n\n[[6]](https://doi.org/10.1002/asl.1072) Iturbide, M., Casanueva, A., Bedia, J., Herrera, S., Milovac, J., Gutiérrez, J.M. (2021). On the need of bias adjustment for more plausible climate change projections of extreme heat. Atmos. Sci. Lett., 23, e1072. https://doi.org/10.1002/asl.1072\n\n[[7]](https://ibicus.readthedocs.io/en/latest/) https://ibicus.readthedocs.io/en/latest/\n\n[[8]](https://doi.org/10.1002/wcc.8) Rummukainen, M. (2010). State-of-the-art with regional climate models. WIREs Clim Change, 1: 82-96. https://doi.org/10.1002/wcc.8\n\n[[9]](https://doi.org/10.1088/1748-9326/aacc77) Silje Lund Sørland et al. (2018). Bias patterns and climate change signals in GCM-RCM model chains. Environ. Res. Lett. 13 074017. https://doi.org/10.1088/1748-9326/aacc77", "text_with_prefix": "EQC Quality Assessment: \"Bias in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q01 | Category: Climate_Projections\nSection: Bias in extreme temperature indices for the reinsurance sector > ℹ️ If you want to know more > References\n---\n[[1]](https://doi.org/10.3390/atmos11020146) Tesselaar, M., Botzen, W.J.W., Aerts, J.C.J.H. (2020). Impacts of Climate Change and Remote Natural Catastrophes on EU Flood Insurance Markets: An Analysis of Soft and Hard Reinsurance Markets for Flood Coverage. Atmosphere 2020, 11, 146. https://doi.org/10.3390/atmos11020146\n\n[[2]](https://doi.org/10.5194/nhess-22-659-2022) Rädler, A. T. (2022). Invited perspectives: how does climate change affect the risk of natural hazards? Challenges and step changes from the reinsurance perspective. Nat. Hazards Earth Syst. Sci., 22, 659–664. https://doi.org/10.5194/nhess-22-659-2022\n\n[[3]](https://www.ipcc.ch/site/assets/uploads/2018/03/wg2TARchap8.pdf) Vellinga, P., Mills, E., Bowers, L., Berz, G.A., Huq, S., Kozak, L.M., Paultikof, J., Schanzenbacker, B., Shida, S., Soler, G., Benson, C., Bidan, P., Bruce, J.W., Huyck, P.M., Lemcke, G., Peara, A., Radevsky, R., Schoubroeck, C.V., Dlugolecki, A.F. (2001). Insurance and other financial services. In J. J. McCarthy, O. F. Canziani, N. A. Leary, D. J. Dokken, & K. S. White (Eds.), Climate change 2001: impacts, adaptation, and vulnerability. Contribution of working group 2 to the third assessment report of the intergovernmental panel on climate change. (pp. 417-450). Cambridge University Press.\n\n[[4]](https://doi.org/10.1002/wcc.71) Lemos, M.C. and Rood, R.B. (2010). Climate projections and their impact on policy and practice. WIREs Clim Chg, 1: 670-682. https://doi.org/10.1002/wcc.71\n\n[[5]](https://doi.org/10.1002/wcc.579) Nissan H, Goddard L, de Perez EC, et al. (2019). On the use and misuse of climate change projections in international development. WIREs Clim Change, 10:e579. https://doi.org/10.1002/wcc.579\n\n[[6]](https://doi.org/10.1002/asl.1072) Iturbide, M., Casanueva, A., Bedia, J., Herrera, S., Milovac, J., Gutiérrez, J.M. (2021). On the need of bias adjustment for more plausible climate change projections of extreme heat. Atmos. Sci. Lett., 23, e1072. https://doi.org/10.1002/asl.1072\n\n[[7]](https://ibicus.readthedocs.io/en/latest/) https://ibicus.readthedocs.io/en/latest/\n\n[[8]](https://doi.org/10.1002/wcc.8) Rummukainen, M. (2010). State-of-the-art with regional climate models. WIREs Clim Change, 1: 82-96. https://doi.org/10.1002/wcc.8\n\n[[9]](https://doi.org/10.1088/1748-9326/aacc77) Silje Lund Sørland et al. (2018). Bias patterns and climate change signals in GCM-RCM model chains. Environ. Res. Lett. 13 074017. https://doi.org/10.1088/1748-9326/aacc77"} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q02__ab86df3153b8", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q02", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q02", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector", "title": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector", "chunk_index": 0, "token_count": 108, "text_raw": "Production date: 9-07-2024\n\nProduced by: CMCC foundation - Euro-Mediterranean Center on Climate Change. Albert Martinez Boti.", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q02 | Category: Climate_Projections\nSection: Uncertainty in projected changes in extreme temperature indices for the reinsurance sector\n---\nProduction date: 9-07-2024\n\nProduced by: CMCC foundation - Euro-Mediterranean Center on Climate Change. Albert Martinez Boti."} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q02__51a655590769", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q02", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q02", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector > Quality assessment question", "title": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector", "chunk_index": 1, "token_count": 667, "text_raw": "* **What are the projected future changes and associated uncertainties in air temperature extremes in Europe?**\n* **Do I get different results than if I use CMIP6 projections?**\n\nClimate change has a major impact on the reinsurance market [[1]](https://doi.org/10.3390/atmos11020146)[[2]](https://doi.org/10.5194/nhess-22-659-2022). In the third assessment report of the IPCC, hot temperature extremes were already presented as relevant to insurance and related services [[3]](https://www.ipcc.ch/site/assets/uploads/2018/03/wg2TARchap8.pdf). Consequently, the need for reliable regional and global climate projections has become paramount, offering valuable insights for optimising reinsurance strategies in the face of a changing climate landscape. Nonetheless, despite their pivotal role, uncertainties inherent in these projections can potentially lead to misuse [[4]](https://doi.org/10.1002/wcc.71)[[5]](https://doi.org/10.1002/wcc.579). This underscores the importance of accurately calculating and accounting for uncertainties to ensure their appropriate consideration. This notebook utilises data from a subset of models from **[CORDEX](https://cds.climate.copernicus.eu/datasets/projections-cordex-domains-single-levels?tab=overview)** Regional Climate Models (RCMs) and explores the uncertainty in future projections of maximum temperature-based extreme indices by considering the ensemble inter-model spread of projected changes. Two maximum temperature-based indices from [ECA&D](https://www.ecad.eu/indicesextremes/) indices (one of physical nature and the other of statistical nature) are computed using the [icclim](https://icclim.readthedocs.io/en/stable/) Python package. The first index, identified by the ETCCDI short name 'SU', quantifies the occurrence of summer days (i.e., with daily maximum temperatures exceeding 25°C) within a year or a season (JJA in this notebook). The second index, labeled 'TX90p', describes the number of days with daily maximum temperatures exceeding the daily 90th percentile of maximum temperature for a 5-day moving window. For this notebook, the daily 90th percentile threshold is calculated for the historical period spanning from 1971 to 2000. The index calculations, though, are performed over the future period from 2015 to 2099, following the Representative Concentration Pathway RCP 8.5. It is important to note that the results presented here pertain to a specific subset of the CORDEX ensemble and may not be generalisable to the entire dataset. Also note that a separate assessment examines the representation of climatology and trends of these indices for the same models during the historical period (1971-2000).", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q02 | Category: Climate_Projections\nSection: Uncertainty in projected changes in extreme temperature indices for the reinsurance sector > Quality assessment question\n---\n* **What are the projected future changes and associated uncertainties in air temperature extremes in Europe?**\n* **Do I get different results than if I use CMIP6 projections?**\n\nClimate change has a major impact on the reinsurance market [[1]](https://doi.org/10.3390/atmos11020146)[[2]](https://doi.org/10.5194/nhess-22-659-2022). In the third assessment report of the IPCC, hot temperature extremes were already presented as relevant to insurance and related services [[3]](https://www.ipcc.ch/site/assets/uploads/2018/03/wg2TARchap8.pdf). Consequently, the need for reliable regional and global climate projections has become paramount, offering valuable insights for optimising reinsurance strategies in the face of a changing climate landscape. Nonetheless, despite their pivotal role, uncertainties inherent in these projections can potentially lead to misuse [[4]](https://doi.org/10.1002/wcc.71)[[5]](https://doi.org/10.1002/wcc.579). This underscores the importance of accurately calculating and accounting for uncertainties to ensure their appropriate consideration. This notebook utilises data from a subset of models from **[CORDEX](https://cds.climate.copernicus.eu/datasets/projections-cordex-domains-single-levels?tab=overview)** Regional Climate Models (RCMs) and explores the uncertainty in future projections of maximum temperature-based extreme indices by considering the ensemble inter-model spread of projected changes. Two maximum temperature-based indices from [ECA&D](https://www.ecad.eu/indicesextremes/) indices (one of physical nature and the other of statistical nature) are computed using the [icclim](https://icclim.readthedocs.io/en/stable/) Python package. The first index, identified by the ETCCDI short name 'SU', quantifies the occurrence of summer days (i.e., with daily maximum temperatures exceeding 25°C) within a year or a season (JJA in this notebook). The second index, labeled 'TX90p', describes the number of days with daily maximum temperatures exceeding the daily 90th percentile of maximum temperature for a 5-day moving window. For this notebook, the daily 90th percentile threshold is calculated for the historical period spanning from 1971 to 2000. The index calculations, though, are performed over the future period from 2015 to 2099, following the Representative Concentration Pathway RCP 8.5. It is important to note that the results presented here pertain to a specific subset of the CORDEX ensemble and may not be generalisable to the entire dataset. Also note that a separate assessment examines the representation of climatology and trends of these indices for the same models during the historical period (1971-2000)."} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q02__aed6324bd4d7", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q02", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q02", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector > Quality assessment statement", "title": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector", "chunk_index": 2, "token_count": 670, "text_raw": "These are the key outcomes of this assessment\n\n* All models within the considered subset agree on projecting positive future trends (2015-2099) for both indices across Europe during the temporal aggregation of JJA, with a particularly notable positive projected trend in the Mediterranean Basin for 'TX90p'. This finding is consistent with the results of Josep Cos et al. (2022) [[6]](https://doi.org/10.5194/esd-13-321-2022), who evaluated the Mediterranean climate change hotspot using CMIP6 projections. While certain regions exhibit near-zero trends for the 'SU' index, possibly due to threshold temperature constraints, others show higher values, highlighting the importance of considering both statistically and physically based extreme indices for comprehensive assessments. In fact this arise limitations of the 'SU' index, indicating the potential need to select a higher (or lower for northern areas) threshold to capture the changes in these regions. Such regions, where the changes may be just as impactful or even more so than areas with a significant increase in days above 25 degrees, require careful consideration.\n\n* The magnitude of the trend is smaller and shows more spatial variability than in another CMIP6 future trend assessment (\"Uncertainty in projected changes in extreme temperature indices for the reinsurance sector\").\n\n* The findings underscore the need for proactive measures to optimise reinsurance protections in the face of projected increases in air temperature extremes for the future period spanning from 2015 to 2099. While all considered models show a positive trend for these indices, the spatial variation in the magnitude of these trends and their uncertainty, quantified by the inter-model spread, underscores the need for careful consideration.\n\n* A larger GCM-RCM matrix should be considered when addressing specific cases to enhance the robustness of the analysis and account for uncertainties [[7]](https://doi.org/10.1002/wcc.8)[[8]](https://doi.org/10.1088/1748-9326/aacc77)\n```\n\nattachment:6243ed90-3ae0-4340-bff3-5af1cead6e6b.png\n---\nalt: trend_future_TX90p_CORDEX\nwidth: 900px\n---\nNumber of days with daily maximum temperatures exceeding the daily 90th percentile of maximum temperature for a 5-day moving window ('TX90p') for the temporal aggregation of 'JJA'. Trend for the future period (2015-2099). For this index, the reference daily 90th percentile threshold is calculated based on the historical period (1971-2000). The layout includes data corresponding to: (a) the ensemble median (understood as the median of the trend values of the chosen subset of models calculated for each grid cell) and (b) the ensemble spread (derived as the standard deviation of the distribution of the chosen subset of models).\n```", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q02 | Category: Climate_Projections\nSection: Uncertainty in projected changes in extreme temperature indices for the reinsurance sector > Quality assessment statement\n---\nThese are the key outcomes of this assessment\n\n* All models within the considered subset agree on projecting positive future trends (2015-2099) for both indices across Europe during the temporal aggregation of JJA, with a particularly notable positive projected trend in the Mediterranean Basin for 'TX90p'. This finding is consistent with the results of Josep Cos et al. (2022) [[6]](https://doi.org/10.5194/esd-13-321-2022), who evaluated the Mediterranean climate change hotspot using CMIP6 projections. While certain regions exhibit near-zero trends for the 'SU' index, possibly due to threshold temperature constraints, others show higher values, highlighting the importance of considering both statistically and physically based extreme indices for comprehensive assessments. In fact this arise limitations of the 'SU' index, indicating the potential need to select a higher (or lower for northern areas) threshold to capture the changes in these regions. Such regions, where the changes may be just as impactful or even more so than areas with a significant increase in days above 25 degrees, require careful consideration.\n\n* The magnitude of the trend is smaller and shows more spatial variability than in another CMIP6 future trend assessment (\"Uncertainty in projected changes in extreme temperature indices for the reinsurance sector\").\n\n* The findings underscore the need for proactive measures to optimise reinsurance protections in the face of projected increases in air temperature extremes for the future period spanning from 2015 to 2099. While all considered models show a positive trend for these indices, the spatial variation in the magnitude of these trends and their uncertainty, quantified by the inter-model spread, underscores the need for careful consideration.\n\n* A larger GCM-RCM matrix should be considered when addressing specific cases to enhance the robustness of the analysis and account for uncertainties [[7]](https://doi.org/10.1002/wcc.8)[[8]](https://doi.org/10.1088/1748-9326/aacc77)\n```\n\nattachment:6243ed90-3ae0-4340-bff3-5af1cead6e6b.png\n---\nalt: trend_future_TX90p_CORDEX\nwidth: 900px\n---\nNumber of days with daily maximum temperatures exceeding the daily 90th percentile of maximum temperature for a 5-day moving window ('TX90p') for the temporal aggregation of 'JJA'. Trend for the future period (2015-2099). For this index, the reference daily 90th percentile threshold is calculated based on the historical period (1971-2000). The layout includes data corresponding to: (a) the ensemble median (understood as the median of the trend values of the chosen subset of models calculated for each grid cell) and (b) the ensemble spread (derived as the standard deviation of the distribution of the chosen subset of models).\n```"} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q02__dc1fa23067b1", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q02", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q02", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector > Methodology", "title": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector", "chunk_index": 3, "token_count": 404, "text_raw": "This notebook offers an assessment of the projected changes and their associated uncertainties using a subset of 9 models from [CORDEX](https://cds.climate.copernicus.eu/datasets/projections-cordex-domains-single-levels?tab=overview). The uncertainty is examined by analysing the ensemble inter-model spread of projected changes for the maximum-temperature-based indices 'SU' and 'TX90p,' calculated over the temporal aggregation of JJA for the future period spanning from 2015 to 2099. In particular, spatial patterns of climate projected trends are examined and displayed for each model individually and for the ensemble median (calculated for each grid cell), alongside the ensemble inter-model spread to account for projected uncertainty. Additionally, spatially-averaged trend values are analysed and presented using box plots to provide an overview of trend behavior across the distribution of the chosen subset of models when averaged across Europe.\n\nThe analysis and results follow the next outline:\n\n**[](section-1)**\n * [](section-1.1)\n * [](section-1.2)\n * [](section-1.3)\n * [](section-1.4)\n * [](section-1.5)\n * [](section-1.6)\n\n**[](section-2)**\n * [](section-2.1)\n * [](section-2.2)\n * [](section-2.3)\n\n**[](section-3)**\n * [](section-3.1)\n * [](section-3.2)\n * [](section-3.3)\n * [](section-3.4)\n * [](section-3.5)", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q02 | Category: Climate_Projections\nSection: Uncertainty in projected changes in extreme temperature indices for the reinsurance sector > Methodology\n---\nThis notebook offers an assessment of the projected changes and their associated uncertainties using a subset of 9 models from [CORDEX](https://cds.climate.copernicus.eu/datasets/projections-cordex-domains-single-levels?tab=overview). The uncertainty is examined by analysing the ensemble inter-model spread of projected changes for the maximum-temperature-based indices 'SU' and 'TX90p,' calculated over the temporal aggregation of JJA for the future period spanning from 2015 to 2099. In particular, spatial patterns of climate projected trends are examined and displayed for each model individually and for the ensemble median (calculated for each grid cell), alongside the ensemble inter-model spread to account for projected uncertainty. Additionally, spatially-averaged trend values are analysed and presented using box plots to provide an overview of trend behavior across the distribution of the chosen subset of models when averaged across Europe.\n\nThe analysis and results follow the next outline:\n\n**[](section-1)**\n * [](section-1.1)\n * [](section-1.2)\n * [](section-1.3)\n * [](section-1.4)\n * [](section-1.5)\n * [](section-1.6)\n\n**[](section-2)**\n * [](section-2.1)\n * [](section-2.2)\n * [](section-2.3)\n\n**[](section-3)**\n * [](section-3.1)\n * [](section-3.2)\n * [](section-3.3)\n * [](section-3.4)\n * [](section-3.5)"} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q02__9f3e20f5452d", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q02", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q02", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector > Analysis and results > 1. Parameters, requests and functions definition > 1.2. Define Parameters", "title": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector", "chunk_index": 4, "token_count": 422, "text_raw": "In the \"Define Parameters\" section, various customisable options for the notebook are specified. Most of the parameters chosen are the same as those used in other assessments ([](./climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q01.ipynb)), being them: \n- The initial and ending year used for the future projections period can be specified by changing the parametes `future_slice` (2015-2099 is chosen for consistency between CORDEX and CMIP6).\n- `historical_slice` determines the historical period used (1971 to 2000 is choosen to allow comparison to CORDEX models in other assessments).\n- The `timeseries` set the temporal aggregation. For instance, selecting \"DJF\" implies considering only the winter season.\n- `collection_id` provides the choice between Global Climate Models CMIP6 or Regional Climate Models CORDEX. Although the code allows choosing between CMIP6 or CORDEX, the example provided in this notebook deals with CORDEX RCMs.\n- `area` allows specifying the geographical domain of interest.\n- The `interpolation_method` parameter allows selecting the interpolation method when regridding is performed over the indices.\n- The `chunk` selection allows the user to define if dividing into chunks when downloading the data on their local machine. Although it does not significantly affect the analysis, it is recommended to keep the default value for optimal performance.\n\nChoose annual or seasonal timeseries\nChoose CORDEX or CMIP6\nInterpolation method\nArea to show\nChunks for download\n\n(section-1.3)=", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q02 | Category: Climate_Projections\nSection: Uncertainty in projected changes in extreme temperature indices for the reinsurance sector > Analysis and results > 1. Parameters, requests and functions definition > 1.2. Define Parameters\n---\nIn the \"Define Parameters\" section, various customisable options for the notebook are specified. Most of the parameters chosen are the same as those used in other assessments ([](./climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q01.ipynb)), being them: \n- The initial and ending year used for the future projections period can be specified by changing the parametes `future_slice` (2015-2099 is chosen for consistency between CORDEX and CMIP6).\n- `historical_slice` determines the historical period used (1971 to 2000 is choosen to allow comparison to CORDEX models in other assessments).\n- The `timeseries` set the temporal aggregation. For instance, selecting \"DJF\" implies considering only the winter season.\n- `collection_id` provides the choice between Global Climate Models CMIP6 or Regional Climate Models CORDEX. Although the code allows choosing between CMIP6 or CORDEX, the example provided in this notebook deals with CORDEX RCMs.\n- `area` allows specifying the geographical domain of interest.\n- The `interpolation_method` parameter allows selecting the interpolation method when regridding is performed over the indices.\n- The `chunk` selection allows the user to define if dividing into chunks when downloading the data on their local machine. Although it does not significantly affect the analysis, it is recommended to keep the default value for optimal performance.\n\nChoose annual or seasonal timeseries\nChoose CORDEX or CMIP6\nInterpolation method\nArea to show\nChunks for download\n\n(section-1.3)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q02__51edd394adf1", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q02", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q02", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector > Analysis and results > 1. Parameters, requests and functions definition > 1.3. Define models", "title": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector", "chunk_index": 5, "token_count": 343, "text_raw": "The following climate analyses are performed considering a subset of GCMs from CMIP6. Models names are listed in the parameters below. Some variable-dependent parameters are also selected, as the `index_names` parameter, which specifies the maximum-temperature-based indices ('SU' and 'TX90p' in our case) from the [icclim](https://icclim.readthedocs.io/en/stable/) Python package.\n\nWhen choosing Cordex models, it is crucial to consider the availability of RCMs for the selected GCM and the specified region. The listed RCMs, for instance, are accessible for the GCM “mpi_m_mpi_esm_lr” in the “europe” cordex_domain, and they are the same as those used in other assessments (\"CORDEX Climate Projections: evaluating bias in extreme temperature indices for the reinsurance sector\"). To confirm the available combinations, refer to the [CORDEX](https://cds.climate.copernicus.eu/datasets/projections-cordex-domains-single-levels?tab=overview) CDS catalogue entry.\n\nDefine dictionaries to use in titles and caption\nDefine dictionaries to use in titles and caption\n\n(section-1.4)=", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q02 | Category: Climate_Projections\nSection: Uncertainty in projected changes in extreme temperature indices for the reinsurance sector > Analysis and results > 1. Parameters, requests and functions definition > 1.3. Define models\n---\nThe following climate analyses are performed considering a subset of GCMs from CMIP6. Models names are listed in the parameters below. Some variable-dependent parameters are also selected, as the `index_names` parameter, which specifies the maximum-temperature-based indices ('SU' and 'TX90p' in our case) from the [icclim](https://icclim.readthedocs.io/en/stable/) Python package.\n\nWhen choosing Cordex models, it is crucial to consider the availability of RCMs for the selected GCM and the specified region. The listed RCMs, for instance, are accessible for the GCM “mpi_m_mpi_esm_lr” in the “europe” cordex_domain, and they are the same as those used in other assessments (\"CORDEX Climate Projections: evaluating bias in extreme temperature indices for the reinsurance sector\"). To confirm the available combinations, refer to the [CORDEX](https://cds.climate.copernicus.eu/datasets/projections-cordex-domains-single-levels?tab=overview) CDS catalogue entry.\n\nDefine dictionaries to use in titles and caption\nDefine dictionaries to use in titles and caption\n\n(section-1.4)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q02__297c3d340e6c", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q02", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q02", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector > Analysis and results > 1. Parameters, requests and functions definition > 1.4. Define land-sea mask request", "title": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector", "chunk_index": 6, "token_count": 148, "text_raw": "Within this notebook, ERA5 will be used to download the land-sea mask when plotting. In this section, we set the required parameters for the cds-api data-request of ERA5 land-sea mask.\n\n(section-1.5)=", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q02 | Category: Climate_Projections\nSection: Uncertainty in projected changes in extreme temperature indices for the reinsurance sector > Analysis and results > 1. Parameters, requests and functions definition > 1.4. Define land-sea mask request\n---\nWithin this notebook, ERA5 will be used to download the land-sea mask when plotting. In this section, we set the required parameters for the cds-api data-request of ERA5 land-sea mask.\n\n(section-1.5)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q02__331dc90b5296", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q02", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q02", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector > Analysis and results > 1. Parameters, requests and functions definition > 1.5. Define model requests", "title": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector", "chunk_index": 7, "token_count": 224, "text_raw": "In this section we set the required parameters for the cds-api data-request.\n\nThe `get_cordex_years` function is employed to choose suitable data chunks for CORDEX data requests.\n\nWhen `Weights = True`, spatial weighting is applied for calculations requiring spatial data aggregation. This is particularly relevant for CMIP6 GCMs with regular lon-lat grids that do not consider varying surface extensions at different latitudes. In contrast, CORDEX RCMs, using rotated grids, inherently account for different cell surfaces based on latitude, eliminating the need for a latitude cosine multiplicative factor (`Weights = False`).\n\n(section-1.6)=", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q02 | Category: Climate_Projections\nSection: Uncertainty in projected changes in extreme temperature indices for the reinsurance sector > Analysis and results > 1. Parameters, requests and functions definition > 1.5. Define model requests\n---\nIn this section we set the required parameters for the cds-api data-request.\n\nThe `get_cordex_years` function is employed to choose suitable data chunks for CORDEX data requests.\n\nWhen `Weights = True`, spatial weighting is applied for calculations requiring spatial data aggregation. This is particularly relevant for CMIP6 GCMs with regular lon-lat grids that do not consider varying surface extensions at different latitudes. In contrast, CORDEX RCMs, using rotated grids, inherently account for different cell surfaces based on latitude, eliminating the need for a latitude cosine multiplicative factor (`Weights = False`).\n\n(section-1.6)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q02__e0c4e82aabcf", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q02", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q02", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector > Analysis and results > 1. Parameters, requests and functions definition > 1.6. Functions to cache", "title": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector", "chunk_index": 8, "token_count": 304, "text_raw": "In this section, functions that will be executed in the caching phase are defined. Caching is the process of storing copies of files in a temporary storage location, so that they can be accessed more quickly. This process also checks if the user has already downloaded a file, avoiding redundant downloads.\n\nFunctions description:\n\n- The `select_timeseries` function subsets the dataset based on the chosen `timeseries` parameter.\n\n- The `compute_indices` function utilises the icclim package to calculate the maximum-temperature-based indices.\n\n- The `compute_trends` function employs the Mann-Kendall test for trend calculation.\n\n- Finally, the `compute_indices_and_trends` function calculates the maximum-temperature-based indices for the corresponding temporal aggregation using the `compute_indices` function, determines the indices mean for the future period (2015-2099), obtain the trends using the `compute_trends` function, and offers an option for regridding to `model_regrid`.\n\nOriginal bounds for conservative interpolation\n\n(section-2)=", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q02 | Category: Climate_Projections\nSection: Uncertainty in projected changes in extreme temperature indices for the reinsurance sector > Analysis and results > 1. Parameters, requests and functions definition > 1.6. Functions to cache\n---\nIn this section, functions that will be executed in the caching phase are defined. Caching is the process of storing copies of files in a temporary storage location, so that they can be accessed more quickly. This process also checks if the user has already downloaded a file, avoiding redundant downloads.\n\nFunctions description:\n\n- The `select_timeseries` function subsets the dataset based on the chosen `timeseries` parameter.\n\n- The `compute_indices` function utilises the icclim package to calculate the maximum-temperature-based indices.\n\n- The `compute_trends` function employs the Mann-Kendall test for trend calculation.\n\n- Finally, the `compute_indices_and_trends` function calculates the maximum-temperature-based indices for the corresponding temporal aggregation using the `compute_indices` function, determines the indices mean for the future period (2015-2099), obtain the trends using the `compute_trends` function, and offers an option for regridding to `model_regrid`.\n\nOriginal bounds for conservative interpolation\n\n(section-2)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q02__0eb2dadb7042", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q02", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q02", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector > Analysis and results > 2. Downloading and processing > 2.1. Download and transform the regridding model", "title": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector", "chunk_index": 9, "token_count": 294, "text_raw": "In this section, the `download.download_and_transform` function from the 'c3s_eqc_automatic_quality_control' package is employed to download daily data from the selected CORDEX regridding model, compute the maximum-temperature-based indices for the selected temporal aggregation, calculate the mean and trend over the future projections period (2015-2099), and cache the result (to avoid redundant downloads and processing).\n\nThe regridding model is intended here as the model whose grid will be used to interpolate the others. This ensures all models share a common grid, facilitating the calculation of median values for each cell point. The regridding model within this notebook is \"clmcom_eth_cosmo_crclim\" but a different one can be selected by just modifying the `model_regrid` parameter at [](section-1.3). It is key to highlight the importance of the chosen target grid depending on the specific application.\n\n(section-2.2)=", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q02 | Category: Climate_Projections\nSection: Uncertainty in projected changes in extreme temperature indices for the reinsurance sector > Analysis and results > 2. Downloading and processing > 2.1. Download and transform the regridding model\n---\nIn this section, the `download.download_and_transform` function from the 'c3s_eqc_automatic_quality_control' package is employed to download daily data from the selected CORDEX regridding model, compute the maximum-temperature-based indices for the selected temporal aggregation, calculate the mean and trend over the future projections period (2015-2099), and cache the result (to avoid redundant downloads and processing).\n\nThe regridding model is intended here as the model whose grid will be used to interpolate the others. This ensures all models share a common grid, facilitating the calculation of median values for each cell point. The regridding model within this notebook is \"clmcom_eth_cosmo_crclim\" but a different one can be selected by just modifying the `model_regrid` parameter at [](section-1.3). It is key to highlight the importance of the chosen target grid depending on the specific application.\n\n(section-2.2)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q02__f81aa4d6f6b2", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q02", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q02", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector > Analysis and results > 2. Downloading and processing > 2.2. Download and transform models", "title": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector", "chunk_index": 10, "token_count": 326, "text_raw": "In this section, the `download.download_and_transform` function from the 'c3s_eqc_automatic_quality_control' package is employed to download daily data from the CORDDEX models, compute the maximum-temperature-based indices for the selected temporal aggregation, calculate the mean and trend over the future period (2015-2099), interpolate to the regridding model's grid (only for the cases in which it is specified, in the other cases, the original model's grid is mantained), and cache the result (to avoid redundant downloads and processing).\n\nOriginal model\nInterpolated model\n\n```text\nmodel='clmcom_clm_cclm4_8_17'\nmodel='clmcom_eth_cosmo_crclim'\nmodel='cnrm_aladin63'\nmodel='dmi_hirham5'\nmodel='knmi_racmo22e'\nmodel='mohc_hadrem3_ga7_05'\nmodel='mpi_csc_remo2009'\nmodel='smhi_rca4'\nmodel='uhoh_wrf361h'\n```\n\n(section-2.3)=", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q02 | Category: Climate_Projections\nSection: Uncertainty in projected changes in extreme temperature indices for the reinsurance sector > Analysis and results > 2. Downloading and processing > 2.2. Download and transform models\n---\nIn this section, the `download.download_and_transform` function from the 'c3s_eqc_automatic_quality_control' package is employed to download daily data from the CORDDEX models, compute the maximum-temperature-based indices for the selected temporal aggregation, calculate the mean and trend over the future period (2015-2099), interpolate to the regridding model's grid (only for the cases in which it is specified, in the other cases, the original model's grid is mantained), and cache the result (to avoid redundant downloads and processing).\n\nOriginal model\nInterpolated model\n\n```text\nmodel='clmcom_clm_cclm4_8_17'\nmodel='clmcom_eth_cosmo_crclim'\nmodel='cnrm_aladin63'\nmodel='dmi_hirham5'\nmodel='knmi_racmo22e'\nmodel='mohc_hadrem3_ga7_05'\nmodel='mpi_csc_remo2009'\nmodel='smhi_rca4'\nmodel='uhoh_wrf361h'\n```\n\n(section-2.3)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q02__010e9ce99bfa", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q02", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q02", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector > Analysis and results > 2. Downloading and processing > 2.3. Apply land-sea mask, change attributes and cut the region to show", "title": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector", "chunk_index": 11, "token_count": 244, "text_raw": "This section performs the following tasks:\n\n1. Cut the region of interest.\n2. Downloads the sea mask for ERA5.\n3. Regrids ERA5's mask to the `model_regrid` grid and applies it to the regridded data\n4. Regrids the ERA5 land-sea mask to the model's original grid and applies it to them.\n5. Change some variable attributes for plotting purposes.\n\n**Note:** `ds_interpolated` contains data from the models regridded to the regridding model's grid. `model_datasets` contain the same data but in the original grid of each model.\n\nEdit attributes\n\n(section-3)=", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q02 | Category: Climate_Projections\nSection: Uncertainty in projected changes in extreme temperature indices for the reinsurance sector > Analysis and results > 2. Downloading and processing > 2.3. Apply land-sea mask, change attributes and cut the region to show\n---\nThis section performs the following tasks:\n\n1. Cut the region of interest.\n2. Downloads the sea mask for ERA5.\n3. Regrids ERA5's mask to the `model_regrid` grid and applies it to the regridded data\n4. Regrids the ERA5 land-sea mask to the model's original grid and applies it to them.\n5. Change some variable attributes for plotting purposes.\n\n**Note:** `ds_interpolated` contains data from the models regridded to the regridding model's grid. `model_datasets` contain the same data but in the original grid of each model.\n\nEdit attributes\n\n(section-3)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q02__a3feb9e7aa59", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q02", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q02", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector > Analysis and results > 3. Plot and describe results", "title": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector", "chunk_index": 12, "token_count": 208, "text_raw": "This section will display the following results:\n\n- Maps representing the spatial distribution of the **future trends** (2015-2099) of the indices 'SU' and 'TX90p' for each model individually, the ensemble median (understood as the median of the trend values of the chosen subset of models calculated for each grid cell), and the ensemble spread (derived as the standard deviation of the distribution of the chosen subset of models).\n- **Boxplots** which represent statistical distributions (PDF) built on the spatially-averaged future trend from each considered model.\n\n(section-3.1)=", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q02 | Category: Climate_Projections\nSection: Uncertainty in projected changes in extreme temperature indices for the reinsurance sector > Analysis and results > 3. Plot and describe results\n---\nThis section will display the following results:\n\n- Maps representing the spatial distribution of the **future trends** (2015-2099) of the indices 'SU' and 'TX90p' for each model individually, the ensemble median (understood as the median of the trend values of the chosen subset of models calculated for each grid cell), and the ensemble spread (derived as the standard deviation of the distribution of the chosen subset of models).\n- **Boxplots** which represent statistical distributions (PDF) built on the spatially-averaged future trend from each considered model.\n\n(section-3.1)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q02__79050a3cec57", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q02", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q02", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector > Analysis and results > 3. Plot and describe results > 3.1. Define plotting functions", "title": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector", "chunk_index": 13, "token_count": 640, "text_raw": "The functions presented here are used to plot the trends calculated over the future period (2015-2099) for each of the indices ('SU' and 'TX90p').\n\nFor a selected index, two layout types will be displayed, depending on the chosen function:\n\n1. Layout including the ensemble median and the ensemble spread for the trend: `plot_ensemble()` is used.\n2. Layout including every model trend: `plot_models()` is employed.\n\n`trend==True` allows displaying trend values over the future period, while `trend==False` show mean values. In this notebook, which focuses on the future period, only trend values will be shown, and, consequently, `trend==True`. When the `trend` argument is set to True, regions with no significance are hatched. For individual models, a grid point is considered to have a statistically significant trend when the p-value is lower than 0.05 (in such cases, no hatching is shown). However, for determining trend significance for the ensemble median (understood as the median of the trend values of the chosen subset of models calculated for each grid cell), reliance is placed on agreement categories, following the advanced approach proposed in AR6 [IPCC](https://www.ipcc.ch/report/ar6/wg1/downloads/report/IPCC_AR6_WGI_Atlas.pdf) on pages 1945-1950. The `hatch_p_value_ensemble()` function is used to distinguish, for each grid point, between three possible cases:\n\n1. If more than 66% of the models are statistically significant (p-value < 0.05) and more than 80% of the models share the same sign, we consider the ensemble median trend to be statistically significant, and there is agreement on the sign. To represent this, no hatching is used.\n2. If less than 66% of the models are statistically significant, regardless of agreement on the sign of the trend, hatching is applied (indicating that the ensemble median trend is not statistically significant).\n3. If more than 66% of the models are statistically significant but less than 80% of the models share the same sign, we consider the ensemble median trend to be statistically significant, but there is no agreement on the sign of the trend. This is represented using crosses.\n\nDefine function to plot the caption of the figures (for the ensemble case)\nAdd caption to the figure\nAdd each line to the figure\nend captioning\nDefine function to plot the caption of the figures (for the individual models case)\nAdd caption to the figure\nAdd each line to the figure\n\n(section-3.2)=", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q02 | Category: Climate_Projections\nSection: Uncertainty in projected changes in extreme temperature indices for the reinsurance sector > Analysis and results > 3. Plot and describe results > 3.1. Define plotting functions\n---\nThe functions presented here are used to plot the trends calculated over the future period (2015-2099) for each of the indices ('SU' and 'TX90p').\n\nFor a selected index, two layout types will be displayed, depending on the chosen function:\n\n1. Layout including the ensemble median and the ensemble spread for the trend: `plot_ensemble()` is used.\n2. Layout including every model trend: `plot_models()` is employed.\n\n`trend==True` allows displaying trend values over the future period, while `trend==False` show mean values. In this notebook, which focuses on the future period, only trend values will be shown, and, consequently, `trend==True`. When the `trend` argument is set to True, regions with no significance are hatched. For individual models, a grid point is considered to have a statistically significant trend when the p-value is lower than 0.05 (in such cases, no hatching is shown). However, for determining trend significance for the ensemble median (understood as the median of the trend values of the chosen subset of models calculated for each grid cell), reliance is placed on agreement categories, following the advanced approach proposed in AR6 [IPCC](https://www.ipcc.ch/report/ar6/wg1/downloads/report/IPCC_AR6_WGI_Atlas.pdf) on pages 1945-1950. The `hatch_p_value_ensemble()` function is used to distinguish, for each grid point, between three possible cases:\n\n1. If more than 66% of the models are statistically significant (p-value < 0.05) and more than 80% of the models share the same sign, we consider the ensemble median trend to be statistically significant, and there is agreement on the sign. To represent this, no hatching is used.\n2. If less than 66% of the models are statistically significant, regardless of agreement on the sign of the trend, hatching is applied (indicating that the ensemble median trend is not statistically significant).\n3. If more than 66% of the models are statistically significant but less than 80% of the models share the same sign, we consider the ensemble median trend to be statistically significant, but there is no agreement on the sign of the trend. This is represented using crosses.\n\nDefine function to plot the caption of the figures (for the ensemble case)\nAdd caption to the figure\nAdd each line to the figure\nend captioning\nDefine function to plot the caption of the figures (for the individual models case)\nAdd caption to the figure\nAdd each line to the figure\n\n(section-3.2)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q02__273bbe4e4bcc", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q02", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q02", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector > Analysis and results > 3. Plot and describe results > 3.2. Plot ensemble maps", "title": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector", "chunk_index": 14, "token_count": 273, "text_raw": "In this section, we invoke the `plot_ensemble()` function to visualise the trend calculated over the future period (2015-2099) for the model ensemble across Europe. Note that the model data used in this section has previously been interpolated to the \"regridding model\" grid (`\"clmcom_eth_cosmo_crclim\"` for this notebook).\n\nSpecifically, for each of the indices ('SU' and 'TX90p'), this section presents a single layout including trend values of the future period (2015-2099) for: (a) the ensemble median (understood as the median of the trend values of the chosen subset of models calculated for each grid cell) and (b) the ensemble spread (derived as the standard deviation of the distribution of the chosen subset of models).\n\nFig number counter\nCommon title\n\n(section-3.3)=", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q02 | Category: Climate_Projections\nSection: Uncertainty in projected changes in extreme temperature indices for the reinsurance sector > Analysis and results > 3. Plot and describe results > 3.2. Plot ensemble maps\n---\nIn this section, we invoke the `plot_ensemble()` function to visualise the trend calculated over the future period (2015-2099) for the model ensemble across Europe. Note that the model data used in this section has previously been interpolated to the \"regridding model\" grid (`\"clmcom_eth_cosmo_crclim\"` for this notebook).\n\nSpecifically, for each of the indices ('SU' and 'TX90p'), this section presents a single layout including trend values of the future period (2015-2099) for: (a) the ensemble median (understood as the median of the trend values of the chosen subset of models calculated for each grid cell) and (b) the ensemble spread (derived as the standard deviation of the distribution of the chosen subset of models).\n\nFig number counter\nCommon title\n\n(section-3.3)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q02__1c649ab53e56", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q02", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q02", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector > Analysis and results > 3. Plot and describe results > 3.3. Plot model maps", "title": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector", "chunk_index": 15, "token_count": 189, "text_raw": "In this section, we invoke the `plot_models()` function to visualise the trend over the future period (2015-2099) for every model individually. Note that the model data used in this section maintains its original grid.\n\nSpecifically, for each of the indices ('SU' and 'TX90p'), this section presents a single layout including the trend for the future period (2015-2099) of every model.\n\n(section-3.4)=", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q02 | Category: Climate_Projections\nSection: Uncertainty in projected changes in extreme temperature indices for the reinsurance sector > Analysis and results > 3. Plot and describe results > 3.3. Plot model maps\n---\nIn this section, we invoke the `plot_models()` function to visualise the trend over the future period (2015-2099) for every model individually. Note that the model data used in this section maintains its original grid.\n\nSpecifically, for each of the indices ('SU' and 'TX90p'), this section presents a single layout including the trend for the future period (2015-2099) of every model.\n\n(section-3.4)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q02__d0b788350d24", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q02", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q02", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector > Analysis and results > 3. Plot and describe results > 3.4. Boxplots of the future trend", "title": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector", "chunk_index": 16, "token_count": 398, "text_raw": "Finally, we present boxplots representing the ensemble distribution of each climate model trend calculated over the future period (2015-2099) across Europe.\n\nDots represent the spatially-averaged future trend over the selected region (change of the number of days per decade) for each model (grey) and the ensemble mean (blue). The ensemble median is shown as a green line. Note that the spatially averaged values are calculated for each model from its original grid (i.e., no interpolated data has been used here).\n\nThe boxplot visually illustrates the distribution of trends among the climate models, with the box covering the first quartile (Q1 = 25th percentile) to the third quartile (Q3 = 75th percentile), and a green line indicating the ensemble median (Q2 = 50th percentile). Whiskers extend from the edges of the box to show the full data range.\n\nEnsemble mean\n\n
\n
\n

Fig 5. Boxplots illustrating the future trends of the ensemble distribution for: (a) the 'SU' index and (b) the 'TX90p' index. The distribution is created by considering spatially averaged trends across Europe. The ensemble mean and the ensemble median trends are both included. Outliers in the distribution are denoted by a grey circle with a black contour.

\n
\n\n(section-3.5)=", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q02 | Category: Climate_Projections\nSection: Uncertainty in projected changes in extreme temperature indices for the reinsurance sector > Analysis and results > 3. Plot and describe results > 3.4. Boxplots of the future trend\n---\nFinally, we present boxplots representing the ensemble distribution of each climate model trend calculated over the future period (2015-2099) across Europe.\n\nDots represent the spatially-averaged future trend over the selected region (change of the number of days per decade) for each model (grey) and the ensemble mean (blue). The ensemble median is shown as a green line. Note that the spatially averaged values are calculated for each model from its original grid (i.e., no interpolated data has been used here).\n\nThe boxplot visually illustrates the distribution of trends among the climate models, with the box covering the first quartile (Q1 = 25th percentile) to the third quartile (Q3 = 75th percentile), and a green line indicating the ensemble median (Q2 = 50th percentile). Whiskers extend from the edges of the box to show the full data range.\n\nEnsemble mean\n\n
\n
\n

Fig 5. Boxplots illustrating the future trends of the ensemble distribution for: (a) the 'SU' index and (b) the 'TX90p' index. The distribution is created by considering spatially averaged trends across Europe. The ensemble mean and the ensemble median trends are both included. Outliers in the distribution are denoted by a grey circle with a black contour.

\n
\n\n(section-3.5)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q02__3be655a0e6a5", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q02", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q02", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector > Analysis and results > 3. Plot and describe results > 3.5. Results summary and discussion", "title": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector", "chunk_index": 17, "token_count": 778, "text_raw": "- For the temporal aggregation of JJA, future trends (2015-2099) exhibit positive values over Europe for both considered indices. For the number of summer days index (SU), the southern part of the Mediterranean Basin and the northernmost regions of Europe have trends near 0. This may be due to the fact that, for the future period, the number of summer days remains constant near 0 for the northern parts of Europe (where the threshold temperature of 25°C may be too high to be reached). Meanwhile, for the southern part of the Mediterranean Basin, this phenomenon may be attributed to the fact that the threshold is already surpassed for the entire JJA season. This emphasises the necessity to consider both statistically and physically based extreme indices for a proper assessment.\n\n- Regional differences can also be observed for the 'TX90p' index. Notably, values are higher for the Mediterranean Basin, consistent with the findings of [[6]](https://doi.org/10.5194/esd-13-321-2022), who assessed the Mediterranean climate change hotspot using CMIP6 projections. These regional differences appear to be greater than those observed for the CMIP6 future trend analysis of the same index (\"CMIP6 Climate Projections: evaluating uncertainty in projected changes in extreme temperature indices for the reinsurance sector\").\n\n- The boxplots show spatially averaged positive trends over Europe. For the 'SU' index, the ensemble median trends reach a value of 1.4 days per decade. For the 'TX90p' index, the ensemble median is nearly 5 days per decade. The interquantile range of the ensemble spans from 1.3 to 1.7 days per decade for the 'SU' index and from 4.5 to more than 5.2 days per decade for the 'TX90p' index. As for the historical analyses, the spatially-averaged trend values are lower than those obtained for the CMIP6 exercise.\n\n- What do the results mean for users? Are the biases relevant?\n\n- The findings underscore the need for proactive measures to optimise reinsurance protections in the face of projected increases in air temperature extremes for the future period spanning from 2015 to 2099. While all considered models show a positive trend for these indices, the spatial variation in the magnitude of these trends and their uncertainty, quantified by the inter-model spread, underscores the need for careful consideration.\n\n- It is also important to point out that the magnitude of the trend is lower than that obtained in another CMIP6 assessment that covers the same period and seasonal aggregation (\"CMIP6 Climate Projections: evaluating uncertainty in projected changes in extreme temperature indices for the reinsurance sector\").\n\n- Finally, it should also be taken into account the biases of the trend obtained for a CORDEX assessment (\"CORDEX Climate Projections: evaluating bias in extreme temperature indices for the reinsurance sector\"), which showed an important underestimation for the trend during the historical period (1971-2000).\n\n\n
\nRESULTS NOTE:
\nIt is important to note that the results presented are specific to the 9 models chosen, and users should aim to assess as wide a range of models as possible before making a sub-selection.", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q02 | Category: Climate_Projections\nSection: Uncertainty in projected changes in extreme temperature indices for the reinsurance sector > Analysis and results > 3. Plot and describe results > 3.5. Results summary and discussion\n---\n- For the temporal aggregation of JJA, future trends (2015-2099) exhibit positive values over Europe for both considered indices. For the number of summer days index (SU), the southern part of the Mediterranean Basin and the northernmost regions of Europe have trends near 0. This may be due to the fact that, for the future period, the number of summer days remains constant near 0 for the northern parts of Europe (where the threshold temperature of 25°C may be too high to be reached). Meanwhile, for the southern part of the Mediterranean Basin, this phenomenon may be attributed to the fact that the threshold is already surpassed for the entire JJA season. This emphasises the necessity to consider both statistically and physically based extreme indices for a proper assessment.\n\n- Regional differences can also be observed for the 'TX90p' index. Notably, values are higher for the Mediterranean Basin, consistent with the findings of [[6]](https://doi.org/10.5194/esd-13-321-2022), who assessed the Mediterranean climate change hotspot using CMIP6 projections. These regional differences appear to be greater than those observed for the CMIP6 future trend analysis of the same index (\"CMIP6 Climate Projections: evaluating uncertainty in projected changes in extreme temperature indices for the reinsurance sector\").\n\n- The boxplots show spatially averaged positive trends over Europe. For the 'SU' index, the ensemble median trends reach a value of 1.4 days per decade. For the 'TX90p' index, the ensemble median is nearly 5 days per decade. The interquantile range of the ensemble spans from 1.3 to 1.7 days per decade for the 'SU' index and from 4.5 to more than 5.2 days per decade for the 'TX90p' index. As for the historical analyses, the spatially-averaged trend values are lower than those obtained for the CMIP6 exercise.\n\n- What do the results mean for users? Are the biases relevant?\n\n- The findings underscore the need for proactive measures to optimise reinsurance protections in the face of projected increases in air temperature extremes for the future period spanning from 2015 to 2099. While all considered models show a positive trend for these indices, the spatial variation in the magnitude of these trends and their uncertainty, quantified by the inter-model spread, underscores the need for careful consideration.\n\n- It is also important to point out that the magnitude of the trend is lower than that obtained in another CMIP6 assessment that covers the same period and seasonal aggregation (\"CMIP6 Climate Projections: evaluating uncertainty in projected changes in extreme temperature indices for the reinsurance sector\").\n\n- Finally, it should also be taken into account the biases of the trend obtained for a CORDEX assessment (\"CORDEX Climate Projections: evaluating bias in extreme temperature indices for the reinsurance sector\"), which showed an important underestimation for the trend during the historical period (1971-2000).\n\n\n
\nRESULTS NOTE:
\nIt is important to note that the results presented are specific to the 9 models chosen, and users should aim to assess as wide a range of models as possible before making a sub-selection."} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q02__0f3898c23f1b", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q02", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q02", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector > ℹ️ If you want to know more > Key resources", "title": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector", "chunk_index": 18, "token_count": 251, "text_raw": "Some key resources and further reading were linked throughout this assessment.\n\nThe CDS catalogue entries for the data used were:\n* CORDEX regional climate model data on single levels (Daily mean - Maximum 2m temperature in the last 24 hours): https://cds.climate.copernicus.eu/datasets/projections-cordex-domains-single-levels?tab=overview\n\nCode libraries used:\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)\n* [icclim](https://icclim.readthedocs.io/en/stable/) Python package", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q02 | Category: Climate_Projections\nSection: Uncertainty in projected changes in extreme temperature indices for the reinsurance sector > ℹ️ If you want to know more > Key resources\n---\nSome key resources and further reading were linked throughout this assessment.\n\nThe CDS catalogue entries for the data used were:\n* CORDEX regional climate model data on single levels (Daily mean - Maximum 2m temperature in the last 24 hours): https://cds.climate.copernicus.eu/datasets/projections-cordex-domains-single-levels?tab=overview\n\nCode libraries used:\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)\n* [icclim](https://icclim.readthedocs.io/en/stable/) Python package"} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q02__c9f6180ac2a0", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q02", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q02", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector > ℹ️ If you want to know more > References", "title": "Uncertainty in projected changes in extreme temperature indices for the reinsurance sector", "chunk_index": 19, "token_count": 930, "text_raw": "[[1]](https://doi.org/10.3390/atmos11020146) Tesselaar, M., Botzen, W.J.W., Aerts, J.C.J.H. (2020). Impacts of Climate Change and Remote Natural Catastrophes on EU Flood Insurance Markets: An Analysis of Soft and Hard Reinsurance Markets for Flood Coverage. Atmosphere 2020, 11, 146. https://doi.org/10.3390/atmos11020146\n\n[[2]](https://doi.org/10.5194/nhess-22-659-2022) Rädler, A. T. (2022). Invited perspectives: how does climate change affect the risk of natural hazards? Challenges and step changes from the reinsurance perspective. Nat. Hazards Earth Syst. Sci., 22, 659–664. https://doi.org/10.5194/nhess-22-659-2022\n\n[[3]](https://www.ipcc.ch/site/assets/uploads/2018/03/wg2TARchap8.pdf) Vellinga, P., Mills, E., Bowers, L., Berz, G.A., Huq, S., Kozak, L.M., Paultikof, J., Schanzenbacker, B., Shida, S., Soler, G., Benson, C., Bidan, P., Bruce, J.W., Huyck, P.M., Lemcke, G., Peara, A., Radevsky, R., Schoubroeck, C.V., Dlugolecki, A.F. (2001). Insurance and other financial services. In J. J. McCarthy, O. F. Canziani, N. A. Leary, D. J. Dokken, & K. S. White (Eds.), Climate change 2001: impacts, adaptation, and vulnerability. Contribution of working group 2 to the third assessment report of the intergovernmental panel on climate change. (pp. 417-450). Cambridge University Press.\n\n[[4]](https://doi.org/10.1002/wcc.71) Lemos, M.C. and Rood, R.B. (2010). Climate projections and their impact on policy and practice. WIREs Clim Chg, 1: 670-682. https://doi.org/10.1002/wcc.71\n\n[[5]](https://doi.org/10.1002/wcc.579) Nissan H, Goddard L, de Perez EC, et al. (2019). On the use and misuse of climate change projections in international development. WIREs Clim Change, 10:e579. https://doi.org/10.1002/wcc.579\n\n[[6]](https://doi.org/10.5194/esd-13-321-2022) Cos, J., Doblas-Reyes, F., Jury, M., Marcos, R., Bretonnière, P.-A., and Samsó, M. (2022). The Mediterranean climate change hotspot in the CMIP5 and CMIP6 projections, Earth Syst. Dynam., 13, 321–340. https://doi.org/10.5194/esd-13-321-2022\n\n[[7]](https://doi.org/10.1002/wcc.8) Rummukainen, M. (2010). State-of-the-art with regional climate models. WIREs Clim Change, 1: 82-96. https://doi.org/10.1002/wcc.8\n\n[[8]](https://doi.org/10.1088/1748-9326/aacc77) Silje Lund Sørland et al. (2018). Bias patterns and climate change signals in GCM-RCM model chains. Environ. Res. Lett. 13 074017. https://doi.org/10.1088/1748-9326/aacc77", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty in projected changes in extreme temperature indices for the reinsurance sector\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q02 | Category: Climate_Projections\nSection: Uncertainty in projected changes in extreme temperature indices for the reinsurance sector > ℹ️ If you want to know more > References\n---\n[[1]](https://doi.org/10.3390/atmos11020146) Tesselaar, M., Botzen, W.J.W., Aerts, J.C.J.H. (2020). Impacts of Climate Change and Remote Natural Catastrophes on EU Flood Insurance Markets: An Analysis of Soft and Hard Reinsurance Markets for Flood Coverage. Atmosphere 2020, 11, 146. https://doi.org/10.3390/atmos11020146\n\n[[2]](https://doi.org/10.5194/nhess-22-659-2022) Rädler, A. T. (2022). Invited perspectives: how does climate change affect the risk of natural hazards? Challenges and step changes from the reinsurance perspective. Nat. Hazards Earth Syst. Sci., 22, 659–664. https://doi.org/10.5194/nhess-22-659-2022\n\n[[3]](https://www.ipcc.ch/site/assets/uploads/2018/03/wg2TARchap8.pdf) Vellinga, P., Mills, E., Bowers, L., Berz, G.A., Huq, S., Kozak, L.M., Paultikof, J., Schanzenbacker, B., Shida, S., Soler, G., Benson, C., Bidan, P., Bruce, J.W., Huyck, P.M., Lemcke, G., Peara, A., Radevsky, R., Schoubroeck, C.V., Dlugolecki, A.F. (2001). Insurance and other financial services. In J. J. McCarthy, O. F. Canziani, N. A. Leary, D. J. Dokken, & K. S. White (Eds.), Climate change 2001: impacts, adaptation, and vulnerability. Contribution of working group 2 to the third assessment report of the intergovernmental panel on climate change. (pp. 417-450). Cambridge University Press.\n\n[[4]](https://doi.org/10.1002/wcc.71) Lemos, M.C. and Rood, R.B. (2010). Climate projections and their impact on policy and practice. WIREs Clim Chg, 1: 670-682. https://doi.org/10.1002/wcc.71\n\n[[5]](https://doi.org/10.1002/wcc.579) Nissan H, Goddard L, de Perez EC, et al. (2019). On the use and misuse of climate change projections in international development. WIREs Clim Change, 10:e579. https://doi.org/10.1002/wcc.579\n\n[[6]](https://doi.org/10.5194/esd-13-321-2022) Cos, J., Doblas-Reyes, F., Jury, M., Marcos, R., Bretonnière, P.-A., and Samsó, M. (2022). The Mediterranean climate change hotspot in the CMIP5 and CMIP6 projections, Earth Syst. Dynam., 13, 321–340. https://doi.org/10.5194/esd-13-321-2022\n\n[[7]](https://doi.org/10.1002/wcc.8) Rummukainen, M. (2010). State-of-the-art with regional climate models. WIREs Clim Change, 1: 82-96. https://doi.org/10.1002/wcc.8\n\n[[8]](https://doi.org/10.1088/1748-9326/aacc77) Silje Lund Sørland et al. (2018). Bias patterns and climate change signals in GCM-RCM model chains. Environ. Res. Lett. 13 074017. https://doi.org/10.1088/1748-9326/aacc77"} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q03__040aef108a65", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q03", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q03", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Bias in precipitation-based indices for impact models", "title": "Bias in precipitation-based indices for impact models", "chunk_index": 0, "token_count": 96, "text_raw": "Production date: 29-05-2024\n\nProduced by: CMCC foundation - Euro-Mediterranean Center on Climate Change. Albert Martinez Boti.", "text_with_prefix": "EQC Quality Assessment: \"Bias in precipitation-based indices for impact models\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q03 | Category: Climate_Projections\nSection: Bias in precipitation-based indices for impact models\n---\nProduction date: 29-05-2024\n\nProduced by: CMCC foundation - Euro-Mediterranean Center on Climate Change. Albert Martinez Boti."} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q03__6f5a63a8913b", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q03", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q03", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Bias in precipitation-based indices for impact models > Quality assessment question", "title": "Bias in precipitation-based indices for impact models", "chunk_index": 1, "token_count": 536, "text_raw": "* **How well do CORDEX projections represent climatology and trends of precipitation-based indices in Europe?**\n\nSoil erosion stands as one of the primary environmental concerns in Europe [[1]](https://doi.org/10.1016/j.envsci.2015.08.012). Its accelerated occurrence can precipitate a decline in ecosystem stability, land productivity, and overall land degradation, resulting in diminished income for farmers [[2]](https://doi.org/10.1016/j.landusepol.2012.11.007). This notebook is designed to evaluate the effectiveness of a specific set of models from **[CORDEX](https://cds.climate.copernicus.eu/datasets/projections-cordex-domains-single-levels?tab=overview)** Regional Climate Models (RCMs) in accurately replicating historical precipitation extreme indices. These selected rainfall indicators are recognised as valuable proxies for rainfall erosivity, a factor directly linked to empirical calculations of soil loss, as outlined in an application available through the old CDS platform ([dataset documentation](https://dast.copernicus-climate.eu/documents/sis-soil-erosion/C3S_D429d.2.3.2_Product_User_Guide_v1.2.pdf)). The five precipitation-based indices used here are calculated using the **[icclim](https://icclim.readthedocs.io/en/stable/)** package, them being:\n- Spell length of days with precipitation greater than 1 mm (also known as \"Maximum consecutive wet days\" - \"CWD\")\n- Number of heavy precipitation days (Precip >= 20mm - \"R20mm\")\n- Number of wet days (Precip >= 1mm - \"RR1\")\n- Maximum 1-day total precipitation (\"RX1day\")\n- Maximum 5-day total precipitation (\"RX5day\").\n\nWithin this notebook, these calculations are performed over the temporal aggregation of DJF and for the historical period spanning from 1971 to 2000. It is important to note that the results presented here pertain to a specific subset of the CORDEX ensemble and may not be generalisable to the entire dataset. Also note that a separate assessment examines the representation of trends of these indices for the same models during a fixed future period (2015-2099).", "text_with_prefix": "EQC Quality Assessment: \"Bias in precipitation-based indices for impact models\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q03 | Category: Climate_Projections\nSection: Bias in precipitation-based indices for impact models > Quality assessment question\n---\n* **How well do CORDEX projections represent climatology and trends of precipitation-based indices in Europe?**\n\nSoil erosion stands as one of the primary environmental concerns in Europe [[1]](https://doi.org/10.1016/j.envsci.2015.08.012). Its accelerated occurrence can precipitate a decline in ecosystem stability, land productivity, and overall land degradation, resulting in diminished income for farmers [[2]](https://doi.org/10.1016/j.landusepol.2012.11.007). This notebook is designed to evaluate the effectiveness of a specific set of models from **[CORDEX](https://cds.climate.copernicus.eu/datasets/projections-cordex-domains-single-levels?tab=overview)** Regional Climate Models (RCMs) in accurately replicating historical precipitation extreme indices. These selected rainfall indicators are recognised as valuable proxies for rainfall erosivity, a factor directly linked to empirical calculations of soil loss, as outlined in an application available through the old CDS platform ([dataset documentation](https://dast.copernicus-climate.eu/documents/sis-soil-erosion/C3S_D429d.2.3.2_Product_User_Guide_v1.2.pdf)). The five precipitation-based indices used here are calculated using the **[icclim](https://icclim.readthedocs.io/en/stable/)** package, them being:\n- Spell length of days with precipitation greater than 1 mm (also known as \"Maximum consecutive wet days\" - \"CWD\")\n- Number of heavy precipitation days (Precip >= 20mm - \"R20mm\")\n- Number of wet days (Precip >= 1mm - \"RR1\")\n- Maximum 1-day total precipitation (\"RX1day\")\n- Maximum 5-day total precipitation (\"RX5day\").\n\nWithin this notebook, these calculations are performed over the temporal aggregation of DJF and for the historical period spanning from 1971 to 2000. It is important to note that the results presented here pertain to a specific subset of the CORDEX ensemble and may not be generalisable to the entire dataset. Also note that a separate assessment examines the representation of trends of these indices for the same models during a fixed future period (2015-2099)."} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q03__66f3e825a335", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q03", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q03", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Bias in precipitation-based indices for impact models > Quality assessment statement", "title": "Bias in precipitation-based indices for impact models", "chunk_index": 2, "token_count": 670, "text_raw": "These are the key outcomes of this assessment\n\n* The assessment of CORDEX projections regarding the climatology and trend of precipitation-based indices across Europe requires a comprehensive evaluation of the historical period.\n\n* The subset of Regional Climate Models (RCMs) considered in this notebook for the seasonal aggregation of DJF generally struggle to replicate both the observed climatology and historical trends of the selected precipitation-based indices, with particular difficulty in capturing the trends. However, the models manage to accurately capture the climatology and trend directions for certain regions and indices. For instance, the models reflect an increasing trend for the CWD index in the western part of the Iberian Peninsula and an upward trend for RR1 in central Europe. Users should take into account these regional variations, the specific indices they intend to use, the differing bias behaviours between climatology and trends, and the necessity of bias-correcting data before using it for hydrological applications [[3]](https://doi.org/10.1016/j.jhydrol.2012.05.052).\n\n* Precipitation trends in observational and reanalysis datasets are generally less robust and more challenging to detect compared to temperature trends. This limitation is also evident in model simulations.\n\n* While ERA5 serves as a reference dataset in this notebook, it is important to recognise its inherent uncertainties. Although uncertainties exist for all variables, these biases are especially relevant for variables like surface wind or precipitation, as shown in [[4]](https://doi.org/10.1002/qj.3616) and [[5]](https://doi.org/10.3390/atmos12111462). This highlights the need for cautious interpretation, particularly in the context of precipitation extremes.\n\n* A larger GCM-RCM matrix should be considered when addressing specific cases to enhance the robustness of the analysis and account for uncertainties [[6]](https://doi.org/10.1002/wcc.8)[[7]](https://doi.org/10.1088/1748-9326/aacc77).\n```\n\nattachment:4890e24e-847d-40bd-91e3-ea3202916a49.png\n---\nalt: mean_bias_RR1\nwidth: 900px\n---\nNumber of wet days (Precip >= 1mm) ('RRI') for the temporal aggreggtion of 'DJF'. Mean bias for the historical period (1971 - 2000) of each individual CORDEX model.\n```\n\n\n
\nNOTE on the DJF selection:
\nIt is important to note that using seasonal temporal aggregations offers only partial insights into the dynamics. For more comprehensive results, it is advisable to also consider other seasons and annual aggregations. However, for the sake of efficiency and to avoid making the notebook too heavy, we have opted to prioritise seasonal aggregations.", "text_with_prefix": "EQC Quality Assessment: \"Bias in precipitation-based indices for impact models\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q03 | Category: Climate_Projections\nSection: Bias in precipitation-based indices for impact models > Quality assessment statement\n---\nThese are the key outcomes of this assessment\n\n* The assessment of CORDEX projections regarding the climatology and trend of precipitation-based indices across Europe requires a comprehensive evaluation of the historical period.\n\n* The subset of Regional Climate Models (RCMs) considered in this notebook for the seasonal aggregation of DJF generally struggle to replicate both the observed climatology and historical trends of the selected precipitation-based indices, with particular difficulty in capturing the trends. However, the models manage to accurately capture the climatology and trend directions for certain regions and indices. For instance, the models reflect an increasing trend for the CWD index in the western part of the Iberian Peninsula and an upward trend for RR1 in central Europe. Users should take into account these regional variations, the specific indices they intend to use, the differing bias behaviours between climatology and trends, and the necessity of bias-correcting data before using it for hydrological applications [[3]](https://doi.org/10.1016/j.jhydrol.2012.05.052).\n\n* Precipitation trends in observational and reanalysis datasets are generally less robust and more challenging to detect compared to temperature trends. This limitation is also evident in model simulations.\n\n* While ERA5 serves as a reference dataset in this notebook, it is important to recognise its inherent uncertainties. Although uncertainties exist for all variables, these biases are especially relevant for variables like surface wind or precipitation, as shown in [[4]](https://doi.org/10.1002/qj.3616) and [[5]](https://doi.org/10.3390/atmos12111462). This highlights the need for cautious interpretation, particularly in the context of precipitation extremes.\n\n* A larger GCM-RCM matrix should be considered when addressing specific cases to enhance the robustness of the analysis and account for uncertainties [[6]](https://doi.org/10.1002/wcc.8)[[7]](https://doi.org/10.1088/1748-9326/aacc77).\n```\n\nattachment:4890e24e-847d-40bd-91e3-ea3202916a49.png\n---\nalt: mean_bias_RR1\nwidth: 900px\n---\nNumber of wet days (Precip >= 1mm) ('RRI') for the temporal aggreggtion of 'DJF'. Mean bias for the historical period (1971 - 2000) of each individual CORDEX model.\n```\n\n\n
\nNOTE on the DJF selection:
\nIt is important to note that using seasonal temporal aggregations offers only partial insights into the dynamics. For more comprehensive results, it is advisable to also consider other seasons and annual aggregations. However, for the sake of efficiency and to avoid making the notebook too heavy, we have opted to prioritise seasonal aggregations."} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q03__65e1b5722501", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q03", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q03", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Bias in precipitation-based indices for impact models > Methodology", "title": "Bias in precipitation-based indices for impact models", "chunk_index": 3, "token_count": 472, "text_raw": "This notebook provides an assessment of the systematic errors (trend and climatology) in a subset of 9 models from [CORDEX](https://cds.climate.copernicus.eu/datasets/projections-cordex-domains-single-levels?tab=overview). The analysis involves comparing model predictions with ERA5 reanalysis data using several precipitation-based indices. These indices, calculated over the DJF period for the historical period spanning from 1971 to 2000, include:\n\n- Spell length of days with precipitation greater than 1 mm (also known as \"Maximum consecutive wet days\" - \"CWD\")\n- Number of heavy precipitation days (Precip >= 20mm - \"R20mm\")\n- Number of wet days (Precip >= 1mm - \"RR1\")\n- Maximum 1-day total precipitation (\"RX1day\")\n- Maximum 5-day total precipitation (\"RX5day\").\n\nIn particular, spatial patterns of climate mean and trend, along with biases, are examined and displayed for each model and the ensemble median (calculated for each grid cell). Additionally, spatially-averaged trend values are analysed and presented using box plots to provide an overview of trend behavior across the distribution of the chosen subset of models when averaged across Europe.\n\nThe analysis and results follow the next outline:\n\n**[](section-1)**\n * [](section-1.1)\n * [](section-1.2)\n * [](section-1.3)\n * [](section-1.4)\n * [](section-1.5)\n * [](section-1.6)\n\n**[](section-2)**\n * [](section-2.1)\n * [](section-2.2)\n * [](section-2.3)\n\n**[](section-3)**\n * [](section-3.1)\n * [](section-3.2)\n * [](section-3.3)\n * [](section-3.4)\n * [](section-3.5)\n * [](section-3.6)", "text_with_prefix": "EQC Quality Assessment: \"Bias in precipitation-based indices for impact models\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q03 | Category: Climate_Projections\nSection: Bias in precipitation-based indices for impact models > Methodology\n---\nThis notebook provides an assessment of the systematic errors (trend and climatology) in a subset of 9 models from [CORDEX](https://cds.climate.copernicus.eu/datasets/projections-cordex-domains-single-levels?tab=overview). The analysis involves comparing model predictions with ERA5 reanalysis data using several precipitation-based indices. These indices, calculated over the DJF period for the historical period spanning from 1971 to 2000, include:\n\n- Spell length of days with precipitation greater than 1 mm (also known as \"Maximum consecutive wet days\" - \"CWD\")\n- Number of heavy precipitation days (Precip >= 20mm - \"R20mm\")\n- Number of wet days (Precip >= 1mm - \"RR1\")\n- Maximum 1-day total precipitation (\"RX1day\")\n- Maximum 5-day total precipitation (\"RX5day\").\n\nIn particular, spatial patterns of climate mean and trend, along with biases, are examined and displayed for each model and the ensemble median (calculated for each grid cell). Additionally, spatially-averaged trend values are analysed and presented using box plots to provide an overview of trend behavior across the distribution of the chosen subset of models when averaged across Europe.\n\nThe analysis and results follow the next outline:\n\n**[](section-1)**\n * [](section-1.1)\n * [](section-1.2)\n * [](section-1.3)\n * [](section-1.4)\n * [](section-1.5)\n * [](section-1.6)\n\n**[](section-2)**\n * [](section-2.1)\n * [](section-2.2)\n * [](section-2.3)\n\n**[](section-3)**\n * [](section-3.1)\n * [](section-3.2)\n * [](section-3.3)\n * [](section-3.4)\n * [](section-3.5)\n * [](section-3.6)"} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q03__fc04fe029999", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q03", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q03", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Bias in precipitation-based indices for impact models > Analysis and results > 1. Parameters, requests and functions definition > 1.2. Define Parameters", "title": "Bias in precipitation-based indices for impact models", "chunk_index": 4, "token_count": 335, "text_raw": "In the \"Define Parameters\" section, various customisable options for the notebook are specified:\n\n- The initial and ending year used for the historical period can be specified by changing the parameters `year_start` and `year_stop` (1971-2000 is chosen for consistency between CORDEX and CMIP6).\n- The `timeseries` set the temporal aggregation. For instance, selecting \"DJF\" implies considering only the winter season.\n- `collection_id` provides the choice between Global Climate Models CMIP6 or Regional Climate Models CORDEX. Although the code allows choosing between CMIP6 or CORDEX, the example provided in this notebook deals with CORDEX RCMs.\n- `area` allows specifying the geographical domain of interest.\n- The `interpolation_method` parameter allows selecting the interpolation method when regridding is performed over the indices.\n- The `chunk` selection allows the user to define if dividing into chunks when downloading the data on their local machine. Although it does not significantly affect the analysis, it is recommended to keep the default value for optimal performance.\n\nChoose annual or seasonal timeseries\nChoose CORDEX or CMIP6\nInterpolation method\nArea to show\nChunks for download\n\n(section-1.3)=", "text_with_prefix": "EQC Quality Assessment: \"Bias in precipitation-based indices for impact models\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q03 | Category: Climate_Projections\nSection: Bias in precipitation-based indices for impact models > Analysis and results > 1. Parameters, requests and functions definition > 1.2. Define Parameters\n---\nIn the \"Define Parameters\" section, various customisable options for the notebook are specified:\n\n- The initial and ending year used for the historical period can be specified by changing the parameters `year_start` and `year_stop` (1971-2000 is chosen for consistency between CORDEX and CMIP6).\n- The `timeseries` set the temporal aggregation. For instance, selecting \"DJF\" implies considering only the winter season.\n- `collection_id` provides the choice between Global Climate Models CMIP6 or Regional Climate Models CORDEX. Although the code allows choosing between CMIP6 or CORDEX, the example provided in this notebook deals with CORDEX RCMs.\n- `area` allows specifying the geographical domain of interest.\n- The `interpolation_method` parameter allows selecting the interpolation method when regridding is performed over the indices.\n- The `chunk` selection allows the user to define if dividing into chunks when downloading the data on their local machine. Although it does not significantly affect the analysis, it is recommended to keep the default value for optimal performance.\n\nChoose annual or seasonal timeseries\nChoose CORDEX or CMIP6\nInterpolation method\nArea to show\nChunks for download\n\n(section-1.3)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q03__1d4100d5ae74", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q03", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q03", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Bias in precipitation-based indices for impact models > Analysis and results > 1. Parameters, requests and functions definition > 1.3. Define models", "title": "Bias in precipitation-based indices for impact models", "chunk_index": 5, "token_count": 314, "text_raw": "The following climate analyses are performed considering a subset of GCMs from CMIP6. Models names are listed in the parameters below. Some variable-dependent parameters are also selected, as the `index_names` parameter, which specifies the precipitation-based indices ('CWD', 'R20mm', 'RR1', 'RX1day' and 'RX5day' in our case) from the [icclim](https://icclim.readthedocs.io/en/stable/) Python package.\n\nWhen choosing Cordex models, it is crucial to consider the availability of RCMs for the selected GCM and the specified region. The listed RCMs, for instance, are accessible for the GCM “mpi_m_mpi_esm_lr” in the “europe” cordex_domain. To confirm the available combinations, refer to the [CORDEX](https://cds.climate.copernicus.eu/datasets/projections-cordex-domains-single-levels?tab=overview) CDS catalogue entry.\n\nDefine dictionaries to use in titles and caption\nDefine dictionaries to use in titles and caption\n\n(section-1.4)=", "text_with_prefix": "EQC Quality Assessment: \"Bias in precipitation-based indices for impact models\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q03 | Category: Climate_Projections\nSection: Bias in precipitation-based indices for impact models > Analysis and results > 1. Parameters, requests and functions definition > 1.3. Define models\n---\nThe following climate analyses are performed considering a subset of GCMs from CMIP6. Models names are listed in the parameters below. Some variable-dependent parameters are also selected, as the `index_names` parameter, which specifies the precipitation-based indices ('CWD', 'R20mm', 'RR1', 'RX1day' and 'RX5day' in our case) from the [icclim](https://icclim.readthedocs.io/en/stable/) Python package.\n\nWhen choosing Cordex models, it is crucial to consider the availability of RCMs for the selected GCM and the specified region. The listed RCMs, for instance, are accessible for the GCM “mpi_m_mpi_esm_lr” in the “europe” cordex_domain. To confirm the available combinations, refer to the [CORDEX](https://cds.climate.copernicus.eu/datasets/projections-cordex-domains-single-levels?tab=overview) CDS catalogue entry.\n\nDefine dictionaries to use in titles and caption\nDefine dictionaries to use in titles and caption\n\n(section-1.4)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q03__4382fcb98ca8", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q03", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q03", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Bias in precipitation-based indices for impact models > Analysis and results > 1. Parameters, requests and functions definition > 1.4. Define ERA5 request", "title": "Bias in precipitation-based indices for impact models", "chunk_index": 6, "token_count": 123, "text_raw": "Within this notebook, ERA5 serves as the reference product. In this section, we set the required parameters for the cds-api data-request of ERA5.\n\n(section-1.5)=", "text_with_prefix": "EQC Quality Assessment: \"Bias in precipitation-based indices for impact models\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q03 | Category: Climate_Projections\nSection: Bias in precipitation-based indices for impact models > Analysis and results > 1. Parameters, requests and functions definition > 1.4. Define ERA5 request\n---\nWithin this notebook, ERA5 serves as the reference product. In this section, we set the required parameters for the cds-api data-request of ERA5.\n\n(section-1.5)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q03__331dc90b5296", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q03", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q03", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Bias in precipitation-based indices for impact models > Analysis and results > 1. Parameters, requests and functions definition > 1.5. Define model requests", "title": "Bias in precipitation-based indices for impact models", "chunk_index": 7, "token_count": 212, "text_raw": "In this section we set the required parameters for the cds-api data-request.\n\nThe `get_cordex_years` function is employed to choose suitable data chunks for CORDEX data requests.\n\nWhen `Weights = True`, spatial weighting is applied for calculations requiring spatial data aggregation. This is particularly relevant for CMIP6 GCMs with regular lon-lat grids that do not consider varying surface extensions at different latitudes. In contrast, CORDEX RCMs, using rotated grids, inherently account for different cell surfaces based on latitude, eliminating the need for a latitude cosine multiplicative factor (`Weights = False`).\n\n(section-1.6)=", "text_with_prefix": "EQC Quality Assessment: \"Bias in precipitation-based indices for impact models\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q03 | Category: Climate_Projections\nSection: Bias in precipitation-based indices for impact models > Analysis and results > 1. Parameters, requests and functions definition > 1.5. Define model requests\n---\nIn this section we set the required parameters for the cds-api data-request.\n\nThe `get_cordex_years` function is employed to choose suitable data chunks for CORDEX data requests.\n\nWhen `Weights = True`, spatial weighting is applied for calculations requiring spatial data aggregation. This is particularly relevant for CMIP6 GCMs with regular lon-lat grids that do not consider varying surface extensions at different latitudes. In contrast, CORDEX RCMs, using rotated grids, inherently account for different cell surfaces based on latitude, eliminating the need for a latitude cosine multiplicative factor (`Weights = False`).\n\n(section-1.6)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q03__ff6e71f10b1a", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q03", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q03", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Bias in precipitation-based indices for impact models > Analysis and results > 1. Parameters, requests and functions definition > 1.6. Functions to cache", "title": "Bias in precipitation-based indices for impact models", "chunk_index": 8, "token_count": 302, "text_raw": "In this section, functions that will be executed in the caching phase are defined. Caching is the process of storing copies of files in a temporary storage location, so that they can be accessed more quickly. This process also checks if the user has already downloaded a file, avoiding redundant downloads.\n\nFunctions description:\n\n- The `select_timeseries` function subsets the dataset based on the chosen `timeseries` parameter.\n\n- The `compute_indices` function utilises the icclim package to calculate the precipitation-based indices.\n\n- The `compute_trends` function employs the Mann-Kendall test for trend calculation.\n\n- Finally, the `compute_indices_and_trends` accumulates the daily precipitation (only if we are dealing with ERA5), calculates the precipitation-based indices for the corresponding temporal aggregation using the `compute_indices` function, determines the indices mean over the historical period (1971-2000), obtain the trends using the `compute_trends` function, and offers an option for regridding to ERA5 if required.\n\nOriginal bounds for conservative interpolation\n\n(section-2)=", "text_with_prefix": "EQC Quality Assessment: \"Bias in precipitation-based indices for impact models\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q03 | Category: Climate_Projections\nSection: Bias in precipitation-based indices for impact models > Analysis and results > 1. Parameters, requests and functions definition > 1.6. Functions to cache\n---\nIn this section, functions that will be executed in the caching phase are defined. Caching is the process of storing copies of files in a temporary storage location, so that they can be accessed more quickly. This process also checks if the user has already downloaded a file, avoiding redundant downloads.\n\nFunctions description:\n\n- The `select_timeseries` function subsets the dataset based on the chosen `timeseries` parameter.\n\n- The `compute_indices` function utilises the icclim package to calculate the precipitation-based indices.\n\n- The `compute_trends` function employs the Mann-Kendall test for trend calculation.\n\n- Finally, the `compute_indices_and_trends` accumulates the daily precipitation (only if we are dealing with ERA5), calculates the precipitation-based indices for the corresponding temporal aggregation using the `compute_indices` function, determines the indices mean over the historical period (1971-2000), obtain the trends using the `compute_trends` function, and offers an option for regridding to ERA5 if required.\n\nOriginal bounds for conservative interpolation\n\n(section-2)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q03__62b3c526f18f", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q03", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q03", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Bias in precipitation-based indices for impact models > Analysis and results > 2. Downloading and processing > 2.1. Download and transform ERA5", "title": "Bias in precipitation-based indices for impact models", "chunk_index": 9, "token_count": 179, "text_raw": "In this section, the `download.download_and_transform` function from the 'c3s_eqc_automatic_quality_control' package is employed to download ERA5 reference data, accumulate daily precipitation from hourly data, compute the precipitation-based indices for the selected temporal aggregation (\"DJF\" in this example), calculate the mean and trend over the historical period (1971-2000) and cache the result (to avoid redundant downloads and processing).\n\n(section-2.2)=", "text_with_prefix": "EQC Quality Assessment: \"Bias in precipitation-based indices for impact models\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q03 | Category: Climate_Projections\nSection: Bias in precipitation-based indices for impact models > Analysis and results > 2. Downloading and processing > 2.1. Download and transform ERA5\n---\nIn this section, the `download.download_and_transform` function from the 'c3s_eqc_automatic_quality_control' package is employed to download ERA5 reference data, accumulate daily precipitation from hourly data, compute the precipitation-based indices for the selected temporal aggregation (\"DJF\" in this example), calculate the mean and trend over the historical period (1971-2000) and cache the result (to avoid redundant downloads and processing).\n\n(section-2.2)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q03__2583eb5aab98", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q03", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q03", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Bias in precipitation-based indices for impact models > Analysis and results > 2. Downloading and processing > 2.2. Download and transform models", "title": "Bias in precipitation-based indices for impact models", "chunk_index": 10, "token_count": 308, "text_raw": "In this section, the `download.download_and_transform` function from the 'c3s_eqc_automatic_quality_control' package is employed to download daily data from the CORDEX models, compute the precipitation-based indices for the selected temporal aggregation, calculate the mean and trend over the historical period (1971-2005), interpolate to ERA5's grid (only for the cases in which it is specified, in the other cases, the original model's grid is mantained), and cache the result (to avoid redundant downloads and processing).\n\nOriginal model\nInterpolated model\n\n```text\nmodel='clmcom_clm_cclm4_8_17'\nmodel='clmcom_eth_cosmo_crclim'\nmodel='cnrm_aladin63'\nmodel='dmi_hirham5'\nmodel='knmi_racmo22e'\nmodel='mohc_hadrem3_ga7_05'\nmodel='mpi_csc_remo2009'\nmodel='smhi_rca4'\nmodel='uhoh_wrf361h'\n```\n\n(section-2.3)=", "text_with_prefix": "EQC Quality Assessment: \"Bias in precipitation-based indices for impact models\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q03 | Category: Climate_Projections\nSection: Bias in precipitation-based indices for impact models > Analysis and results > 2. Downloading and processing > 2.2. Download and transform models\n---\nIn this section, the `download.download_and_transform` function from the 'c3s_eqc_automatic_quality_control' package is employed to download daily data from the CORDEX models, compute the precipitation-based indices for the selected temporal aggregation, calculate the mean and trend over the historical period (1971-2005), interpolate to ERA5's grid (only for the cases in which it is specified, in the other cases, the original model's grid is mantained), and cache the result (to avoid redundant downloads and processing).\n\nOriginal model\nInterpolated model\n\n```text\nmodel='clmcom_clm_cclm4_8_17'\nmodel='clmcom_eth_cosmo_crclim'\nmodel='cnrm_aladin63'\nmodel='dmi_hirham5'\nmodel='knmi_racmo22e'\nmodel='mohc_hadrem3_ga7_05'\nmodel='mpi_csc_remo2009'\nmodel='smhi_rca4'\nmodel='uhoh_wrf361h'\n```\n\n(section-2.3)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q03__7f450810a8f3", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q03", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q03", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Bias in precipitation-based indices for impact models > Analysis and results > 2. Downloading and processing > 2.3. Apply land-sea mask, change attributes and cut the region to show", "title": "Bias in precipitation-based indices for impact models", "chunk_index": 11, "token_count": 260, "text_raw": "This section performs the following tasks:\n\n1. Cut the region of interest.\n2. Downloads the sea mask for ERA5.\n3. Applies the sea mask to both ERA5 data and the model data, which were previously regridded to ERA5's grid (i.e., it applies the ERA5 sea mask to `ds_interpolated`).\n4. Regrids the ERA5 land-sea mask to the model's grid and applies it to them.\n5. Change some variable attributes for plotting purposes.\n\n**Note:** `ds_interpolated` contains data from the models (mean and trend over the historical period, p-value of the trends...) regridded to ERA5. `model_datasets` contain the same data but in the original grid of each model.\n\nEdit attributes\n\n(section-3)=", "text_with_prefix": "EQC Quality Assessment: \"Bias in precipitation-based indices for impact models\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q03 | Category: Climate_Projections\nSection: Bias in precipitation-based indices for impact models > Analysis and results > 2. Downloading and processing > 2.3. Apply land-sea mask, change attributes and cut the region to show\n---\nThis section performs the following tasks:\n\n1. Cut the region of interest.\n2. Downloads the sea mask for ERA5.\n3. Applies the sea mask to both ERA5 data and the model data, which were previously regridded to ERA5's grid (i.e., it applies the ERA5 sea mask to `ds_interpolated`).\n4. Regrids the ERA5 land-sea mask to the model's grid and applies it to them.\n5. Change some variable attributes for plotting purposes.\n\n**Note:** `ds_interpolated` contains data from the models (mean and trend over the historical period, p-value of the trends...) regridded to ERA5. `model_datasets` contain the same data but in the original grid of each model.\n\nEdit attributes\n\n(section-3)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q03__2f8a3a8512d7", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q03", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q03", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Bias in precipitation-based indices for impact models > Analysis and results > 3. Plot and describe results", "title": "Bias in precipitation-based indices for impact models", "chunk_index": 12, "token_count": 326, "text_raw": "This section will display the following results:\n\n- Maps representing the spatial distribution of the **historical mean values** (1971-2000) of the indices ('CWD', 'R20mm', 'RR1', 'RX1day' and 'RX5day') for ERA5, each model individually, the ensemble median (understood as the median of the mean values of the chosen subset of models calculated for each grid cell), and the ensemble spread (derived as the standard deviation of the distribution of the chosen subset of models).\n- Maps representing the spatial distribution of the **historical trends** (1971-2000) of the considered indices. Similar to the first analysis, this includes ERA5, each model individually, the ensemble median (understood as the median of the trend values of the chosen subset of models calculated for each grid cell), and the ensemble spread (derived as the standard deviation of the distribution of the chosen subset of models).\n- **Bias maps of the historical mean values**.\n- **Trend bias maps**. \n- **Boxplots** representing statistical distributions (PDF) built on the spatially-averaged historical trends from each considered model, displayed together with ERA5.\n\n(section-3.1)=", "text_with_prefix": "EQC Quality Assessment: \"Bias in precipitation-based indices for impact models\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q03 | Category: Climate_Projections\nSection: Bias in precipitation-based indices for impact models > Analysis and results > 3. Plot and describe results\n---\nThis section will display the following results:\n\n- Maps representing the spatial distribution of the **historical mean values** (1971-2000) of the indices ('CWD', 'R20mm', 'RR1', 'RX1day' and 'RX5day') for ERA5, each model individually, the ensemble median (understood as the median of the mean values of the chosen subset of models calculated for each grid cell), and the ensemble spread (derived as the standard deviation of the distribution of the chosen subset of models).\n- Maps representing the spatial distribution of the **historical trends** (1971-2000) of the considered indices. Similar to the first analysis, this includes ERA5, each model individually, the ensemble median (understood as the median of the trend values of the chosen subset of models calculated for each grid cell), and the ensemble spread (derived as the standard deviation of the distribution of the chosen subset of models).\n- **Bias maps of the historical mean values**.\n- **Trend bias maps**. \n- **Boxplots** representing statistical distributions (PDF) built on the spatially-averaged historical trends from each considered model, displayed together with ERA5.\n\n(section-3.1)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q03__25795597a7e2", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q03", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q03", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Bias in precipitation-based indices for impact models > Analysis and results > 3. Plot and describe results > 3.1. Define plotting functions", "title": "Bias in precipitation-based indices for impact models", "chunk_index": 13, "token_count": 666, "text_raw": "The functions presented here are used to plot the mean values and trends calculated over the historical period (1971-2000) for each of the indices ('CWD', 'R20mm', 'RR1', 'RX1day' and 'RX5day').\n\nFor a selected index, three layout types can be displayed, depending on the chosen function:\n\n1. Layout including the reference ERA5 product, the ensemble median, the bias of the ensemble median, and the ensemble spread: `plot_ensemble()` is used.\n2. Layout including every model (for the trend and mean values): `plot_models()` is employed.\n3. Layout including the bias of every model (for the trend and mean values): `plot_models()` is used again.\n\n`trend==True` argument allows displaying trend values over the historical period, while `trend==False` will show mean values. When the `trend` argument is set to `True`, regions with no significance are hatched. For individual models and ERA5, a grid point is considered to have a statistically significant trend when the p-value is lower than 0.05 (in such cases, no hatching is shown). However, for determining trend significance for the ensemble median (understood as the median of the trend values of the chosen subset of models calculated for each grid cell), reliance is placed on agreement categories, following the advanced approach proposed in AR6 [IPCC](https://www.ipcc.ch/report/ar6/wg1/downloads/report/IPCC_AR6_WGI_Atlas.pdf) on pages 1945-1950. The `hatch_p_value_ensemble()` function is used to distinguish, for each grid point, between three possible cases:\n\n1. If more than 66% of the models are statistically significant (p-value < 0.05) and more than 80% of the models share the same sign, we consider the ensemble median trend to be statistically significant, and there is agreement on the sign. To represent this, no hatching is used.\n2. If less than 66% of the models are statistically significant, regardless of agreement on the sign of the trend, hatching is applied (indicating that the ensemble median trend is not statistically significant).\n3. If more than 66% of the models are statistically significant but less than 80% of the models share the same sign, we consider the ensemble median trend to be statistically significant, but there is no agreement on the sign of the trend. This is represented using crosses.\n\nDefine function to plot the caption of the figures (for the ensemble case)\nAdd caption to the figure\nAdd each line to the figure\nend captioning\nDefine function to plot the caption of the figures (for the individual models case)\nAdd caption to the figure\nAdd each line to the figure\n\n(section-3.2)=", "text_with_prefix": "EQC Quality Assessment: \"Bias in precipitation-based indices for impact models\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q03 | Category: Climate_Projections\nSection: Bias in precipitation-based indices for impact models > Analysis and results > 3. Plot and describe results > 3.1. Define plotting functions\n---\nThe functions presented here are used to plot the mean values and trends calculated over the historical period (1971-2000) for each of the indices ('CWD', 'R20mm', 'RR1', 'RX1day' and 'RX5day').\n\nFor a selected index, three layout types can be displayed, depending on the chosen function:\n\n1. Layout including the reference ERA5 product, the ensemble median, the bias of the ensemble median, and the ensemble spread: `plot_ensemble()` is used.\n2. Layout including every model (for the trend and mean values): `plot_models()` is employed.\n3. Layout including the bias of every model (for the trend and mean values): `plot_models()` is used again.\n\n`trend==True` argument allows displaying trend values over the historical period, while `trend==False` will show mean values. When the `trend` argument is set to `True`, regions with no significance are hatched. For individual models and ERA5, a grid point is considered to have a statistically significant trend when the p-value is lower than 0.05 (in such cases, no hatching is shown). However, for determining trend significance for the ensemble median (understood as the median of the trend values of the chosen subset of models calculated for each grid cell), reliance is placed on agreement categories, following the advanced approach proposed in AR6 [IPCC](https://www.ipcc.ch/report/ar6/wg1/downloads/report/IPCC_AR6_WGI_Atlas.pdf) on pages 1945-1950. The `hatch_p_value_ensemble()` function is used to distinguish, for each grid point, between three possible cases:\n\n1. If more than 66% of the models are statistically significant (p-value < 0.05) and more than 80% of the models share the same sign, we consider the ensemble median trend to be statistically significant, and there is agreement on the sign. To represent this, no hatching is used.\n2. If less than 66% of the models are statistically significant, regardless of agreement on the sign of the trend, hatching is applied (indicating that the ensemble median trend is not statistically significant).\n3. If more than 66% of the models are statistically significant but less than 80% of the models share the same sign, we consider the ensemble median trend to be statistically significant, but there is no agreement on the sign of the trend. This is represented using crosses.\n\nDefine function to plot the caption of the figures (for the ensemble case)\nAdd caption to the figure\nAdd each line to the figure\nend captioning\nDefine function to plot the caption of the figures (for the individual models case)\nAdd caption to the figure\nAdd each line to the figure\n\n(section-3.2)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q03__0f19c3500601", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q03", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q03", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Bias in precipitation-based indices for impact models > Analysis and results > 3. Plot and describe results > 3.2. Plot ensemble maps", "title": "Bias in precipitation-based indices for impact models", "chunk_index": 14, "token_count": 372, "text_raw": "In this section, we invoke the `plot_ensemble()` function to visualise the mean values and trends calculated over the historical period (1971-2000) for the model ensemble and ERA5 reference product across Europe. Note that the model data used in this section has previously been interpolated to the ERA5 grid.\n\nSpecifically, for each of the indices ('CWD', 'R20mm', 'RR1', 'RX1day' and 'RX5day'), this section presents two layouts:\n\n1. Mean values of the historical period (1971-2000) for: (a) the reference ERA5 product, (b) the ensemble median (understood as the median of the mean values of the chosen subset of models calculated for each grid cell), (c) the bias of the ensemble median, and (d) the ensemble spread (derived as the standard deviation of the distribution of the chosen subset of models).\n\n2. Trend values of the historical period (1971-2000) for: (a) the reference ERA5 product, (b) the ensemble median (understood as the median of the trend values of the chosen subset of models calculated for each grid cell), (c) the bias of the ensemble median, and (d) the ensemble spread (derived as the standard deviation of the distribution of the chosen subset of models).\n\nFig number counter\nCommon title\n\n(section-3.3)=", "text_with_prefix": "EQC Quality Assessment: \"Bias in precipitation-based indices for impact models\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q03 | Category: Climate_Projections\nSection: Bias in precipitation-based indices for impact models > Analysis and results > 3. Plot and describe results > 3.2. Plot ensemble maps\n---\nIn this section, we invoke the `plot_ensemble()` function to visualise the mean values and trends calculated over the historical period (1971-2000) for the model ensemble and ERA5 reference product across Europe. Note that the model data used in this section has previously been interpolated to the ERA5 grid.\n\nSpecifically, for each of the indices ('CWD', 'R20mm', 'RR1', 'RX1day' and 'RX5day'), this section presents two layouts:\n\n1. Mean values of the historical period (1971-2000) for: (a) the reference ERA5 product, (b) the ensemble median (understood as the median of the mean values of the chosen subset of models calculated for each grid cell), (c) the bias of the ensemble median, and (d) the ensemble spread (derived as the standard deviation of the distribution of the chosen subset of models).\n\n2. Trend values of the historical period (1971-2000) for: (a) the reference ERA5 product, (b) the ensemble median (understood as the median of the trend values of the chosen subset of models calculated for each grid cell), (c) the bias of the ensemble median, and (d) the ensemble spread (derived as the standard deviation of the distribution of the chosen subset of models).\n\nFig number counter\nCommon title\n\n(section-3.3)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q03__38bb4ac7f33a", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q03", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q03", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Bias in precipitation-based indices for impact models > Analysis and results > 3. Plot and describe results > 3.3. Plot model maps", "title": "Bias in precipitation-based indices for impact models", "chunk_index": 15, "token_count": 226, "text_raw": "In this section, we invoke the `plot_models()` function to visualise the mean values and trends calculated over the historical period (1971-2000) for every model individually across Europe. Note that the model data used in this section maintains its original grid.\n\nSpecifically, for each of the indices ('CWD', 'R20mm', 'RR1', 'RX1day' and 'RX5day'), this section presents two layouts:\n\n1. A layout including the historical mean (1971-2000) of every model (only for the 'SU' index).\n\n2. A layout including the historical trend (1971-2000) of every model.\n\n(section-3.4)=", "text_with_prefix": "EQC Quality Assessment: \"Bias in precipitation-based indices for impact models\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q03 | Category: Climate_Projections\nSection: Bias in precipitation-based indices for impact models > Analysis and results > 3. Plot and describe results > 3.3. Plot model maps\n---\nIn this section, we invoke the `plot_models()` function to visualise the mean values and trends calculated over the historical period (1971-2000) for every model individually across Europe. Note that the model data used in this section maintains its original grid.\n\nSpecifically, for each of the indices ('CWD', 'R20mm', 'RR1', 'RX1day' and 'RX5day'), this section presents two layouts:\n\n1. A layout including the historical mean (1971-2000) of every model (only for the 'SU' index).\n\n2. A layout including the historical trend (1971-2000) of every model.\n\n(section-3.4)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q03__e86673d645e5", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q03", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q03", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Bias in precipitation-based indices for impact models > Analysis and results > 3. Plot and describe results > 3.4. Plot bias maps", "title": "Bias in precipitation-based indices for impact models", "chunk_index": 16, "token_count": 240, "text_raw": "In this section, we invoke the `plot_models()` function to visualise the bias for the mean values and trends calculated over the historical period (1971-2000) for every model individually across Europe. Note that the model data used in this section has previously been interpolated to the ERA5 grid.\n\nSpecifically, for each of the indices ('CWD', 'R20mm', 'RR1', 'RX1day' and 'RX5day'), this section presents two layouts:\n\n1. A layout including the bias for the historical mean (1971-2000) of every model (only for the 'SU' index).\n\n2. A layout including the bias for the historical trend (1971-2000) of every model.\n\n(section-3.5)=", "text_with_prefix": "EQC Quality Assessment: \"Bias in precipitation-based indices for impact models\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q03 | Category: Climate_Projections\nSection: Bias in precipitation-based indices for impact models > Analysis and results > 3. Plot and describe results > 3.4. Plot bias maps\n---\nIn this section, we invoke the `plot_models()` function to visualise the bias for the mean values and trends calculated over the historical period (1971-2000) for every model individually across Europe. Note that the model data used in this section has previously been interpolated to the ERA5 grid.\n\nSpecifically, for each of the indices ('CWD', 'R20mm', 'RR1', 'RX1day' and 'RX5day'), this section presents two layouts:\n\n1. A layout including the bias for the historical mean (1971-2000) of every model (only for the 'SU' index).\n\n2. A layout including the bias for the historical trend (1971-2000) of every model.\n\n(section-3.5)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q03__c9ab48be90a7", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q03", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q03", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Bias in precipitation-based indices for impact models > Analysis and results > 3. Plot and describe results > 3.5. Boxplots of the historical trend", "title": "Bias in precipitation-based indices for impact models", "chunk_index": 17, "token_count": 404, "text_raw": "In this last section, we compare the trends of the climate models with the reference trend from ERA5.\n\nDots represent the spatially-averaged historical trend over the selected region (change of the number of days per decade) for each model (grey), the ensemble mean (blue), and the reference product (orange). The ensemble median is shown as a green line. Note that the spatially averaged values are calculated for each model from its original grid (i.e., no interpolated data has been used here).\n\nThe boxplot visually illustrates the distribution of trends among the climate models, with the box covering the first quartile (Q1 = 25th percentile) to the third quartile (Q3 = 75th percentile), and a green line indicating the ensemble median (Q2 = 50th percentile). Whiskers extend from the edges of the box to show the full data range.\n\nEnsemble mean\n\n
\n
\n

Fig 31. Boxplots illustrating the historical trends of the ensemble distribution and ERA5 for the precipitation-based indices: (a)'CWD', (b)'R20mm', (c)'RR1', (d)'RX1day' and (e)'RX5day'. The distribution is created by considering spatially averaged trends across Europe. The ensemble mean and the ensemble median trends are both included. Outliers in the distribution are denoted by a grey circle with a black contour.

\n
\n\n(section-3.6)=", "text_with_prefix": "EQC Quality Assessment: \"Bias in precipitation-based indices for impact models\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q03 | Category: Climate_Projections\nSection: Bias in precipitation-based indices for impact models > Analysis and results > 3. Plot and describe results > 3.5. Boxplots of the historical trend\n---\nIn this last section, we compare the trends of the climate models with the reference trend from ERA5.\n\nDots represent the spatially-averaged historical trend over the selected region (change of the number of days per decade) for each model (grey), the ensemble mean (blue), and the reference product (orange). The ensemble median is shown as a green line. Note that the spatially averaged values are calculated for each model from its original grid (i.e., no interpolated data has been used here).\n\nThe boxplot visually illustrates the distribution of trends among the climate models, with the box covering the first quartile (Q1 = 25th percentile) to the third quartile (Q3 = 75th percentile), and a green line indicating the ensemble median (Q2 = 50th percentile). Whiskers extend from the edges of the box to show the full data range.\n\nEnsemble mean\n\n
\n
\n

Fig 31. Boxplots illustrating the historical trends of the ensemble distribution and ERA5 for the precipitation-based indices: (a)'CWD', (b)'R20mm', (c)'RR1', (d)'RX1day' and (e)'RX5day'. The distribution is created by considering spatially averaged trends across Europe. The ensemble mean and the ensemble median trends are both included. Outliers in the distribution are denoted by a grey circle with a black contour.

\n
\n\n(section-3.6)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q03__e4eaf1da5ef0", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q03", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q03", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Bias in precipitation-based indices for impact models > Analysis and results > 3. Plot and describe results > 3.6. Results summary and discussion", "title": "Bias in precipitation-based indices for impact models", "chunk_index": 18, "token_count": 958, "text_raw": "- The bias behaviour varies depending on the index and whether we are dealing with trends or climatologies. Generally speaking, the subset of models selected overestimates the magnitudes of climatologies for DJF, except in the northwest of Great Britain, the west coast of the Scandinavian Peninsula, parts of the Mediterranean Basin, and the east of Europe for the CWD index, where there is underestimation. For the RR1 index, the models also underestimate the magnitudes for the west coast of the Scandinavian Peninsula and parts of the Mediterranean Basin.\n\n- For the historical period and the seasonal aggregation of DJF, ERA5 exhibits trends that are highly dependent on the region and index. Indices like CWD, R20mm, and RX1day show scattered spatial patterns that are generally not well-reproduced by the considered models, except in some specific areas. The spatial patterns for the trends of RR1 and RX5day are less scattered and are slightly better captured than those of the other indices.\n\n- The boxplots displaying spatially-averaged values reveal diverse outcomes depending on the index under consideration. Notably, significant inter-model variability is evident for certain indices, whose interquantile range encompasses trends of varying directions, and in some cases, even exhibits a trend opposite to that shown by the ERA5 data. This discrepancy is particularly noticeable for the R20mm, RR1, and RX1day indices, where the interquantile range showcases a trend contrary to that portrayed by ERA5. Meanwhile, the RX5day index displays an interquantile range centered around zero, with the trend sign dependent on the model. In contrast, the CWD index demonstrates a trend direction consistent with ERA5.\n\n- It is crucial to emphasise that the boxplots present spatially-averaged values, and their interpretation should be approached with caution, as the outcomes can vary significantly when considering different regions across Europe. For regional analyses, it is essential to focus solely on the region of interest, as the influence of other European regions may distort the results, leading to conclusions that do not accurately reflect the specific area under study.\n\n- Although ERA5 is used as our reference dataset, it is important to be aware of its inherent uncertainties. While uncertainties affect all variables, biases are particularly pronounced for variables such as surface wind and precipitation, as demonstrated in [[4]](https://doi.org/10.1002/qj.3616) and [[5]](https://doi.org/10.3390/atmos12111462). This underscores the importance of careful interpretation, especially regarding precipitation extremes.\n\n- A larger GCM-RCM matrix should be considered when addressing specific cases to enhance the robustness of the analysis and account for uncertainties [[6]](https://doi.org/10.1002/wcc.8)[[7]](https://doi.org/10.1088/1748-9326/aacc77).\n\n- What do the results mean for users? Are the biases relevant?\n\n- The subset of Regional Climate Models (RCMs) considered for this notebook and for this seasonal aggregation struggle to replicate both the observed climatology and the trends during the historical period, with particular difficulty in capturing the trends.\n\n- However, for specific regions and indices, the models successfully capture the climatology and even the direction of the trend. Examples include the increasing trend for the CWD index in the western part of the Iberian Peninsula and the positive trend for RR1 in central Europe. Users should consider these regional variations, the indices they aim to use, the differences in bias behavior between climatology and trends, and the necessity of bias-correcting data before using it for hydrological applications [[3]](https://doi.org/10.1016/j.jhydrol.2012.05.052).\n\n- Precipitation trends in observational and reanalysis datasets are generally less robust and more challenging to detect compared to temperature trends. This limitation is also evident in model simulations.\n\n\n
\nRESULTS NOTE:
\nIt is important to note that the results presented are specific to the 9 models chosen, and users should aim to assess as wide a range of models as possible before making a sub-selection.", "text_with_prefix": "EQC Quality Assessment: \"Bias in precipitation-based indices for impact models\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q03 | Category: Climate_Projections\nSection: Bias in precipitation-based indices for impact models > Analysis and results > 3. Plot and describe results > 3.6. Results summary and discussion\n---\n- The bias behaviour varies depending on the index and whether we are dealing with trends or climatologies. Generally speaking, the subset of models selected overestimates the magnitudes of climatologies for DJF, except in the northwest of Great Britain, the west coast of the Scandinavian Peninsula, parts of the Mediterranean Basin, and the east of Europe for the CWD index, where there is underestimation. For the RR1 index, the models also underestimate the magnitudes for the west coast of the Scandinavian Peninsula and parts of the Mediterranean Basin.\n\n- For the historical period and the seasonal aggregation of DJF, ERA5 exhibits trends that are highly dependent on the region and index. Indices like CWD, R20mm, and RX1day show scattered spatial patterns that are generally not well-reproduced by the considered models, except in some specific areas. The spatial patterns for the trends of RR1 and RX5day are less scattered and are slightly better captured than those of the other indices.\n\n- The boxplots displaying spatially-averaged values reveal diverse outcomes depending on the index under consideration. Notably, significant inter-model variability is evident for certain indices, whose interquantile range encompasses trends of varying directions, and in some cases, even exhibits a trend opposite to that shown by the ERA5 data. This discrepancy is particularly noticeable for the R20mm, RR1, and RX1day indices, where the interquantile range showcases a trend contrary to that portrayed by ERA5. Meanwhile, the RX5day index displays an interquantile range centered around zero, with the trend sign dependent on the model. In contrast, the CWD index demonstrates a trend direction consistent with ERA5.\n\n- It is crucial to emphasise that the boxplots present spatially-averaged values, and their interpretation should be approached with caution, as the outcomes can vary significantly when considering different regions across Europe. For regional analyses, it is essential to focus solely on the region of interest, as the influence of other European regions may distort the results, leading to conclusions that do not accurately reflect the specific area under study.\n\n- Although ERA5 is used as our reference dataset, it is important to be aware of its inherent uncertainties. While uncertainties affect all variables, biases are particularly pronounced for variables such as surface wind and precipitation, as demonstrated in [[4]](https://doi.org/10.1002/qj.3616) and [[5]](https://doi.org/10.3390/atmos12111462). This underscores the importance of careful interpretation, especially regarding precipitation extremes.\n\n- A larger GCM-RCM matrix should be considered when addressing specific cases to enhance the robustness of the analysis and account for uncertainties [[6]](https://doi.org/10.1002/wcc.8)[[7]](https://doi.org/10.1088/1748-9326/aacc77).\n\n- What do the results mean for users? Are the biases relevant?\n\n- The subset of Regional Climate Models (RCMs) considered for this notebook and for this seasonal aggregation struggle to replicate both the observed climatology and the trends during the historical period, with particular difficulty in capturing the trends.\n\n- However, for specific regions and indices, the models successfully capture the climatology and even the direction of the trend. Examples include the increasing trend for the CWD index in the western part of the Iberian Peninsula and the positive trend for RR1 in central Europe. Users should consider these regional variations, the indices they aim to use, the differences in bias behavior between climatology and trends, and the necessity of bias-correcting data before using it for hydrological applications [[3]](https://doi.org/10.1016/j.jhydrol.2012.05.052).\n\n- Precipitation trends in observational and reanalysis datasets are generally less robust and more challenging to detect compared to temperature trends. This limitation is also evident in model simulations.\n\n\n
\nRESULTS NOTE:
\nIt is important to note that the results presented are specific to the 9 models chosen, and users should aim to assess as wide a range of models as possible before making a sub-selection."} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q03__f5f515e25dc1", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q03", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q03", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Bias in precipitation-based indices for impact models > ℹ️ If you want to know more > Key resources", "title": "Bias in precipitation-based indices for impact models", "chunk_index": 19, "token_count": 274, "text_raw": "Some key resources and further reading were linked throughout this assessment.\n\nThe CDS catalogue entries for the data used were:\n* CORDEX regional climate model data on single levels (Daily mean - Mean precipitation flux): https://cds.climate.copernicus.eu/datasets/projections-cordex-domains-single-levels?tab=overview\n* ERA5 hourly data on single levels from 1940 to present (Total precipitation): https://cds.climate.copernicus.eu/datasets/reanalysis-era5-single-levels?tab=overview\n\nCode libraries used:\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)\n* [icclim](https://icclim.readthedocs.io/en/stable/) Python package", "text_with_prefix": "EQC Quality Assessment: \"Bias in precipitation-based indices for impact models\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q03 | Category: Climate_Projections\nSection: Bias in precipitation-based indices for impact models > ℹ️ If you want to know more > Key resources\n---\nSome key resources and further reading were linked throughout this assessment.\n\nThe CDS catalogue entries for the data used were:\n* CORDEX regional climate model data on single levels (Daily mean - Mean precipitation flux): https://cds.climate.copernicus.eu/datasets/projections-cordex-domains-single-levels?tab=overview\n* ERA5 hourly data on single levels from 1940 to present (Total precipitation): https://cds.climate.copernicus.eu/datasets/reanalysis-era5-single-levels?tab=overview\n\nCode libraries used:\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)\n* [icclim](https://icclim.readthedocs.io/en/stable/) Python package"} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q03__8747c7c9c070", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q03", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q03", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Bias in precipitation-based indices for impact models > ℹ️ If you want to know more > References", "title": "Bias in precipitation-based indices for impact models", "chunk_index": 20, "token_count": 720, "text_raw": "[[1]](https://doi.org/10.1016/j.envsci.2015.08.012) Panagos, P., Borrelli, P., Poesen, J., Ballabio, C., Lugato, E., Meusburger, K., Montanarella, L., Alewell, C. (2015). The new assessment of soil loss by water erosion in Europe. Environ. Sci. Policy, 54, pp. 438-447. https://doi.org/10.1016/j.envsci.2015.08.012\n\n[[2]](https://doi.org/10.1016/j.landusepol.2012.11.007) Salvati, L., Carlucci, M. (2013). The impact of mediterranean land degradation on agricultural income: a short-term scenario. Land Use Policy, 32, pp. 302-308. https://doi.org/10.1016/j.landusepol.2012.11.007\n\n[[3]](https://doi.org/10.1016/j.jhydrol.2012.05.052) Teutschbein, C., Seibert, J. (2012). Bias correction of regional climate model simulations for hydrological climate-change impact studies: Review and evaluation of different methods. Journal of Hydrology,\nVolumes 456–457, pp. 12-29. https://doi.org/10.1016/j.jhydrol.2012.05.052\n\n[[4]](https://doi.org/10.1002/qj.3616) Ramon, J., Lledó L., Torralba, V., Soret, A., Doblas-Reyes, F.J. (2019). What global reanalysis best represents near-surface winds?. Q J R Meteorol Soc. 2019; 145: 3236–3251. https://doi.org/10.1002/qj.3616\n\n[[5]](https://doi.org/10.3390/atmos12111462) Hassler, B. and Lauer, A. (2021). Comparison of Reanalysis and Observational Precipitation Datasets Including ERA5 and WFDE5. Atmosphere, 12, 1462. https://doi.org/10.3390/atmos12111462\n\n[[6]](https://doi.org/10.1002/wcc.8) Rummukainen, M. (2010). State-of-the-art with regional climate models. WIREs Clim Change, 1: 82-96. https://doi.org/10.1002/wcc.8\n\n[[7]](https://doi.org/10.1088/1748-9326/aacc77) Silje Lund Sørland et al. (2018). Bias patterns and climate change signals in GCM-RCM model chains. Environ. Res. Lett. 13 074017. https://doi.org/10.1088/1748-9326/aacc77", "text_with_prefix": "EQC Quality Assessment: \"Bias in precipitation-based indices for impact models\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q03 | Category: Climate_Projections\nSection: Bias in precipitation-based indices for impact models > ℹ️ If you want to know more > References\n---\n[[1]](https://doi.org/10.1016/j.envsci.2015.08.012) Panagos, P., Borrelli, P., Poesen, J., Ballabio, C., Lugato, E., Meusburger, K., Montanarella, L., Alewell, C. (2015). The new assessment of soil loss by water erosion in Europe. Environ. Sci. Policy, 54, pp. 438-447. https://doi.org/10.1016/j.envsci.2015.08.012\n\n[[2]](https://doi.org/10.1016/j.landusepol.2012.11.007) Salvati, L., Carlucci, M. (2013). The impact of mediterranean land degradation on agricultural income: a short-term scenario. Land Use Policy, 32, pp. 302-308. https://doi.org/10.1016/j.landusepol.2012.11.007\n\n[[3]](https://doi.org/10.1016/j.jhydrol.2012.05.052) Teutschbein, C., Seibert, J. (2012). Bias correction of regional climate model simulations for hydrological climate-change impact studies: Review and evaluation of different methods. Journal of Hydrology,\nVolumes 456–457, pp. 12-29. https://doi.org/10.1016/j.jhydrol.2012.05.052\n\n[[4]](https://doi.org/10.1002/qj.3616) Ramon, J., Lledó L., Torralba, V., Soret, A., Doblas-Reyes, F.J. (2019). What global reanalysis best represents near-surface winds?. Q J R Meteorol Soc. 2019; 145: 3236–3251. https://doi.org/10.1002/qj.3616\n\n[[5]](https://doi.org/10.3390/atmos12111462) Hassler, B. and Lauer, A. (2021). Comparison of Reanalysis and Observational Precipitation Datasets Including ERA5 and WFDE5. Atmosphere, 12, 1462. https://doi.org/10.3390/atmos12111462\n\n[[6]](https://doi.org/10.1002/wcc.8) Rummukainen, M. (2010). State-of-the-art with regional climate models. WIREs Clim Change, 1: 82-96. https://doi.org/10.1002/wcc.8\n\n[[7]](https://doi.org/10.1088/1748-9326/aacc77) Silje Lund Sørland et al. (2018). Bias patterns and climate change signals in GCM-RCM model chains. Environ. Res. Lett. 13 074017. https://doi.org/10.1088/1748-9326/aacc77"} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q04__040aef108a65", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q04", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q04", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Projected changes in precipitation-based indices for impact models", "title": "Projected changes in precipitation-based indices for impact models", "chunk_index": 0, "token_count": 99, "text_raw": "Production date: 29-05-2024\n\nProduced by: CMCC foundation - Euro-Mediterranean Center on Climate Change. Albert Martinez Boti.", "text_with_prefix": "EQC Quality Assessment: \"Projected changes in precipitation-based indices for impact models\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q04 | Category: Climate_Projections\nSection: Projected changes in precipitation-based indices for impact models\n---\nProduction date: 29-05-2024\n\nProduced by: CMCC foundation - Euro-Mediterranean Center on Climate Change. Albert Martinez Boti."} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q04__9fc2832f015d", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q04", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q04", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Projected changes in precipitation-based indices for impact models > Quality assessment question", "title": "Projected changes in precipitation-based indices for impact models", "chunk_index": 1, "token_count": 558, "text_raw": "* **What are the projected future changes and associated uncertainties of precipitation-based indices in Europe?**\n\nSoil erosion stands as one of the primary environmental concerns in Europe [[1]](https://doi.org/10.1016/j.envsci.2015.08.012). Its accelerated occurrence can precipitate a decline in ecosystem stability, land productivity, and overall land degradation, resulting in diminished income for farmers [[2]](https://doi.org/10.1016/j.landusepol.2012.11.007). This notebook is designed to evaluate the uncertainty in future projections of a specific set of models from **[CORDEX](https://cds.climate.copernicus.eu/datasets/projections-cordex-domains-single-levels?tab=overview)** Regional Climate Models (RCMs) by considering the ensemble inter-model spread of projected changes. These selected rainfall indicators are recognised as valuable proxies for rainfall erosivity, a factor directly linked to empirical calculations of soil loss, as outlined in an application available through the old CDS platform ([dataset documentation](https://dast.copernicus-climate.eu/documents/sis-soil-erosion/C3S_D429d.2.3.2_Product_User_Guide_v1.2.pdf)) The five precipitation-based indices used here are calculated using the **[icclim](https://icclim.readthedocs.io/en/stable/)** package, them being:\n\n- Spell length of days with precipitation greater than 1 mm (also known as \"Maximum consecutive wet days\" - \"CWD\")\n- Number of heavy precipitation days (Precip >= 20mm - \"R20mm\")\n- Number of wet days (Precip >= 1mm - \"RR1\")\n- Maximum 1-day total precipitation (\"RX1day\")\n- Maximum 5-day total precipitation (\"RX5day\").\n\nWithin this notebook, these calculations are performed over the temporal aggregation of DJF and for the future period spanning from 2015 to 2099, following the Representative Concentration Pathway RCP 8.5. It is important to note that the results presented here pertain to a specific subset of the CORDEX ensemble and may not be generalisable to the entire dataset. Also note that a separate assessment examines the representation of climatology and trends of these indices for the same models during the historical period (1971-2000).", "text_with_prefix": "EQC Quality Assessment: \"Projected changes in precipitation-based indices for impact models\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q04 | Category: Climate_Projections\nSection: Projected changes in precipitation-based indices for impact models > Quality assessment question\n---\n* **What are the projected future changes and associated uncertainties of precipitation-based indices in Europe?**\n\nSoil erosion stands as one of the primary environmental concerns in Europe [[1]](https://doi.org/10.1016/j.envsci.2015.08.012). Its accelerated occurrence can precipitate a decline in ecosystem stability, land productivity, and overall land degradation, resulting in diminished income for farmers [[2]](https://doi.org/10.1016/j.landusepol.2012.11.007). This notebook is designed to evaluate the uncertainty in future projections of a specific set of models from **[CORDEX](https://cds.climate.copernicus.eu/datasets/projections-cordex-domains-single-levels?tab=overview)** Regional Climate Models (RCMs) by considering the ensemble inter-model spread of projected changes. These selected rainfall indicators are recognised as valuable proxies for rainfall erosivity, a factor directly linked to empirical calculations of soil loss, as outlined in an application available through the old CDS platform ([dataset documentation](https://dast.copernicus-climate.eu/documents/sis-soil-erosion/C3S_D429d.2.3.2_Product_User_Guide_v1.2.pdf)) The five precipitation-based indices used here are calculated using the **[icclim](https://icclim.readthedocs.io/en/stable/)** package, them being:\n\n- Spell length of days with precipitation greater than 1 mm (also known as \"Maximum consecutive wet days\" - \"CWD\")\n- Number of heavy precipitation days (Precip >= 20mm - \"R20mm\")\n- Number of wet days (Precip >= 1mm - \"RR1\")\n- Maximum 1-day total precipitation (\"RX1day\")\n- Maximum 5-day total precipitation (\"RX5day\").\n\nWithin this notebook, these calculations are performed over the temporal aggregation of DJF and for the future period spanning from 2015 to 2099, following the Representative Concentration Pathway RCP 8.5. It is important to note that the results presented here pertain to a specific subset of the CORDEX ensemble and may not be generalisable to the entire dataset. Also note that a separate assessment examines the representation of climatology and trends of these indices for the same models during the historical period (1971-2000)."} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q04__676a4f069e1e", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q04", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q04", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Projected changes in precipitation-based indices for impact models > Quality assessment statement", "title": "Projected changes in precipitation-based indices for impact models", "chunk_index": 2, "token_count": 715, "text_raw": "These are the key outcomes of this assessment\n\n* For the selected subset of models, future projections suggest diversity on the sign of the trend of precipitation-based indices for the DJF aggregation depending on the index selection and the region under consideration.\n\n* The spatially-averaged values displayed by boxplots illustrate the diversity of outcomes, emphasising the importance of regional analyses. There is a significant inter-model variability, suggesting uncertainties in future projections. It is essential to note that the interpretation of boxplot results should be approached cautiously, considering the potential distortion of outcomes due to the influence of regions with trends that behave differently.\n\n* A larger GCM-RCM matrix should be considered when addressing specific cases to enhance the robustness of the analysis and account for uncertainties [[3]](https://doi.org/10.1002/wcc.8)[[4]](https://doi.org/10.1088/1748-9326/aacc77).\n\n* The outcomes of this notebook are highly dependent on the index and region. Projected changes for RR1, RX1day, and RX5day show similar spatial patterns among the subset of considered models. In contrast, CWD and R20mm exhibit more scattered spatial patterns, and the regions with significant trend signals appear to have large uncertainties (inter-model spread). Results obtained from another assessment of the same indices and temporal aggregation (DJF) for the historical period 1971-2000, which evaluated biases (\"CORDEX Climate Projections: evaluating bias in precipitation-based indices for impact models\"), should also be taken into consideration. The high uncertainties in both assessments encourage users to bias-correct precipitation-based indices before using them for specific applications [[5]](https://doi.org/10.1016/j.jhydrol.2012.05.052).\n\n* Model agreement on precipitation-based indices projections is lower than on maximum temperature indices. This difference is evident when comparing the results of this assessment with those obtained in \"RCM uncertainty in projected changes in extreme temperature indices for the reinsurance sector\"\n```\n\nattachment:13536fdf-aabd-48d2-b1a4-3c84b6fb58ac.png\n---\nalt: trend_future_RX5day\nwidth: 900px\n---\nMaximum 5-day total precipitation ('RX5day') for the temporal aggregation of 'DJF'. Trend for the future period (2015-2099). The layout includes data corresponding to: (a) the ensemble median (understood as the median of the trend values of the chosen subset of models calculated for each grid cell) and (b) the ensemble spread (derived as the standard deviation of the distribution of the chosen subset of models).\n```\n\n\n
\nNOTE on the DJF selection:
\nIt is important to note that using seasonal temporal aggregations offers only partial insights into the dynamics. For more comprehensive results, it is advisable to also consider other seasons and annual aggregations. However, for the sake of efficiency and to avoid making the notebook too heavy, we have opted to prioritise seasonal aggregations.", "text_with_prefix": "EQC Quality Assessment: \"Projected changes in precipitation-based indices for impact models\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q04 | Category: Climate_Projections\nSection: Projected changes in precipitation-based indices for impact models > Quality assessment statement\n---\nThese are the key outcomes of this assessment\n\n* For the selected subset of models, future projections suggest diversity on the sign of the trend of precipitation-based indices for the DJF aggregation depending on the index selection and the region under consideration.\n\n* The spatially-averaged values displayed by boxplots illustrate the diversity of outcomes, emphasising the importance of regional analyses. There is a significant inter-model variability, suggesting uncertainties in future projections. It is essential to note that the interpretation of boxplot results should be approached cautiously, considering the potential distortion of outcomes due to the influence of regions with trends that behave differently.\n\n* A larger GCM-RCM matrix should be considered when addressing specific cases to enhance the robustness of the analysis and account for uncertainties [[3]](https://doi.org/10.1002/wcc.8)[[4]](https://doi.org/10.1088/1748-9326/aacc77).\n\n* The outcomes of this notebook are highly dependent on the index and region. Projected changes for RR1, RX1day, and RX5day show similar spatial patterns among the subset of considered models. In contrast, CWD and R20mm exhibit more scattered spatial patterns, and the regions with significant trend signals appear to have large uncertainties (inter-model spread). Results obtained from another assessment of the same indices and temporal aggregation (DJF) for the historical period 1971-2000, which evaluated biases (\"CORDEX Climate Projections: evaluating bias in precipitation-based indices for impact models\"), should also be taken into consideration. The high uncertainties in both assessments encourage users to bias-correct precipitation-based indices before using them for specific applications [[5]](https://doi.org/10.1016/j.jhydrol.2012.05.052).\n\n* Model agreement on precipitation-based indices projections is lower than on maximum temperature indices. This difference is evident when comparing the results of this assessment with those obtained in \"RCM uncertainty in projected changes in extreme temperature indices for the reinsurance sector\"\n```\n\nattachment:13536fdf-aabd-48d2-b1a4-3c84b6fb58ac.png\n---\nalt: trend_future_RX5day\nwidth: 900px\n---\nMaximum 5-day total precipitation ('RX5day') for the temporal aggregation of 'DJF'. Trend for the future period (2015-2099). The layout includes data corresponding to: (a) the ensemble median (understood as the median of the trend values of the chosen subset of models calculated for each grid cell) and (b) the ensemble spread (derived as the standard deviation of the distribution of the chosen subset of models).\n```\n\n\n
\nNOTE on the DJF selection:
\nIt is important to note that using seasonal temporal aggregations offers only partial insights into the dynamics. For more comprehensive results, it is advisable to also consider other seasons and annual aggregations. However, for the sake of efficiency and to avoid making the notebook too heavy, we have opted to prioritise seasonal aggregations."} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q04__8746203f35de", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q04", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q04", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Projected changes in precipitation-based indices for impact models > Methodology", "title": "Projected changes in precipitation-based indices for impact models", "chunk_index": 3, "token_count": 472, "text_raw": "This notebook offers an assessment of the projected changes and their associated uncertainties using a subset of 9 models from [CORDEX](https://cds.climate.copernicus.eu/datasets/projections-cordex-domains-single-levels?tab=overview). The analysis involves evaluating the ensemble inter-model spread of projected changes using several precipitation-based indices. These indices, calculated over the DJF period for the future period spanning from 2015 to 2099, include:\n\n- Spell length of days with precipitation greater than 1 mm (also known as \"Maximum consecutive wet days\" - \"CWD\")\n- Number of heavy precipitation days (Precip >= 20mm - \"R20mm\")\n- Number of wet days (Precip >= 1mm - \"RR1\")\n- Maximum 1-day total precipitation (\"RX1day\")\n- Maximum 5-day total precipitation (\"RX5day\").\n\nIn particular, spatial patterns of climate projected trends are examined and displayed for each model individually and for the ensemble median (calculated for each grid cell), alongside the ensemble inter-model spread to account for projected uncertainty. Additionally, spatially-averaged trend values are analysed and presented using box plots to provide an overview of trend behavior across the distribution of the chosen subset of models when averaged across Europe.\n\nThe analysis and results follow the next outline:\n\n**[](section-1)**\n * [](section-1.1)\n * [](section-1.2)\n * [](section-1.3)\n * [](section-1.4)\n * [](section-1.5)\n * [](section-1.6)\n\n**[](section-2)**\n * [](section-2.1)\n * [](section-2.2)\n * [](section-2.3)\n\n**[](section-3)**\n * [](section-3.1)\n * [](section-3.2)\n * [](section-3.3)\n * [](section-3.4)\n * [](section-3.5)", "text_with_prefix": "EQC Quality Assessment: \"Projected changes in precipitation-based indices for impact models\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q04 | Category: Climate_Projections\nSection: Projected changes in precipitation-based indices for impact models > Methodology\n---\nThis notebook offers an assessment of the projected changes and their associated uncertainties using a subset of 9 models from [CORDEX](https://cds.climate.copernicus.eu/datasets/projections-cordex-domains-single-levels?tab=overview). The analysis involves evaluating the ensemble inter-model spread of projected changes using several precipitation-based indices. These indices, calculated over the DJF period for the future period spanning from 2015 to 2099, include:\n\n- Spell length of days with precipitation greater than 1 mm (also known as \"Maximum consecutive wet days\" - \"CWD\")\n- Number of heavy precipitation days (Precip >= 20mm - \"R20mm\")\n- Number of wet days (Precip >= 1mm - \"RR1\")\n- Maximum 1-day total precipitation (\"RX1day\")\n- Maximum 5-day total precipitation (\"RX5day\").\n\nIn particular, spatial patterns of climate projected trends are examined and displayed for each model individually and for the ensemble median (calculated for each grid cell), alongside the ensemble inter-model spread to account for projected uncertainty. Additionally, spatially-averaged trend values are analysed and presented using box plots to provide an overview of trend behavior across the distribution of the chosen subset of models when averaged across Europe.\n\nThe analysis and results follow the next outline:\n\n**[](section-1)**\n * [](section-1.1)\n * [](section-1.2)\n * [](section-1.3)\n * [](section-1.4)\n * [](section-1.5)\n * [](section-1.6)\n\n**[](section-2)**\n * [](section-2.1)\n * [](section-2.2)\n * [](section-2.3)\n\n**[](section-3)**\n * [](section-3.1)\n * [](section-3.2)\n * [](section-3.3)\n * [](section-3.4)\n * [](section-3.5)"} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q04__2b197440b069", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q04", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q04", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Projected changes in precipitation-based indices for impact models > Analysis and results > 1. Parameters, requests and functions definition > 1.2. Define Parameters", "title": "Projected changes in precipitation-based indices for impact models", "chunk_index": 4, "token_count": 369, "text_raw": "In the \"Define Parameters\" section, various customisable options for the notebook are specified. Most of the parameters chosen are the same as those used in another assessment (\"CORDEX Climate Projections: evaluating bias in precipitation-based indices for impact models\"), being them:\n\n- The initial and ending year used for the future projections period can be specified by changing the parametes `future_slice` (2015-2099 is chosen for consistency between CORDEX and CMIP6).\n- The `timeseries` set the temporal aggregation. For instance, selecting \"DJF\" implies considering only the winter season.\n- `collection_id` provides the choice between Global Climate Models CMIP6 or Regional Climate Models CORDEX. Although the code allows choosing between CMIP6 or CORDEX, the example provided in this notebook deals with CORDEX RCMs.\n- `area` allows specifying the geographical domain of interest.\n- The `interpolation_method` parameter allows selecting the interpolation method when regridding is performed over the indices.\n- The `chunk` selection allows the user to define if dividing into chunks when downloading the data on their local machine. Although it does not significantly affect the analysis, it is recommended to keep the default value for optimal performance.\n\nChoose annual or seasonal timeseries\nChoose CORDEX or CMIP6\nInterpolation method\nArea to show\nChunks for download\n\n(section-1.3)=", "text_with_prefix": "EQC Quality Assessment: \"Projected changes in precipitation-based indices for impact models\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q04 | Category: Climate_Projections\nSection: Projected changes in precipitation-based indices for impact models > Analysis and results > 1. Parameters, requests and functions definition > 1.2. Define Parameters\n---\nIn the \"Define Parameters\" section, various customisable options for the notebook are specified. Most of the parameters chosen are the same as those used in another assessment (\"CORDEX Climate Projections: evaluating bias in precipitation-based indices for impact models\"), being them:\n\n- The initial and ending year used for the future projections period can be specified by changing the parametes `future_slice` (2015-2099 is chosen for consistency between CORDEX and CMIP6).\n- The `timeseries` set the temporal aggregation. For instance, selecting \"DJF\" implies considering only the winter season.\n- `collection_id` provides the choice between Global Climate Models CMIP6 or Regional Climate Models CORDEX. Although the code allows choosing between CMIP6 or CORDEX, the example provided in this notebook deals with CORDEX RCMs.\n- `area` allows specifying the geographical domain of interest.\n- The `interpolation_method` parameter allows selecting the interpolation method when regridding is performed over the indices.\n- The `chunk` selection allows the user to define if dividing into chunks when downloading the data on their local machine. Although it does not significantly affect the analysis, it is recommended to keep the default value for optimal performance.\n\nChoose annual or seasonal timeseries\nChoose CORDEX or CMIP6\nInterpolation method\nArea to show\nChunks for download\n\n(section-1.3)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q04__a844b991c6ef", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q04", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q04", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Projected changes in precipitation-based indices for impact models > Analysis and results > 1. Parameters, requests and functions definition > 1.3. Define models", "title": "Projected changes in precipitation-based indices for impact models", "chunk_index": 5, "token_count": 346, "text_raw": "he following climate analyses are performed considering a subset of GCMs from CMIP6. Models names are listed in the parameters below. Some variable-dependent parameters are also selected, as the `index_names` parameter, which specifies the precipitation-based indices ('CWD', 'R20mm', 'RR1', 'RX1day' and 'RX5day' in our case) from the [icclim](https://icclim.readthedocs.io/en/stable/) Python package.\n\nWhen choosing Cordex models, it is crucial to consider the availability of RCMs for the selected GCM and the specified region. The listed RCMs, for instance, are accessible for the GCM “mpi_m_mpi_esm_lr” in the “europe” cordex_domain, and they are the same as those used in another assessment (\"CORDEX Climate Projections: evaluating bias in precipitation-based indices for impact models\"). To confirm the available combinations, refer to the [CORDEX](https://cds.climate.copernicus.eu/datasets/projections-cordex-domains-single-levels?tab=overview) CDS catalogue entry.\n\nDefine dictionaries to use in titles and caption\nDefine dictionaries to use in titles and caption\n\n(section-1.4)=", "text_with_prefix": "EQC Quality Assessment: \"Projected changes in precipitation-based indices for impact models\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q04 | Category: Climate_Projections\nSection: Projected changes in precipitation-based indices for impact models > Analysis and results > 1. Parameters, requests and functions definition > 1.3. Define models\n---\nhe following climate analyses are performed considering a subset of GCMs from CMIP6. Models names are listed in the parameters below. Some variable-dependent parameters are also selected, as the `index_names` parameter, which specifies the precipitation-based indices ('CWD', 'R20mm', 'RR1', 'RX1day' and 'RX5day' in our case) from the [icclim](https://icclim.readthedocs.io/en/stable/) Python package.\n\nWhen choosing Cordex models, it is crucial to consider the availability of RCMs for the selected GCM and the specified region. The listed RCMs, for instance, are accessible for the GCM “mpi_m_mpi_esm_lr” in the “europe” cordex_domain, and they are the same as those used in another assessment (\"CORDEX Climate Projections: evaluating bias in precipitation-based indices for impact models\"). To confirm the available combinations, refer to the [CORDEX](https://cds.climate.copernicus.eu/datasets/projections-cordex-domains-single-levels?tab=overview) CDS catalogue entry.\n\nDefine dictionaries to use in titles and caption\nDefine dictionaries to use in titles and caption\n\n(section-1.4)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q04__297c3d340e6c", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q04", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q04", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Projected changes in precipitation-based indices for impact models > Analysis and results > 1. Parameters, requests and functions definition > 1.4. Define land-sea mask request", "title": "Projected changes in precipitation-based indices for impact models", "chunk_index": 6, "token_count": 139, "text_raw": "Within this notebook, ERA5 will be used to download the land-sea mask when plotting. In this section, we set the required parameters for the cds-api data-request of ERA5 land-sea mask.\n\n(section-1.5)=", "text_with_prefix": "EQC Quality Assessment: \"Projected changes in precipitation-based indices for impact models\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q04 | Category: Climate_Projections\nSection: Projected changes in precipitation-based indices for impact models > Analysis and results > 1. Parameters, requests and functions definition > 1.4. Define land-sea mask request\n---\nWithin this notebook, ERA5 will be used to download the land-sea mask when plotting. In this section, we set the required parameters for the cds-api data-request of ERA5 land-sea mask.\n\n(section-1.5)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q04__331dc90b5296", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q04", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q04", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Projected changes in precipitation-based indices for impact models > Analysis and results > 1. Parameters, requests and functions definition > 1.5. Define model requests", "title": "Projected changes in precipitation-based indices for impact models", "chunk_index": 7, "token_count": 215, "text_raw": "In this section we set the required parameters for the cds-api data-request.\n\nThe `get_cordex_years` function is employed to choose suitable data chunks for CORDEX data requests.\n\nWhen `Weights = True`, spatial weighting is applied for calculations requiring spatial data aggregation. This is particularly relevant for CMIP6 GCMs with regular lon-lat grids that do not consider varying surface extensions at different latitudes. In contrast, CORDEX RCMs, using rotated grids, inherently account for different cell surfaces based on latitude, eliminating the need for a latitude cosine multiplicative factor (`Weights = False`).\n\n(section-1.6)=", "text_with_prefix": "EQC Quality Assessment: \"Projected changes in precipitation-based indices for impact models\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q04 | Category: Climate_Projections\nSection: Projected changes in precipitation-based indices for impact models > Analysis and results > 1. Parameters, requests and functions definition > 1.5. Define model requests\n---\nIn this section we set the required parameters for the cds-api data-request.\n\nThe `get_cordex_years` function is employed to choose suitable data chunks for CORDEX data requests.\n\nWhen `Weights = True`, spatial weighting is applied for calculations requiring spatial data aggregation. This is particularly relevant for CMIP6 GCMs with regular lon-lat grids that do not consider varying surface extensions at different latitudes. In contrast, CORDEX RCMs, using rotated grids, inherently account for different cell surfaces based on latitude, eliminating the need for a latitude cosine multiplicative factor (`Weights = False`).\n\n(section-1.6)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q04__6ba3da262d97", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q04", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q04", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Projected changes in precipitation-based indices for impact models > Analysis and results > 1. Parameters, requests and functions definition > 1.6. Functions to cache", "title": "Projected changes in precipitation-based indices for impact models", "chunk_index": 8, "token_count": 291, "text_raw": "In this section, functions that will be executed in the caching phase are defined. Caching is the process of storing copies of files in a temporary storage location, so that they can be accessed more quickly. This process also checks if the user has already downloaded a file, avoiding redundant downloads.\n\nFunctions description:\n\n- The `select_timeseries` function subsets the dataset based on the chosen `timeseries` parameter.\n\n- The `compute_indices` function utilises the icclim package to calculate the precipitation-based indices.\n\n- The `compute_trends` function employs the Mann-Kendall test for trend calculation.\n\n- Finally, the `compute_indices_and_trends` function calculates the precipitation-based indices for the corresponding temporal aggregation using the `compute_indices` function, determines the indices mean for the future period (2015-2099), obtain the trends using the `compute_trends` function, and offers an option for regridding to `model_regrid`.\n\nOriginal bounds for conservative interpolation\n\n(section-2)=", "text_with_prefix": "EQC Quality Assessment: \"Projected changes in precipitation-based indices for impact models\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q04 | Category: Climate_Projections\nSection: Projected changes in precipitation-based indices for impact models > Analysis and results > 1. Parameters, requests and functions definition > 1.6. Functions to cache\n---\nIn this section, functions that will be executed in the caching phase are defined. Caching is the process of storing copies of files in a temporary storage location, so that they can be accessed more quickly. This process also checks if the user has already downloaded a file, avoiding redundant downloads.\n\nFunctions description:\n\n- The `select_timeseries` function subsets the dataset based on the chosen `timeseries` parameter.\n\n- The `compute_indices` function utilises the icclim package to calculate the precipitation-based indices.\n\n- The `compute_trends` function employs the Mann-Kendall test for trend calculation.\n\n- Finally, the `compute_indices_and_trends` function calculates the precipitation-based indices for the corresponding temporal aggregation using the `compute_indices` function, determines the indices mean for the future period (2015-2099), obtain the trends using the `compute_trends` function, and offers an option for regridding to `model_regrid`.\n\nOriginal bounds for conservative interpolation\n\n(section-2)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q04__fcb46a497019", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q04", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q04", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Projected changes in precipitation-based indices for impact models > Analysis and results > 2. Downloading and processing > 2.1. Download and transform the regridding model", "title": "Projected changes in precipitation-based indices for impact models", "chunk_index": 9, "token_count": 281, "text_raw": "In this section, the `download.download_and_transform` function from the 'c3s_eqc_automatic_quality_control' package is employed to download daily data from the selected CORDEX regridding model, compute the precipitation-based indices for the selected temporal aggregation, calculate the mean and trend over the future projections period (2015-2099), and cache the result (to avoid redundant downloads and processing).\n\nThe regridding model is intended here as the model whose grid will be used to interpolate the others. This ensures all models share a common grid, facilitating the calculation of median values for each cell point. The regridding model within this notebook is \"clmcom_eth_cosmo_crclim\" but a different one can be selected by just modifying the model_regrid parameter at [](section-1.3). It is key to highlight the importance of the chosen target grid depending on the specific application.\n\n(section-2.2)=", "text_with_prefix": "EQC Quality Assessment: \"Projected changes in precipitation-based indices for impact models\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q04 | Category: Climate_Projections\nSection: Projected changes in precipitation-based indices for impact models > Analysis and results > 2. Downloading and processing > 2.1. Download and transform the regridding model\n---\nIn this section, the `download.download_and_transform` function from the 'c3s_eqc_automatic_quality_control' package is employed to download daily data from the selected CORDEX regridding model, compute the precipitation-based indices for the selected temporal aggregation, calculate the mean and trend over the future projections period (2015-2099), and cache the result (to avoid redundant downloads and processing).\n\nThe regridding model is intended here as the model whose grid will be used to interpolate the others. This ensures all models share a common grid, facilitating the calculation of median values for each cell point. The regridding model within this notebook is \"clmcom_eth_cosmo_crclim\" but a different one can be selected by just modifying the model_regrid parameter at [](section-1.3). It is key to highlight the importance of the chosen target grid depending on the specific application.\n\n(section-2.2)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q04__5948492b68f1", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q04", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q04", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Projected changes in precipitation-based indices for impact models > Analysis and results > 2. Downloading and processing > 2.2. Download and transform models", "title": "Projected changes in precipitation-based indices for impact models", "chunk_index": 10, "token_count": 315, "text_raw": "In this section, the `download.download_and_transform` function from the 'c3s_eqc_automatic_quality_control' package is employed to download daily data from the CORDDEX models, compute the precipitation-based indices for the selected temporal aggregation, calculate the mean and trend over the future period (2015-2099), interpolate to the regridding model's grid (only for the cases in which it is specified, in the other cases, the original model's grid is mantained), and cache the result (to avoid redundant downloads and processing).\n\nOriginal model\nInterpolated model\n\n```text\nmodel='clmcom_clm_cclm4_8_17'\nmodel='clmcom_eth_cosmo_crclim'\nmodel='cnrm_aladin63'\nmodel='dmi_hirham5'\nmodel='knmi_racmo22e'\nmodel='mohc_hadrem3_ga7_05'\nmodel='mpi_csc_remo2009'\nmodel='smhi_rca4'\nmodel='uhoh_wrf361h'\n```\n\n(section-2.3)=", "text_with_prefix": "EQC Quality Assessment: \"Projected changes in precipitation-based indices for impact models\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q04 | Category: Climate_Projections\nSection: Projected changes in precipitation-based indices for impact models > Analysis and results > 2. Downloading and processing > 2.2. Download and transform models\n---\nIn this section, the `download.download_and_transform` function from the 'c3s_eqc_automatic_quality_control' package is employed to download daily data from the CORDDEX models, compute the precipitation-based indices for the selected temporal aggregation, calculate the mean and trend over the future period (2015-2099), interpolate to the regridding model's grid (only for the cases in which it is specified, in the other cases, the original model's grid is mantained), and cache the result (to avoid redundant downloads and processing).\n\nOriginal model\nInterpolated model\n\n```text\nmodel='clmcom_clm_cclm4_8_17'\nmodel='clmcom_eth_cosmo_crclim'\nmodel='cnrm_aladin63'\nmodel='dmi_hirham5'\nmodel='knmi_racmo22e'\nmodel='mohc_hadrem3_ga7_05'\nmodel='mpi_csc_remo2009'\nmodel='smhi_rca4'\nmodel='uhoh_wrf361h'\n```\n\n(section-2.3)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q04__010e9ce99bfa", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q04", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q04", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Projected changes in precipitation-based indices for impact models > Analysis and results > 2. Downloading and processing > 2.3. Apply land-sea mask, change attributes and cut the region to show", "title": "Projected changes in precipitation-based indices for impact models", "chunk_index": 11, "token_count": 235, "text_raw": "This section performs the following tasks:\n\n1. Cut the region of interest.\n2. Downloads the sea mask for ERA5.\n3. Regrids ERA5's mask to the `model_regrid` grid and applies it to the regridded data\n4. Regrids the ERA5 land-sea mask to the model's original grid and applies it to them.\n5. Change some variable attributes for plotting purposes.\n\n**Note:** `ds_interpolated` contains data from the models regridded to the regridding model's grid. `model_datasets` contain the same data but in the original grid of each model.\n\nEdit attributes\n\n(section-3)=", "text_with_prefix": "EQC Quality Assessment: \"Projected changes in precipitation-based indices for impact models\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q04 | Category: Climate_Projections\nSection: Projected changes in precipitation-based indices for impact models > Analysis and results > 2. Downloading and processing > 2.3. Apply land-sea mask, change attributes and cut the region to show\n---\nThis section performs the following tasks:\n\n1. Cut the region of interest.\n2. Downloads the sea mask for ERA5.\n3. Regrids ERA5's mask to the `model_regrid` grid and applies it to the regridded data\n4. Regrids the ERA5 land-sea mask to the model's original grid and applies it to them.\n5. Change some variable attributes for plotting purposes.\n\n**Note:** `ds_interpolated` contains data from the models regridded to the regridding model's grid. `model_datasets` contain the same data but in the original grid of each model.\n\nEdit attributes\n\n(section-3)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q04__15c269419439", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q04", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q04", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Projected changes in precipitation-based indices for impact models > Analysis and results > 3. Plot and describe results", "title": "Projected changes in precipitation-based indices for impact models", "chunk_index": 12, "token_count": 214, "text_raw": "This section will display the following results:\n\n- Maps representing the spatial distribution of the **future trends** (2015-2099) of the indices ('CWD', 'R20mm', 'RR1', 'RX1day' and 'RX5day') for each model individually, the ensemble median (understood as the median of the trend values of the chosen subset of models calculated for each grid cell), and the ensemble spread (derived as the standard deviation of the distribution of the chosen subset of models).\n- **Boxplots** which represent statistical distributions (PDF) built on the spatially-averaged future trend from each considered model.\n\n(section-3.1)=", "text_with_prefix": "EQC Quality Assessment: \"Projected changes in precipitation-based indices for impact models\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q04 | Category: Climate_Projections\nSection: Projected changes in precipitation-based indices for impact models > Analysis and results > 3. Plot and describe results\n---\nThis section will display the following results:\n\n- Maps representing the spatial distribution of the **future trends** (2015-2099) of the indices ('CWD', 'R20mm', 'RR1', 'RX1day' and 'RX5day') for each model individually, the ensemble median (understood as the median of the trend values of the chosen subset of models calculated for each grid cell), and the ensemble spread (derived as the standard deviation of the distribution of the chosen subset of models).\n- **Boxplots** which represent statistical distributions (PDF) built on the spatially-averaged future trend from each considered model.\n\n(section-3.1)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q04__bebc13baa0f0", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q04", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q04", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Projected changes in precipitation-based indices for impact models > Analysis and results > 3. Plot and describe results > 3.1. Define plotting functions", "title": "Projected changes in precipitation-based indices for impact models", "chunk_index": 13, "token_count": 646, "text_raw": "The functions presented here are used to plot the trends calculated over the future period (2015-2099) for each of the indices ('CWD', 'R20mm', 'RR1', 'RX1day' and 'RX5day').\n\nFor a selected index, two layout types will be displayed, depending on the chosen function:\n\n1. Layout including the ensemble median and the ensemble spread for the trend: `plot_ensemble()` is used.\n2. Layout including every model trend: `plot_models()` is employed.\n\n`trend==True` allows displaying trend values over the future period, while `trend==False` show mean values. In this notebook, which focuses on the future period, only trend values will be shown, and, consequently, `trend==True`. When the `trend` argument is set to True, regions with no significance are hatched. For individual models, a grid point is considered to have a statistically significant trend when the p-value is lower than 0.05 (in such cases, no hatching is shown). However, for determining trend significance for the ensemble median (understood as the median of the trend values of the chosen subset of models calculated for each grid cell), reliance is placed on agreement categories, following the advanced approach proposed in AR6 [IPCC](https://www.ipcc.ch/report/ar6/wg1/downloads/report/IPCC_AR6_WGI_Atlas.pdf) on pages 1945-1950. The `hatch_p_value_ensemble()` function is used to distinguish, for each grid point, between three possible cases:\n\n1. If more than 66% of the models are statistically significant (p-value < 0.05) and more than 80% of the models share the same sign, we consider the ensemble median trend to be statistically significant, and there is agreement on the sign. To represent this, no hatching is used.\n2. If less than 66% of the models are statistically significant, regardless of agreement on the sign of the trend, hatching is applied (indicating that the ensemble median trend is not statistically significant).\n3. If more than 66% of the models are statistically significant but less than 80% of the models share the same sign, we consider the ensemble median trend to be statistically significant, but there is no agreement on the sign of the trend. This is represented using crosses.\n\nDefine function to plot the caption of the figures (for the ensemble case)\nAdd caption to the figure\nAdd each line to the figure\nend captioning\nDefine function to plot the caption of the figures (for the individual models case)\nAdd caption to the figure\nAdd each line to the figure\n\n(section-3.2)=", "text_with_prefix": "EQC Quality Assessment: \"Projected changes in precipitation-based indices for impact models\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q04 | Category: Climate_Projections\nSection: Projected changes in precipitation-based indices for impact models > Analysis and results > 3. Plot and describe results > 3.1. Define plotting functions\n---\nThe functions presented here are used to plot the trends calculated over the future period (2015-2099) for each of the indices ('CWD', 'R20mm', 'RR1', 'RX1day' and 'RX5day').\n\nFor a selected index, two layout types will be displayed, depending on the chosen function:\n\n1. Layout including the ensemble median and the ensemble spread for the trend: `plot_ensemble()` is used.\n2. Layout including every model trend: `plot_models()` is employed.\n\n`trend==True` allows displaying trend values over the future period, while `trend==False` show mean values. In this notebook, which focuses on the future period, only trend values will be shown, and, consequently, `trend==True`. When the `trend` argument is set to True, regions with no significance are hatched. For individual models, a grid point is considered to have a statistically significant trend when the p-value is lower than 0.05 (in such cases, no hatching is shown). However, for determining trend significance for the ensemble median (understood as the median of the trend values of the chosen subset of models calculated for each grid cell), reliance is placed on agreement categories, following the advanced approach proposed in AR6 [IPCC](https://www.ipcc.ch/report/ar6/wg1/downloads/report/IPCC_AR6_WGI_Atlas.pdf) on pages 1945-1950. The `hatch_p_value_ensemble()` function is used to distinguish, for each grid point, between three possible cases:\n\n1. If more than 66% of the models are statistically significant (p-value < 0.05) and more than 80% of the models share the same sign, we consider the ensemble median trend to be statistically significant, and there is agreement on the sign. To represent this, no hatching is used.\n2. If less than 66% of the models are statistically significant, regardless of agreement on the sign of the trend, hatching is applied (indicating that the ensemble median trend is not statistically significant).\n3. If more than 66% of the models are statistically significant but less than 80% of the models share the same sign, we consider the ensemble median trend to be statistically significant, but there is no agreement on the sign of the trend. This is represented using crosses.\n\nDefine function to plot the caption of the figures (for the ensemble case)\nAdd caption to the figure\nAdd each line to the figure\nend captioning\nDefine function to plot the caption of the figures (for the individual models case)\nAdd caption to the figure\nAdd each line to the figure\n\n(section-3.2)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q04__1222f22ce0b6", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q04", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q04", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Projected changes in precipitation-based indices for impact models > Analysis and results > 3. Plot and describe results > 3.2. Plot ensemble maps", "title": "Projected changes in precipitation-based indices for impact models", "chunk_index": 14, "token_count": 279, "text_raw": "In this section, we invoke the `plot_ensemble()` function to visualise the trend calculated over the future period (2015-2099) for the model ensemble across Europe. Note that the model data used in this section has previously been interpolated to the \"regridding model\" grid (`\"clmcom_eth_cosmo_crclim\"` for this notebook).\n\nSpecifically, for each of the indices ('CWD', 'R20mm', 'RR1', 'RX1day' and 'RX5day'), this section presents a single layout including trend values of the future period (2015-2099) for: (a) the ensemble median (understood as the median of the trend values of the chosen subset of models calculated for each grid cell) and (b) the ensemble spread (derived as the standard deviation of the distribution of the chosen subset of models).\n\nFig number counter\nCommon title\n\n(section-3.3)=", "text_with_prefix": "EQC Quality Assessment: \"Projected changes in precipitation-based indices for impact models\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q04 | Category: Climate_Projections\nSection: Projected changes in precipitation-based indices for impact models > Analysis and results > 3. Plot and describe results > 3.2. Plot ensemble maps\n---\nIn this section, we invoke the `plot_ensemble()` function to visualise the trend calculated over the future period (2015-2099) for the model ensemble across Europe. Note that the model data used in this section has previously been interpolated to the \"regridding model\" grid (`\"clmcom_eth_cosmo_crclim\"` for this notebook).\n\nSpecifically, for each of the indices ('CWD', 'R20mm', 'RR1', 'RX1day' and 'RX5day'), this section presents a single layout including trend values of the future period (2015-2099) for: (a) the ensemble median (understood as the median of the trend values of the chosen subset of models calculated for each grid cell) and (b) the ensemble spread (derived as the standard deviation of the distribution of the chosen subset of models).\n\nFig number counter\nCommon title\n\n(section-3.3)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q04__4c184b877633", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q04", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q04", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Projected changes in precipitation-based indices for impact models > Analysis and results > 3. Plot and describe results > 3.3. Plot model maps", "title": "Projected changes in precipitation-based indices for impact models", "chunk_index": 15, "token_count": 198, "text_raw": "In this section, we invoke the `plot_models()` function to visualise the trend calculated over the future period (2015-2099) for every model individually across Europe. Note that the model data used in this section maintains its original grid.\n\nSpecifically, for each of the indices ('CWD', 'R20mm', 'RR1', 'RX1day' and 'RX5day'), this section presents a single layout including the trend for the future period (2015-2099) of every model.\n\n(section-3.4)=", "text_with_prefix": "EQC Quality Assessment: \"Projected changes in precipitation-based indices for impact models\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q04 | Category: Climate_Projections\nSection: Projected changes in precipitation-based indices for impact models > Analysis and results > 3. Plot and describe results > 3.3. Plot model maps\n---\nIn this section, we invoke the `plot_models()` function to visualise the trend calculated over the future period (2015-2099) for every model individually across Europe. Note that the model data used in this section maintains its original grid.\n\nSpecifically, for each of the indices ('CWD', 'R20mm', 'RR1', 'RX1day' and 'RX5day'), this section presents a single layout including the trend for the future period (2015-2099) of every model.\n\n(section-3.4)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q04__e8772c2b395b", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q04", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q04", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Projected changes in precipitation-based indices for impact models > Analysis and results > 3. Plot and describe results > 3.4. Boxplots of the future trend", "title": "Projected changes in precipitation-based indices for impact models", "chunk_index": 16, "token_count": 419, "text_raw": "Finally, we present boxplots representing the ensemble distribution of each climate model trend calculated over the future period (2015-2099) across Europe.\n\nDots represent the spatially-averaged future trend over the selected region (change of the number of days per decade) for each model (grey) and the ensemble mean (blue). The ensemble median is shown as a green line. Note that the spatially averaged values are calculated for each model from its original grid (i.e., no interpolated data has been used here).\n\nThe boxplot visually illustrates the distribution of trends (or bias trends) among the climate models, with the box covering the first quartile (Q1 = 25th percentile) to the third quartile (Q3 = 75th percentile), and a green line indicating the ensemble median (Q2 = 50th percentile). Whiskers extend from the edges of the box to show the full data range.\n\nEnsemble mean\n\n
\n
\n

Fig 11. Boxplots illustrating the future trends (2015-2099) of the ensemble distribution for the precipitation-based indices: (a) 'CWD', (b) R20mm, (c) RR1, (d) RX1day and (e) RX5day. The distribution is created by considering spatially averaged trends across Europe. The ensemble mean and the ensemble median trends are both included. Outliers in the distribution are denoted by a grey circle with a black contour.

\n
\n\n(section-3.5)=", "text_with_prefix": "EQC Quality Assessment: \"Projected changes in precipitation-based indices for impact models\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q04 | Category: Climate_Projections\nSection: Projected changes in precipitation-based indices for impact models > Analysis and results > 3. Plot and describe results > 3.4. Boxplots of the future trend\n---\nFinally, we present boxplots representing the ensemble distribution of each climate model trend calculated over the future period (2015-2099) across Europe.\n\nDots represent the spatially-averaged future trend over the selected region (change of the number of days per decade) for each model (grey) and the ensemble mean (blue). The ensemble median is shown as a green line. Note that the spatially averaged values are calculated for each model from its original grid (i.e., no interpolated data has been used here).\n\nThe boxplot visually illustrates the distribution of trends (or bias trends) among the climate models, with the box covering the first quartile (Q1 = 25th percentile) to the third quartile (Q3 = 75th percentile), and a green line indicating the ensemble median (Q2 = 50th percentile). Whiskers extend from the edges of the box to show the full data range.\n\nEnsemble mean\n\n
\n
\n

Fig 11. Boxplots illustrating the future trends (2015-2099) of the ensemble distribution for the precipitation-based indices: (a) 'CWD', (b) R20mm, (c) RR1, (d) RX1day and (e) RX5day. The distribution is created by considering spatially averaged trends across Europe. The ensemble mean and the ensemble median trends are both included. Outliers in the distribution are denoted by a grey circle with a black contour.

\n
\n\n(section-3.5)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q04__5efe77074174", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q04", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q04", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Projected changes in precipitation-based indices for impact models > Analysis and results > 3. Plot and describe results > 3.5. Results summary and discussion", "title": "Projected changes in precipitation-based indices for impact models", "chunk_index": 17, "token_count": 942, "text_raw": "- The level of agreement among models for precipitation-based indices projections is lower than for maximum temperature indices. Additionally, the trends are clearly less robust compared to maximum temperature. This is evident when comparing the findings of this assessment with those presented in \"CORDEX Climate Projections: Evaluating Uncertainty in Projected Changes in Extreme Temperature Indices for the Reinsurance Sector.\"\n\n- The future trend (calculated over the 2015-2099 period) in precipitation-based indices for the temporal aggregation of DJF largely depends on the considered index.\n\n- The maximum Consecutive Wet Days (CWD) are projected to increase on the Atlantic coasts of the western part of the continent (except in the central and northwest of Scandinavia and the west of Iceland, where it is expected to decrease). Decreases are also expected on the southernmost and eastern coasts of the Mediterranean basin. No significant trend is detected in the rest of the regions.\n\n- The number of heavy precipitation days (R20mm) is expected to increase in the west of the Iberian Peninsula, southwest of France, west of the United Kingdom, south of Norway, Balkan coasts, and some parts of the northern part of Italy and southeast of France. Decreases are expected in the Mediterranean coastal areas of Turkey. No significant trend is present in the model output for the rest of the regions.\n\n- Increases in the number of wet days (RR1) are projected to increase in the Scandinavian Peninsula (in contrast to the decrease expected in the western coastal regions), United Kingdom, center, center-east, and northeast of Europe. Conversely, decreases are expected in the Mediterranean regions of Africa and the eastern part of the Mediterranean basin.\n\n- The future trend of the maximum 1-day total precipitation (RX1day) and maximum 5-day total precipitation (RX5day) have a similar spatial pattern. An increase is projected for most of Europe (indicating more extreme precipitation events), with the exception of the Mediterranean regions of Africa and the southeastern part of the Mediterranean basin, and the center-west and northwestern Atlantic coastal regions of Scandinavia where a decrease is expected.\n\n- The boxplots displaying spatially-averaged values of the future trend for the temporal aggregation of winter (DJF) reveal diverse outcomes depending on the index under consideration. Notably, significant inter-model variability is evident for certain indices, whose interquartile range encompasses trends of varying directions. These are the cases of CWD and RR1. The other indices show a positive trend with outliers in some cases (RX5day).\n\n- It is important to emphasise that the boxplots present spatially-averaged values, and their interpretation should be approached with caution, as the outcomes can vary significantly when considering different regions across Europe. For regional analyses, it is essential to focus solely on the region of interest, as the influence of other European regions may distort the results, leading to conclusions that do not accurately reflect the specific area under study.\n\n- What do the results mean for users? Are the biases relevant?\n - The results of this notebook are strongly influenced by the index and region. Projected changes for RR1, RX1day, and RX5day display similar spatial patterns across the subset of models considered. In contrast, CWD and R20mm show more scattered spatial patterns, with regions that have significant trend signals exhibiting high uncertainty (inter-model spread).\n\n- Findings from another assessment of the same indices and seasonal aggregation (DJF) for the historical period 1971-2000, which evaluated biases (\"CORDEX Climate Projections: evaluating bias in precipitation-based indices for impact models\"), should also be taken into account. The high uncertainties in both assessments suggest that users should bias-correct precipitation-based indices before using them for specific applications [[5]](https://doi.org/10.1016/j.jhydrol.2012.05.052).\n\n\n
\nRESULTS NOTE:
\nIt is important to note that the results presented are specific to the 9 models chosen, and users should aim to assess as wide a range of models as possible before making a sub-selection.", "text_with_prefix": "EQC Quality Assessment: \"Projected changes in precipitation-based indices for impact models\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q04 | Category: Climate_Projections\nSection: Projected changes in precipitation-based indices for impact models > Analysis and results > 3. Plot and describe results > 3.5. Results summary and discussion\n---\n- The level of agreement among models for precipitation-based indices projections is lower than for maximum temperature indices. Additionally, the trends are clearly less robust compared to maximum temperature. This is evident when comparing the findings of this assessment with those presented in \"CORDEX Climate Projections: Evaluating Uncertainty in Projected Changes in Extreme Temperature Indices for the Reinsurance Sector.\"\n\n- The future trend (calculated over the 2015-2099 period) in precipitation-based indices for the temporal aggregation of DJF largely depends on the considered index.\n\n- The maximum Consecutive Wet Days (CWD) are projected to increase on the Atlantic coasts of the western part of the continent (except in the central and northwest of Scandinavia and the west of Iceland, where it is expected to decrease). Decreases are also expected on the southernmost and eastern coasts of the Mediterranean basin. No significant trend is detected in the rest of the regions.\n\n- The number of heavy precipitation days (R20mm) is expected to increase in the west of the Iberian Peninsula, southwest of France, west of the United Kingdom, south of Norway, Balkan coasts, and some parts of the northern part of Italy and southeast of France. Decreases are expected in the Mediterranean coastal areas of Turkey. No significant trend is present in the model output for the rest of the regions.\n\n- Increases in the number of wet days (RR1) are projected to increase in the Scandinavian Peninsula (in contrast to the decrease expected in the western coastal regions), United Kingdom, center, center-east, and northeast of Europe. Conversely, decreases are expected in the Mediterranean regions of Africa and the eastern part of the Mediterranean basin.\n\n- The future trend of the maximum 1-day total precipitation (RX1day) and maximum 5-day total precipitation (RX5day) have a similar spatial pattern. An increase is projected for most of Europe (indicating more extreme precipitation events), with the exception of the Mediterranean regions of Africa and the southeastern part of the Mediterranean basin, and the center-west and northwestern Atlantic coastal regions of Scandinavia where a decrease is expected.\n\n- The boxplots displaying spatially-averaged values of the future trend for the temporal aggregation of winter (DJF) reveal diverse outcomes depending on the index under consideration. Notably, significant inter-model variability is evident for certain indices, whose interquartile range encompasses trends of varying directions. These are the cases of CWD and RR1. The other indices show a positive trend with outliers in some cases (RX5day).\n\n- It is important to emphasise that the boxplots present spatially-averaged values, and their interpretation should be approached with caution, as the outcomes can vary significantly when considering different regions across Europe. For regional analyses, it is essential to focus solely on the region of interest, as the influence of other European regions may distort the results, leading to conclusions that do not accurately reflect the specific area under study.\n\n- What do the results mean for users? Are the biases relevant?\n - The results of this notebook are strongly influenced by the index and region. Projected changes for RR1, RX1day, and RX5day display similar spatial patterns across the subset of models considered. In contrast, CWD and R20mm show more scattered spatial patterns, with regions that have significant trend signals exhibiting high uncertainty (inter-model spread).\n\n- Findings from another assessment of the same indices and seasonal aggregation (DJF) for the historical period 1971-2000, which evaluated biases (\"CORDEX Climate Projections: evaluating bias in precipitation-based indices for impact models\"), should also be taken into account. The high uncertainties in both assessments suggest that users should bias-correct precipitation-based indices before using them for specific applications [[5]](https://doi.org/10.1016/j.jhydrol.2012.05.052).\n\n\n
\nRESULTS NOTE:
\nIt is important to note that the results presented are specific to the 9 models chosen, and users should aim to assess as wide a range of models as possible before making a sub-selection."} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q04__90b81ab873ec", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q04", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q04", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Projected changes in precipitation-based indices for impact models > ℹ️ If you want to know more > Key resources", "title": "Projected changes in precipitation-based indices for impact models", "chunk_index": 18, "token_count": 234, "text_raw": "Some key resources and further reading were linked throughout this assessment.\n\nThe CDS catalogue entries for the data used were:\n* CORDEX regional climate model data on single levels (Daily mean - Mean precipitation flux): https://cds.climate.copernicus.eu/datasets/projections-cordex-domains-single-levels?tab=overview\n\nCode libraries used:\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)\n* [icclim](https://icclim.readthedocs.io/en/stable/) Python package", "text_with_prefix": "EQC Quality Assessment: \"Projected changes in precipitation-based indices for impact models\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q04 | Category: Climate_Projections\nSection: Projected changes in precipitation-based indices for impact models > ℹ️ If you want to know more > Key resources\n---\nSome key resources and further reading were linked throughout this assessment.\n\nThe CDS catalogue entries for the data used were:\n* CORDEX regional climate model data on single levels (Daily mean - Mean precipitation flux): https://cds.climate.copernicus.eu/datasets/projections-cordex-domains-single-levels?tab=overview\n\nCode libraries used:\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)\n* [icclim](https://icclim.readthedocs.io/en/stable/) Python package"} {"chunk_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q04__48b2d05cd6a6", "report_id": "climate_projections-cordex-domains-single-levels_climate-and-weather-extremes_q04", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q04", "aspect_base": "climate-and-weather-extremes", "category": "Climate_Projections", "match_confidence": "exact", "section": "Projected changes in precipitation-based indices for impact models > ℹ️ If you want to know more > References", "title": "Projected changes in precipitation-based indices for impact models", "chunk_index": 19, "token_count": 545, "text_raw": "[[1]](https://doi.org/10.1016/j.envsci.2015.08.012) Panagos, P., Borrelli, P., Poesen, J., Ballabio, C., Lugato, E., Meusburger, K., Montanarella, L., Alewell, C. (2015). The new assessment of soil loss by water erosion in Europe. Environ. Sci. Policy, 54, pp. 438-447. https://doi.org/10.1016/j.envsci.2015.08.012\n\n[[2]](https://doi.org/10.1016/j.landusepol.2012.11.007) Salvati, L., Carlucci, M. (2013). The impact of mediterranean land degradation on agricultural income: a short-term scenario. Land Use Policy, 32, pp. 302-308. https://doi.org/10.1016/j.landusepol.2012.11.007\n\n[[3]](https://doi.org/10.1002/wcc.8) Rummukainen, M. (2010). State-of-the-art with regional climate models. WIREs Clim Change, 1: 82-96. https://doi.org/10.1002/wcc.8\n\n[[4]](https://doi.org/10.1088/1748-9326/aacc77) Silje Lund Sørland et al. (2018). Bias patterns and climate change signals in GCM-RCM model chains. Environ. Res. Lett. 13 074017. https://doi.org/10.1088/1748-9326/aacc77\n\n[[5]](https://doi.org/10.1016/j.jhydrol.2012.05.052) Teutschbein, C., Seibert, J. (2012). Bias correction of regional climate model simulations for hydrological climate-change impact studies: Review and evaluation of different methods. Journal of Hydrology,\nVolumes 456–457, pp. 12-29. https://doi.org/10.1016/j.jhydrol.2012.05.052", "text_with_prefix": "EQC Quality Assessment: \"Projected changes in precipitation-based indices for impact models\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: climate-and-weather-extremes_q04 | Category: Climate_Projections\nSection: Projected changes in precipitation-based indices for impact models > ℹ️ If you want to know more > References\n---\n[[1]](https://doi.org/10.1016/j.envsci.2015.08.012) Panagos, P., Borrelli, P., Poesen, J., Ballabio, C., Lugato, E., Meusburger, K., Montanarella, L., Alewell, C. (2015). The new assessment of soil loss by water erosion in Europe. Environ. Sci. Policy, 54, pp. 438-447. https://doi.org/10.1016/j.envsci.2015.08.012\n\n[[2]](https://doi.org/10.1016/j.landusepol.2012.11.007) Salvati, L., Carlucci, M. (2013). The impact of mediterranean land degradation on agricultural income: a short-term scenario. Land Use Policy, 32, pp. 302-308. https://doi.org/10.1016/j.landusepol.2012.11.007\n\n[[3]](https://doi.org/10.1002/wcc.8) Rummukainen, M. (2010). State-of-the-art with regional climate models. WIREs Clim Change, 1: 82-96. https://doi.org/10.1002/wcc.8\n\n[[4]](https://doi.org/10.1088/1748-9326/aacc77) Silje Lund Sørland et al. (2018). Bias patterns and climate change signals in GCM-RCM model chains. Environ. Res. Lett. 13 074017. https://doi.org/10.1088/1748-9326/aacc77\n\n[[5]](https://doi.org/10.1016/j.jhydrol.2012.05.052) Teutschbein, C., Seibert, J. (2012). Bias correction of regional climate model simulations for hydrological climate-change impact studies: Review and evaluation of different methods. Journal of Hydrology,\nVolumes 456–457, pp. 12-29. https://doi.org/10.1016/j.jhydrol.2012.05.052"} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q05__63abf8af4366", "report_id": "climate_projections-cordex-domains-single-levels_validation_q05", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q05", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CORDEX temperature biases over the Alpine region for the ski and hydrological sectors.", "title": "CORDEX temperature biases over the Alpine region for the ski and hydrological sectors.", "chunk_index": 0, "token_count": 113, "text_raw": "Production date: 31-03-2025\n\nProduced by: CMCC foundation - Euro-Mediterranean Center on Climate Change. Daniele Peano, Lorenzo Sangelantoni and Albert Martinez.", "text_with_prefix": "EQC Quality Assessment: \"CORDEX temperature biases over the Alpine region for the ski and hydrological sectors.\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q05 | Category: Climate_Projections\nSection: CORDEX temperature biases over the Alpine region for the ski and hydrological sectors.\n---\nProduction date: 31-03-2025\n\nProduced by: CMCC foundation - Euro-Mediterranean Center on Climate Change. Daniele Peano, Lorenzo Sangelantoni and Albert Martinez."} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q05__32031711c22b", "report_id": "climate_projections-cordex-domains-single-levels_validation_q05", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q05", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CORDEX temperature biases over the Alpine region for the ski and hydrological sectors. > Quality assessment question", "title": "CORDEX temperature biases over the Alpine region for the ski and hydrological sectors.", "chunk_index": 1, "token_count": 871, "text_raw": "* **Are atmospheric models too cold in the mountain regions?**\n\nDespite mountain areas covering a limited amount of land, they play a paramount role in the world’s water tower by supplying water for both natural and anthropogenic demands (Immerzeel et al., 2020 [[1]](https://doi.org/10.1038/s41586-019-1822-y)). In many mountain regions, various water-related processes (such as the fraction of rain versus snow, snowmelt, and streamflow drought, e.g., Rudisill et al., 2024 [[2]](https://doi.org/10.1175/BAMS-D-23-0082.1)) are linked to temperature conditions, making the mountain areas an indicator of the ongoing climate change.\n\nMountain terrains complicate the proper representation of temperature by models in those areas due to complex topography, slopes, aspect, terrain shadowing, snow-albedo feedbacks, and vegetation variability (*Figure .1*, reproduced from Rudisill et al., 2024 [[2]](https://doi.org/10.1175/BAMS-D-23-0082.1)).\n\nThanks to their higher resolution, Regional Climate Models (RCMs) are expected to represent the dynamical and thermodynamic processes underlying temperature variability over complex orography more accurately than Global Climate Models (GCMs). However, RCMs still present biases over mountain regions. Among these, winter cold biases—especially across peaks and ridges—are especially pronounced (Kotlarski et al., 2014 [[3]](https://doi.org/10.5194/gmd-7-1297-2014), Winter et al., 2017 [[4]](https://doi.org/10.1007/s00382-016-3130-7), Matiu et al., 2019 [[5]](https://doi.org/10.3390/atmos11010046), Vautard et al., 2020 [[6]](https://doi.org/10.1029/2019JD032344), Sy et al., 2024 [[7]](https://doi.org/10.1002/joc.8331)). Warm biases in summer have also been reported mainly in valley areas, but only in studies using convection-permitting RCMs (Rudisill et al., 2024 [[2]](https://doi.org/10.1175/BAMS-D-23-0082.1), Napoli et al., 2023 [[8]](https://doi.org/10.1002/joc.8222), Liu et al., 2017 [[9]](https://doi.org/10.1007/s00382-016-3327-9)); such fine-scale features are not captured by RCMs with coarser resolution. These temperature biases, consequently, impact their ability to properly represent snow-related processes in mountain areas (Matiu et al., 2019 [[5]](https://doi.org/10.3390/atmos11010046)). In fact, biases in temperature and snowpack, have a crucial knock-on effect on surface runoff, undermining the evaluation of the water-supplying capability of mountain ranges in Europe (Immerzeel et al., 2020 [[1]](https://doi.org/10.1038/s41586-019-1822-y)). Those cascading effects are crucial for lowlands water availability (van Tiel et al., 2023 [[10]](https://doi.org/10.1029/2022EF003408)) and inherent economic sectors such as tourism, hydropower, and agriculture (Beniston et al., 2011 [[11]](https://doi.org/10.1016/j.envsci.2010.12.009)).", "text_with_prefix": "EQC Quality Assessment: \"CORDEX temperature biases over the Alpine region for the ski and hydrological sectors.\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q05 | Category: Climate_Projections\nSection: CORDEX temperature biases over the Alpine region for the ski and hydrological sectors. > Quality assessment question\n---\n* **Are atmospheric models too cold in the mountain regions?**\n\nDespite mountain areas covering a limited amount of land, they play a paramount role in the world’s water tower by supplying water for both natural and anthropogenic demands (Immerzeel et al., 2020 [[1]](https://doi.org/10.1038/s41586-019-1822-y)). In many mountain regions, various water-related processes (such as the fraction of rain versus snow, snowmelt, and streamflow drought, e.g., Rudisill et al., 2024 [[2]](https://doi.org/10.1175/BAMS-D-23-0082.1)) are linked to temperature conditions, making the mountain areas an indicator of the ongoing climate change.\n\nMountain terrains complicate the proper representation of temperature by models in those areas due to complex topography, slopes, aspect, terrain shadowing, snow-albedo feedbacks, and vegetation variability (*Figure .1*, reproduced from Rudisill et al., 2024 [[2]](https://doi.org/10.1175/BAMS-D-23-0082.1)).\n\nThanks to their higher resolution, Regional Climate Models (RCMs) are expected to represent the dynamical and thermodynamic processes underlying temperature variability over complex orography more accurately than Global Climate Models (GCMs). However, RCMs still present biases over mountain regions. Among these, winter cold biases—especially across peaks and ridges—are especially pronounced (Kotlarski et al., 2014 [[3]](https://doi.org/10.5194/gmd-7-1297-2014), Winter et al., 2017 [[4]](https://doi.org/10.1007/s00382-016-3130-7), Matiu et al., 2019 [[5]](https://doi.org/10.3390/atmos11010046), Vautard et al., 2020 [[6]](https://doi.org/10.1029/2019JD032344), Sy et al., 2024 [[7]](https://doi.org/10.1002/joc.8331)). Warm biases in summer have also been reported mainly in valley areas, but only in studies using convection-permitting RCMs (Rudisill et al., 2024 [[2]](https://doi.org/10.1175/BAMS-D-23-0082.1), Napoli et al., 2023 [[8]](https://doi.org/10.1002/joc.8222), Liu et al., 2017 [[9]](https://doi.org/10.1007/s00382-016-3327-9)); such fine-scale features are not captured by RCMs with coarser resolution. These temperature biases, consequently, impact their ability to properly represent snow-related processes in mountain areas (Matiu et al., 2019 [[5]](https://doi.org/10.3390/atmos11010046)). In fact, biases in temperature and snowpack, have a crucial knock-on effect on surface runoff, undermining the evaluation of the water-supplying capability of mountain ranges in Europe (Immerzeel et al., 2020 [[1]](https://doi.org/10.1038/s41586-019-1822-y)). Those cascading effects are crucial for lowlands water availability (van Tiel et al., 2023 [[10]](https://doi.org/10.1029/2022EF003408)) and inherent economic sectors such as tourism, hydropower, and agriculture (Beniston et al., 2011 [[11]](https://doi.org/10.1016/j.envsci.2010.12.009))."} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q05__192007c5f0ea", "report_id": "climate_projections-cordex-domains-single-levels_validation_q05", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q05", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CORDEX temperature biases over the Alpine region for the ski and hydrological sectors. > Quality assessment statement", "title": "CORDEX temperature biases over the Alpine region for the ski and hydrological sectors.", "chunk_index": 2, "token_count": 511, "text_raw": "These are the key outcomes of this assessment\n\n* State-of-the-art regional climate models exhibit cold biases in the winter season compared to both station/point-based and gridded observations.\n\n* EURO-CORDEX RCMs exhibit systematic winter cold biases over the Alpine region at elevations above 1000 m, with even the \"hottest\" models underestimating temperatures in mountain areas.\n\n* Winter cold biases introduce additional uncertainties in snow estimates, reducing confidence in the use of RCM temperature projections for applications in the ski industry.\n\n* Temperature and mountain snowpack biases affect estimates of water availability, influencing the ability to assess changes in various hydrological and economic sectors, such as tourism, hydropower, and agriculture.\n\n* The outcomes of this assessment, focused on the Alpine region, advocate caution when using RCM future temperature projections. A key concern is the potential propagation of present-climate biases into a warmer future climate, representing an aspect (the propagation) still largely unexplored.\n\n```\n\nattachment:6360ba57-d679-4630-8214-a5a2ec7098a7.jpg\n---\nwidth: 900px\nalt: Abstract \n---\nDepiction of processes influencing mountain T2m (temperature at 2 metres above ground level). (a) Incident solar radiation at the ground surface depends on terrain aspect and shadowing. Mountain slopes can also reflect SW and emit LW radiation; (b) snow reflects solar radiation, cooling surface temperatures through the snow-albedo feedback; (c) vegetation’s high surface roughness and lower albedos warm surrounding air relative to non-vegetated snow surfaces; (d) moderate synoptic forcing leads to upper-air and boundary layer temperature mixing, with an un-inverted vs elevation profile; (e) katabatic flows develop valley temperature inversions during clear-sky, cloud-free conditions when surface LW cooling is strong. Image reproduced from Figure 1 in Rudisill et al. (2024) [[2]](https://doi.org/10.1175/BAMS-D-23-0082.1), © American Meteorological Society. Used with permission.\n\n```", "text_with_prefix": "EQC Quality Assessment: \"CORDEX temperature biases over the Alpine region for the ski and hydrological sectors.\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q05 | Category: Climate_Projections\nSection: CORDEX temperature biases over the Alpine region for the ski and hydrological sectors. > Quality assessment statement\n---\nThese are the key outcomes of this assessment\n\n* State-of-the-art regional climate models exhibit cold biases in the winter season compared to both station/point-based and gridded observations.\n\n* EURO-CORDEX RCMs exhibit systematic winter cold biases over the Alpine region at elevations above 1000 m, with even the \"hottest\" models underestimating temperatures in mountain areas.\n\n* Winter cold biases introduce additional uncertainties in snow estimates, reducing confidence in the use of RCM temperature projections for applications in the ski industry.\n\n* Temperature and mountain snowpack biases affect estimates of water availability, influencing the ability to assess changes in various hydrological and economic sectors, such as tourism, hydropower, and agriculture.\n\n* The outcomes of this assessment, focused on the Alpine region, advocate caution when using RCM future temperature projections. A key concern is the potential propagation of present-climate biases into a warmer future climate, representing an aspect (the propagation) still largely unexplored.\n\n```\n\nattachment:6360ba57-d679-4630-8214-a5a2ec7098a7.jpg\n---\nwidth: 900px\nalt: Abstract \n---\nDepiction of processes influencing mountain T2m (temperature at 2 metres above ground level). (a) Incident solar radiation at the ground surface depends on terrain aspect and shadowing. Mountain slopes can also reflect SW and emit LW radiation; (b) snow reflects solar radiation, cooling surface temperatures through the snow-albedo feedback; (c) vegetation’s high surface roughness and lower albedos warm surrounding air relative to non-vegetated snow surfaces; (d) moderate synoptic forcing leads to upper-air and boundary layer temperature mixing, with an un-inverted vs elevation profile; (e) katabatic flows develop valley temperature inversions during clear-sky, cloud-free conditions when surface LW cooling is strong. Image reproduced from Figure 1 in Rudisill et al. (2024) [[2]](https://doi.org/10.1175/BAMS-D-23-0082.1), © American Meteorological Society. Used with permission.\n\n```"} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q05__129d70588010", "report_id": "climate_projections-cordex-domains-single-levels_validation_q05", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q05", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CORDEX temperature biases over the Alpine region for the ski and hydrological sectors. > Methodology", "title": "CORDEX temperature biases over the Alpine region for the ski and hydrological sectors.", "chunk_index": 3, "token_count": 198, "text_raw": "A review of the relevant scientific literature documenting state-of-the-art RCMs cold biases over mountainous regions, with a focus on the Alpine region, is conducted. Methods and key findings from a series of studies on this topic are presented and discussed.\n\nThe analysis and results follow the next outline:\n\n**[](climate_projections-cordex-domains-single-levels_validation_q05:section-1)**\n\n**[](climate_projections-cordex-domains-single-levels_validation_q05:section-2)**\n\n**[](climate_projections-cordex-domains-single-levels_validation_q05:section-3)**", "text_with_prefix": "EQC Quality Assessment: \"CORDEX temperature biases over the Alpine region for the ski and hydrological sectors.\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q05 | Category: Climate_Projections\nSection: CORDEX temperature biases over the Alpine region for the ski and hydrological sectors. > Methodology\n---\nA review of the relevant scientific literature documenting state-of-the-art RCMs cold biases over mountainous regions, with a focus on the Alpine region, is conducted. Methods and key findings from a series of studies on this topic are presented and discussed.\n\nThe analysis and results follow the next outline:\n\n**[](climate_projections-cordex-domains-single-levels_validation_q05:section-1)**\n\n**[](climate_projections-cordex-domains-single-levels_validation_q05:section-2)**\n\n**[](climate_projections-cordex-domains-single-levels_validation_q05:section-3)**"} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q05__88f7eb72161b", "report_id": "climate_projections-cordex-domains-single-levels_validation_q05", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q05", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CORDEX temperature biases over the Alpine region for the ski and hydrological sectors. > Analysis and results > 1. Main methods within the literature", "title": "CORDEX temperature biases over the Alpine region for the ski and hydrological sectors.", "chunk_index": 4, "token_count": 1053, "text_raw": "The modelled temperatures are typically evaluated against gridded meteorological observations, reanalysis datasets, or point-based observations (i.e. data from individual weather stations). Usually, the assessments focus on the daily average, minimum (occurring at night), and maximum (occurring during the day) values (e.g., Rudisill et al., 2024 [[2]](https://doi.org/10.1175/BAMS-D-23-0082.1)).\n\nVautard et al. (2021) [[6]](https://doi.org/10.1029/2019JD032344) use the E-OBS17 0.22° dataset (Haylock et al., 2008 [[12]](https://doi.org/10.1029/2008JD010201)) as reference to conduct a comprehensive evaluation of a set of 55 EURO-CORDEX regional climate models. The E-OBS datasets are interpolated onto a common 0.11° resolution grid using a bilinear interpolation to perform their evaluation. Matiu et al. (2019) [[5]](https://doi.org/10.3390/atmos11010046) use a similar approach to evaluate temperature biases in EURO-CORDEX models, based on the E-OBSv20 dataset (Cornes et al., 2018 [[13]](https://doi.org/10.1029/2017JD028200)). Differently from Vautard et al. (2021) [[6]](https://doi.org/10.1029/2019JD032344), their study also focuses on snow biases over the European Alps, linking them to orography, temperature, and precipitation mismatches. The analysis combines gridded observations with station data from the province of Bolzano, Germany, and Switzerland (*Figure .2*), which are also used to validate temperature estimates.\n\nMatiu et al. (2024) [[14]](https://doi.org/10.1007/s00382-024-07376-y) analysed EURO-CORDEX temperature and precipitation data for the Alpine region (*Figure .3*) across different elevations, comparing them with gridded observations from E-OBS and the Alpine Precipitation Grid Dataset (APGD; Isotta et al., 2014 [[15]](https://doi.org/10.1002/joc.3794)). They also conducted an observational intercomparison using MESAN (Landelius et al., 2016 [[16]](https://doi.org/10.1002/qj.2813)), WFDEI (Weedon et al., 2014 [[17]](https://doi.org/10.1002/2014WR015638)), and high-resolution national datasets.\n\nattachment:add9873d-30b4-46ad-a83c-ba72f293890f.png\n---\nwidth: 900px\nalt: figure 2\n---\nMap of European Alps. Terrain map with boundaries of the Alpine Convention (Alpconv), marking the core Alpine region. Points indicate the station locations separated by their characteristic and application based on analysis performed in Matiu et al. (2019) (+: show stations used for evaluating reanalysis-driven RCMs, x: show stations used for evaluating GCM-driven RCMs; stars (+ and x overlaid): show stations used for evaluating both). (Map tiles by Stamen Design, under CC BY 3.0. Data by Open Street Map, under ODbL). Image reproduced from Figure 1 in Matiu et al. (2019) [[5]](https://doi.org/10.3390/atmos11010046), under [CC BY 4.0](https://creativecommons.org/licenses/by/4.0/).\n\n```\n\nattachment:209c2943-7ef2-4da6-8246-c8473fd6239c.png\n---\nwidth: 900px\nalt: figure 3\n---\nElevation map of the study area. The source is the European Digital Elevation Model (EU-DEM), version 1.1, which has been aggregated to 1 km. The black rectangle is the area for which the regional climate model data was cropped. Image reproduced from *Figure 1* in Matiu et al. (2024) [[14]](https://doi.org/10.1007/s00382-024-07376-y), under [CC BY 4.0](https://creativecommons.org/licenses/by/4.0/).\n\n```\n\n(climate_projections-cordex-domains-single-levels_validation_q05:section-2)=", "text_with_prefix": "EQC Quality Assessment: \"CORDEX temperature biases over the Alpine region for the ski and hydrological sectors.\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q05 | Category: Climate_Projections\nSection: CORDEX temperature biases over the Alpine region for the ski and hydrological sectors. > Analysis and results > 1. Main methods within the literature\n---\nThe modelled temperatures are typically evaluated against gridded meteorological observations, reanalysis datasets, or point-based observations (i.e. data from individual weather stations). Usually, the assessments focus on the daily average, minimum (occurring at night), and maximum (occurring during the day) values (e.g., Rudisill et al., 2024 [[2]](https://doi.org/10.1175/BAMS-D-23-0082.1)).\n\nVautard et al. (2021) [[6]](https://doi.org/10.1029/2019JD032344) use the E-OBS17 0.22° dataset (Haylock et al., 2008 [[12]](https://doi.org/10.1029/2008JD010201)) as reference to conduct a comprehensive evaluation of a set of 55 EURO-CORDEX regional climate models. The E-OBS datasets are interpolated onto a common 0.11° resolution grid using a bilinear interpolation to perform their evaluation. Matiu et al. (2019) [[5]](https://doi.org/10.3390/atmos11010046) use a similar approach to evaluate temperature biases in EURO-CORDEX models, based on the E-OBSv20 dataset (Cornes et al., 2018 [[13]](https://doi.org/10.1029/2017JD028200)). Differently from Vautard et al. (2021) [[6]](https://doi.org/10.1029/2019JD032344), their study also focuses on snow biases over the European Alps, linking them to orography, temperature, and precipitation mismatches. The analysis combines gridded observations with station data from the province of Bolzano, Germany, and Switzerland (*Figure .2*), which are also used to validate temperature estimates.\n\nMatiu et al. (2024) [[14]](https://doi.org/10.1007/s00382-024-07376-y) analysed EURO-CORDEX temperature and precipitation data for the Alpine region (*Figure .3*) across different elevations, comparing them with gridded observations from E-OBS and the Alpine Precipitation Grid Dataset (APGD; Isotta et al., 2014 [[15]](https://doi.org/10.1002/joc.3794)). They also conducted an observational intercomparison using MESAN (Landelius et al., 2016 [[16]](https://doi.org/10.1002/qj.2813)), WFDEI (Weedon et al., 2014 [[17]](https://doi.org/10.1002/2014WR015638)), and high-resolution national datasets.\n\nattachment:add9873d-30b4-46ad-a83c-ba72f293890f.png\n---\nwidth: 900px\nalt: figure 2\n---\nMap of European Alps. Terrain map with boundaries of the Alpine Convention (Alpconv), marking the core Alpine region. Points indicate the station locations separated by their characteristic and application based on analysis performed in Matiu et al. (2019) (+: show stations used for evaluating reanalysis-driven RCMs, x: show stations used for evaluating GCM-driven RCMs; stars (+ and x overlaid): show stations used for evaluating both). (Map tiles by Stamen Design, under CC BY 3.0. Data by Open Street Map, under ODbL). Image reproduced from Figure 1 in Matiu et al. (2019) [[5]](https://doi.org/10.3390/atmos11010046), under [CC BY 4.0](https://creativecommons.org/licenses/by/4.0/).\n\n```\n\nattachment:209c2943-7ef2-4da6-8246-c8473fd6239c.png\n---\nwidth: 900px\nalt: figure 3\n---\nElevation map of the study area. The source is the European Digital Elevation Model (EU-DEM), version 1.1, which has been aggregated to 1 km. The black rectangle is the area for which the regional climate model data was cropped. Image reproduced from *Figure 1* in Matiu et al. (2024) [[14]](https://doi.org/10.1007/s00382-024-07376-y), under [CC BY 4.0](https://creativecommons.org/licenses/by/4.0/).\n\n```\n\n(climate_projections-cordex-domains-single-levels_validation_q05:section-2)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q05__82fa07da7aa0", "report_id": "climate_projections-cordex-domains-single-levels_validation_q05", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q05", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CORDEX temperature biases over the Alpine region for the ski and hydrological sectors. > Analysis and results > 2. RCMs temperature biases over Alps", "title": "CORDEX temperature biases over the Alpine region for the ski and hydrological sectors.", "chunk_index": 5, "token_count": 967, "text_raw": "Vautard et al. (2021) [[6]](https://doi.org/10.1029/2019JD032344) assessed the performance of 55 combinations of Regional Climate Models (RCMs) driven by Global Climate Models (GCMs), representing the full CORDEX ensemble available for the European domain as of 2019, which, to our knowledge, remains unchanged. *Figure .4* (adapted from Vautard et al., 2021 [[6]](https://doi.org/10.1029/2019JD032344)) presents the distribution of mean temperature biases for the winter and summer seasons. Specifically, it shows the median, 5th, and 95th percentiles of the mean temperature bias distribution derived from the 55 RCM simulations for each season. Unlike the summer season, during winter, even the 95th percentile of the mean temperature bias among the 55 RCM simulations shows cold biases over high mountain regions, such as Scandinavia and the Alps. This indicates that even the warmest models within the Euro-CORDEX ensemble still exhibit negative biases in mountain regions. Similar biases were also identified by other studies. For example, Kotlarski et al. (2014) [[3]](https://doi.org/10.5194/gmd-7-1297-2014) evaluated EURO-CORDEX regional climate models driven by ERA-Interim over the Alps for the period 1989–2008, identifying similar biases. Blázquez et al. (2023) [[18]](https://doi.org/10.1007/s00382-023-06727-5) analysed temperature and precipitation biases in CORDEX RCM simulations over the Andes, while Sy et al. (2024) [[7]](https://doi.org/10.1002/joc.8331) assessed NA-CORDEX models, reanalysis products, and observational datasets against the U.S. Climate Reference Network in North America. These studies all report cold biases over mountainous regions, emphasizing common RCMs’ issues in representing temperature in complex orography regions.\n\nattachment:238d2b8d-0ca9-4d2b-b9a4-8e0277793201.png\n---\nwidth: 900px\nalt: figure 4\n---\nDistribution of temperature biases (°C) for the winter season (top row) and the summer season (bottom row). Top row, left: median of the mean temperature bias among the 55 RCM simulations, middle: 5th percentile of the RCM temperature bias distribution, and right: 95th percentile of the RCM temperature bias distribution; bottom row: same as top row but for the summer season. Image adapted from *Figure 1* in Vautard et al. (2021) [[6]](https://doi.org/10.1029/2019JD032344), under [CC BY 4.0](https://creativecommons.org/licenses/by/4.0/).\n```\n\nVarious potential sources have been suggested to explain the large cold temperature biases across major mountain regions. Rudisill et al. (2024) [[2]](https://doi.org/10.1175/BAMS-D-23-0082.1) present a comprehensive list of possible mechanisms, including the representation of snow thermodynamics, elevation and orographic discrepancies due to grid spacing, and biases in radiation and wind.\n\nOther studies have explored specific factors contributing to these biases. For example, García-Díez et al. (2015) [[19]](https://doi.org/10.1007/s00382-015-2529-x) remark a link between winter cold biases and snow-covered regions, highlighting a poor representation of snow-atmosphere interaction amplified by albedo feedback as a possible driver of the winter cold biases in RCMs. Moreover, the underestimation of cloud cover and consequent changes in incoming radiative fluxes (shortwave and longwave) may also influence temperature biases in Alpine areas (Vionnet et al., 2016 [[20]](https://doi.org/10.1175/JHM-D-15-0241.1)).", "text_with_prefix": "EQC Quality Assessment: \"CORDEX temperature biases over the Alpine region for the ski and hydrological sectors.\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q05 | Category: Climate_Projections\nSection: CORDEX temperature biases over the Alpine region for the ski and hydrological sectors. > Analysis and results > 2. RCMs temperature biases over Alps\n---\nVautard et al. (2021) [[6]](https://doi.org/10.1029/2019JD032344) assessed the performance of 55 combinations of Regional Climate Models (RCMs) driven by Global Climate Models (GCMs), representing the full CORDEX ensemble available for the European domain as of 2019, which, to our knowledge, remains unchanged. *Figure .4* (adapted from Vautard et al., 2021 [[6]](https://doi.org/10.1029/2019JD032344)) presents the distribution of mean temperature biases for the winter and summer seasons. Specifically, it shows the median, 5th, and 95th percentiles of the mean temperature bias distribution derived from the 55 RCM simulations for each season. Unlike the summer season, during winter, even the 95th percentile of the mean temperature bias among the 55 RCM simulations shows cold biases over high mountain regions, such as Scandinavia and the Alps. This indicates that even the warmest models within the Euro-CORDEX ensemble still exhibit negative biases in mountain regions. Similar biases were also identified by other studies. For example, Kotlarski et al. (2014) [[3]](https://doi.org/10.5194/gmd-7-1297-2014) evaluated EURO-CORDEX regional climate models driven by ERA-Interim over the Alps for the period 1989–2008, identifying similar biases. Blázquez et al. (2023) [[18]](https://doi.org/10.1007/s00382-023-06727-5) analysed temperature and precipitation biases in CORDEX RCM simulations over the Andes, while Sy et al. (2024) [[7]](https://doi.org/10.1002/joc.8331) assessed NA-CORDEX models, reanalysis products, and observational datasets against the U.S. Climate Reference Network in North America. These studies all report cold biases over mountainous regions, emphasizing common RCMs’ issues in representing temperature in complex orography regions.\n\nattachment:238d2b8d-0ca9-4d2b-b9a4-8e0277793201.png\n---\nwidth: 900px\nalt: figure 4\n---\nDistribution of temperature biases (°C) for the winter season (top row) and the summer season (bottom row). Top row, left: median of the mean temperature bias among the 55 RCM simulations, middle: 5th percentile of the RCM temperature bias distribution, and right: 95th percentile of the RCM temperature bias distribution; bottom row: same as top row but for the summer season. Image adapted from *Figure 1* in Vautard et al. (2021) [[6]](https://doi.org/10.1029/2019JD032344), under [CC BY 4.0](https://creativecommons.org/licenses/by/4.0/).\n```\n\nVarious potential sources have been suggested to explain the large cold temperature biases across major mountain regions. Rudisill et al. (2024) [[2]](https://doi.org/10.1175/BAMS-D-23-0082.1) present a comprehensive list of possible mechanisms, including the representation of snow thermodynamics, elevation and orographic discrepancies due to grid spacing, and biases in radiation and wind.\n\nOther studies have explored specific factors contributing to these biases. For example, García-Díez et al. (2015) [[19]](https://doi.org/10.1007/s00382-015-2529-x) remark a link between winter cold biases and snow-covered regions, highlighting a poor representation of snow-atmosphere interaction amplified by albedo feedback as a possible driver of the winter cold biases in RCMs. Moreover, the underestimation of cloud cover and consequent changes in incoming radiative fluxes (shortwave and longwave) may also influence temperature biases in Alpine areas (Vionnet et al., 2016 [[20]](https://doi.org/10.1175/JHM-D-15-0241.1))."} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q05__b65839419963", "report_id": "climate_projections-cordex-domains-single-levels_validation_q05", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q05", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CORDEX temperature biases over the Alpine region for the ski and hydrological sectors. > Analysis and results > 2. RCMs temperature biases over Alps", "title": "CORDEX temperature biases over the Alpine region for the ski and hydrological sectors.", "chunk_index": 6, "token_count": 936, "text_raw": "wave) may also influence temperature biases in Alpine areas (Vionnet et al., 2016 [[20]](https://doi.org/10.1175/JHM-D-15-0241.1)).\n\nRelevant studies have further explored elevation-dependent biases in regional climate models across mountain regions. Notably, Matiu et al. (2024) [[14]](https://doi.org/10.1007/s00382-024-07376-y) evaluated data from the EURO-CORDEX ensemble for the Alpine region and the domain specified in *Figure .3*, comparing it with multiple gridded observational datasets, including E-OBS, APGD, and high-resolution national datasets. Their study indicates that biases in seasonal temperature increase with elevation, while also highlighting that observational datasets themselves are not immune to biases, making the choice of dataset for evaluation critical (also see Shrestha et al., 2023 [[21]](https://doi.org/10.3390/cli11070154)). *Figure .5c* (reproduced from Matiu et al., 2024 [[14]](https://doi.org/10.1007/s00382-024-07376-y)) presents the differences in temperature indices between EURO-CORDEX GCM-RCM combinations and E-OBS for various elevations, comparing summer and winter, with values based on the 1971–2000 period. The cold bias became more pronounced with lower temperatures, particularly in winter and with increasing elevation. Smaller biases were observed below 1000 m, while above this threshold and during winter, all considered RCMs exhibited a strong cold bias. This denotes mountain peaks and ridges as the areas characterised by the most substantial cold biases, as highlighted by Rudisill et al. (2024) [[2]](https://doi.org/10.1175/BAMS-D-23-0082.1).\n\nattachment:f1bf2739-4225-4015-bace-595255b2ae1a.png\n---\nwidth: 900px\nalt: figure 5\n---\n(a) Precipitation indices in APGD and GCM-RCMs. (b) Differences between each GCM-RCM and APGD. (c) Differences in temperature indices between each GCM-RCM combination and E-OBS. Each line is one model combination. Values refer to the 1971–2000 climatology. Image reproduced from *Figure 5* in Matiu et al. (2024) [[14]](https://doi.org/10.1007/s00382-024-07376-y), under [CC BY 4.0](https://creativecommons.org/licenses/by/4.0/).\n\n```\n\nEven though local-scale forcing misrepresentation plays a paramount role in driving the above-mentioned cold biases, large-scale-forcing-related biases can also exacerbate or mitigate temperature biases in mountainous areas (e.g., Lhotka et al., 2018 [[22]](https://doi.org/10.1007/s00704-016-2031-3) and Maraun et al., 2021 [[23]](https://doi.org/10.1029/2020JD032824)), especially during the winter season. In RCMs, these biases may stem from inherited large-scale misrepresentations introduced by the driving global climate model, or from processes internal to the RCM formulation. Importantly, the presence of such biases does not imply that RCM data offer no added value over GCM outputs. On the contrary, RCMs — particularly in regions with complex orography — can provide more realistic representations of local climate processes, and their added value should not be overlooked, even when biases persist. An example of this can be found in Bilbao-Barrenetxea et al., 2024 [[24]](https://doi.org/10.1007/s00382-024-07318-8), which assessed the added value of using EURO-CORDEX over the complex orography of the Pyrenees compared to GCMs.", "text_with_prefix": "EQC Quality Assessment: \"CORDEX temperature biases over the Alpine region for the ski and hydrological sectors.\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q05 | Category: Climate_Projections\nSection: CORDEX temperature biases over the Alpine region for the ski and hydrological sectors. > Analysis and results > 2. RCMs temperature biases over Alps\n---\nwave) may also influence temperature biases in Alpine areas (Vionnet et al., 2016 [[20]](https://doi.org/10.1175/JHM-D-15-0241.1)).\n\nRelevant studies have further explored elevation-dependent biases in regional climate models across mountain regions. Notably, Matiu et al. (2024) [[14]](https://doi.org/10.1007/s00382-024-07376-y) evaluated data from the EURO-CORDEX ensemble for the Alpine region and the domain specified in *Figure .3*, comparing it with multiple gridded observational datasets, including E-OBS, APGD, and high-resolution national datasets. Their study indicates that biases in seasonal temperature increase with elevation, while also highlighting that observational datasets themselves are not immune to biases, making the choice of dataset for evaluation critical (also see Shrestha et al., 2023 [[21]](https://doi.org/10.3390/cli11070154)). *Figure .5c* (reproduced from Matiu et al., 2024 [[14]](https://doi.org/10.1007/s00382-024-07376-y)) presents the differences in temperature indices between EURO-CORDEX GCM-RCM combinations and E-OBS for various elevations, comparing summer and winter, with values based on the 1971–2000 period. The cold bias became more pronounced with lower temperatures, particularly in winter and with increasing elevation. Smaller biases were observed below 1000 m, while above this threshold and during winter, all considered RCMs exhibited a strong cold bias. This denotes mountain peaks and ridges as the areas characterised by the most substantial cold biases, as highlighted by Rudisill et al. (2024) [[2]](https://doi.org/10.1175/BAMS-D-23-0082.1).\n\nattachment:f1bf2739-4225-4015-bace-595255b2ae1a.png\n---\nwidth: 900px\nalt: figure 5\n---\n(a) Precipitation indices in APGD and GCM-RCMs. (b) Differences between each GCM-RCM and APGD. (c) Differences in temperature indices between each GCM-RCM combination and E-OBS. Each line is one model combination. Values refer to the 1971–2000 climatology. Image reproduced from *Figure 5* in Matiu et al. (2024) [[14]](https://doi.org/10.1007/s00382-024-07376-y), under [CC BY 4.0](https://creativecommons.org/licenses/by/4.0/).\n\n```\n\nEven though local-scale forcing misrepresentation plays a paramount role in driving the above-mentioned cold biases, large-scale-forcing-related biases can also exacerbate or mitigate temperature biases in mountainous areas (e.g., Lhotka et al., 2018 [[22]](https://doi.org/10.1007/s00704-016-2031-3) and Maraun et al., 2021 [[23]](https://doi.org/10.1029/2020JD032824)), especially during the winter season. In RCMs, these biases may stem from inherited large-scale misrepresentations introduced by the driving global climate model, or from processes internal to the RCM formulation. Importantly, the presence of such biases does not imply that RCM data offer no added value over GCM outputs. On the contrary, RCMs — particularly in regions with complex orography — can provide more realistic representations of local climate processes, and their added value should not be overlooked, even when biases persist. An example of this can be found in Bilbao-Barrenetxea et al., 2024 [[24]](https://doi.org/10.1007/s00382-024-07318-8), which assessed the added value of using EURO-CORDEX over the complex orography of the Pyrenees compared to GCMs."} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q05__d13cb0678ce7", "report_id": "climate_projections-cordex-domains-single-levels_validation_q05", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q05", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CORDEX temperature biases over the Alpine region for the ski and hydrological sectors. > Analysis and results > 2. RCMs temperature biases over Alps", "title": "CORDEX temperature biases over the Alpine region for the ski and hydrological sectors.", "chunk_index": 7, "token_count": 379, "text_raw": "1007/s00382-024-07318-8), which assessed the added value of using EURO-CORDEX over the complex orography of the Pyrenees compared to GCMs.\n\nIn recent years, several promising high-resolution modelling initiatives (including both regional and global climate models) are paving the path toward a better understanding and eventual reduction of this long-standing biases in climate models (Demory et al., 2020 [[25]](https://doi.org/10.5194/gmd-13-5485-2020), Moreno-Chamarro et al., 2022 [[26]](https://doi.org/10.5194/gmd-15-269-2022), Roberts et al., 2025 [[27]](https://doi.org/10.5194/gmd-18-1307-2025), Sangelantoni et al., 2024 [[28]](https://doi.org/10.1007/s00382-023-06769-9), Sangelantoni et al., 2025 [[29]](https://doi.org/10.1029/2024GL111147) and Soares et al., 2022 [[30]](https://doi.org/10.1007/s00382-022-06593-7)).\n\n(climate_projections-cordex-domains-single-levels_validation_q05:section-3)=", "text_with_prefix": "EQC Quality Assessment: \"CORDEX temperature biases over the Alpine region for the ski and hydrological sectors.\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q05 | Category: Climate_Projections\nSection: CORDEX temperature biases over the Alpine region for the ski and hydrological sectors. > Analysis and results > 2. RCMs temperature biases over Alps\n---\n1007/s00382-024-07318-8), which assessed the added value of using EURO-CORDEX over the complex orography of the Pyrenees compared to GCMs.\n\nIn recent years, several promising high-resolution modelling initiatives (including both regional and global climate models) are paving the path toward a better understanding and eventual reduction of this long-standing biases in climate models (Demory et al., 2020 [[25]](https://doi.org/10.5194/gmd-13-5485-2020), Moreno-Chamarro et al., 2022 [[26]](https://doi.org/10.5194/gmd-15-269-2022), Roberts et al., 2025 [[27]](https://doi.org/10.5194/gmd-18-1307-2025), Sangelantoni et al., 2024 [[28]](https://doi.org/10.1007/s00382-023-06769-9), Sangelantoni et al., 2025 [[29]](https://doi.org/10.1029/2024GL111147) and Soares et al., 2022 [[30]](https://doi.org/10.1007/s00382-022-06593-7)).\n\n(climate_projections-cordex-domains-single-levels_validation_q05:section-3)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q05__b27eff4e3a4e", "report_id": "climate_projections-cordex-domains-single-levels_validation_q05", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q05", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CORDEX temperature biases over the Alpine region for the ski and hydrological sectors. > Analysis and results > 3. Implications for the users", "title": "CORDEX temperature biases over the Alpine region for the ski and hydrological sectors.", "chunk_index": 8, "token_count": 755, "text_raw": "The state-of-the-art regional climate models represent the principal tool to be implemented in estimating past, present, and future conditions of mountain areas, such as the Alps.\n\nBased on the results of this assessment, users need to be cautious when using the temperature projections provided by these tools. Notably, cold biases over the Alps are systematic only in winter across all EURO-CORDEX models above 1000 m, with the magnitude of the bias increasing with elevation. Contrary to what might be expected, these winter biases in mountainous regions cannot be solely attributed to altitude underestimation by models (which would typically lead to warmer biases). Instead, as discussed throughout this notebook, these cold biases arise from complex interactions within the models’ numerical schemes, particularly in the treatment of snow and ice processes, which result in the underestimation of surface temperatures.\n\nBias-adjustment techniques can partially relieve the assessed cold temperature biases. In particular, it can reduce the present-day discrepancies between modelled and observed temperature fields (Matiu et al., 2024 [[14]](https://doi.org/10.1007/s00382-024-07376-y), *Figure .6b*). However, bias-adjustment methods do not guarantee that the adjusted projected results will better match future reality, since the bias may not be stationary in a changing climate (e.g., Schmith et al., 2021 [[31]](https://doi.org/10.5194/hess-25-273-2021)). Special consideration should also be given to the representation of physical relationships among variables subjected to bias adjustment (Gennaretti et al., 2015 [[32]](https://doi.org/10.1002/2015JD023890)), as well as to unphysical artefacts such as the damping or amplification of original simulation trends, particularly in regions with complex orography (Sangelantoni et al., 2019 [[33]](https://doi.org/10.1007/s00704-018-2406-8)). Furthermore, different bias-adjustment methods do not always improve downstream products, as no single technique works reliably across all locations, variables, and applications.\n\nThe assessed temperature and mountain snowpack biases influence the current regional climate models' ability to quantify water availability in mountain areas and, consequently, the capability to estimate changes in resources for local and downstream hydrology and economic sectors, such as tourism, hydropower, and agriculture.\n\nattachment:013623ec-4a51-4ab5-bf67-483d3b6e405a.png\n---\nwidth: 900px\nalt: figure 6\n---\nImpact of different bias adjustment methods on (a) precipitation indices and (b) temperature indices over 1971–2000. Each line is a GCM-RCM combination. Grey dashed lines are raw (uncorrected) models, while solid colored lines are bias-adjusted using different methods (see legend). Image reproduced from *Figure 8* in Matiu et al. (2024) [[14]](https://doi.org/10.1007/s00382-024-07376-y), under [CC BY 4.0](https://creativecommons.org/licenses/by/4.0/).\n\n```", "text_with_prefix": "EQC Quality Assessment: \"CORDEX temperature biases over the Alpine region for the ski and hydrological sectors.\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q05 | Category: Climate_Projections\nSection: CORDEX temperature biases over the Alpine region for the ski and hydrological sectors. > Analysis and results > 3. Implications for the users\n---\nThe state-of-the-art regional climate models represent the principal tool to be implemented in estimating past, present, and future conditions of mountain areas, such as the Alps.\n\nBased on the results of this assessment, users need to be cautious when using the temperature projections provided by these tools. Notably, cold biases over the Alps are systematic only in winter across all EURO-CORDEX models above 1000 m, with the magnitude of the bias increasing with elevation. Contrary to what might be expected, these winter biases in mountainous regions cannot be solely attributed to altitude underestimation by models (which would typically lead to warmer biases). Instead, as discussed throughout this notebook, these cold biases arise from complex interactions within the models’ numerical schemes, particularly in the treatment of snow and ice processes, which result in the underestimation of surface temperatures.\n\nBias-adjustment techniques can partially relieve the assessed cold temperature biases. In particular, it can reduce the present-day discrepancies between modelled and observed temperature fields (Matiu et al., 2024 [[14]](https://doi.org/10.1007/s00382-024-07376-y), *Figure .6b*). However, bias-adjustment methods do not guarantee that the adjusted projected results will better match future reality, since the bias may not be stationary in a changing climate (e.g., Schmith et al., 2021 [[31]](https://doi.org/10.5194/hess-25-273-2021)). Special consideration should also be given to the representation of physical relationships among variables subjected to bias adjustment (Gennaretti et al., 2015 [[32]](https://doi.org/10.1002/2015JD023890)), as well as to unphysical artefacts such as the damping or amplification of original simulation trends, particularly in regions with complex orography (Sangelantoni et al., 2019 [[33]](https://doi.org/10.1007/s00704-018-2406-8)). Furthermore, different bias-adjustment methods do not always improve downstream products, as no single technique works reliably across all locations, variables, and applications.\n\nThe assessed temperature and mountain snowpack biases influence the current regional climate models' ability to quantify water availability in mountain areas and, consequently, the capability to estimate changes in resources for local and downstream hydrology and economic sectors, such as tourism, hydropower, and agriculture.\n\nattachment:013623ec-4a51-4ab5-bf67-483d3b6e405a.png\n---\nwidth: 900px\nalt: figure 6\n---\nImpact of different bias adjustment methods on (a) precipitation indices and (b) temperature indices over 1971–2000. Each line is a GCM-RCM combination. Grey dashed lines are raw (uncorrected) models, while solid colored lines are bias-adjusted using different methods (see legend). Image reproduced from *Figure 8* in Matiu et al. (2024) [[14]](https://doi.org/10.1007/s00382-024-07376-y), under [CC BY 4.0](https://creativecommons.org/licenses/by/4.0/).\n\n```"} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q05__7e76a132c9ae", "report_id": "climate_projections-cordex-domains-single-levels_validation_q05", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q05", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CORDEX temperature biases over the Alpine region for the ski and hydrological sectors. > ℹ️ If you want to know more > Key resources", "title": "CORDEX temperature biases over the Alpine region for the ski and hydrological sectors.", "chunk_index": 9, "token_count": 307, "text_raw": "Some key resources and further reading were linked throughout this assessment.\n\n* Regional climate models - CORDEX: https://cds.climate.copernicus.eu/datasets/projections-cordex-domains-single-levels?tab=overview\n\n* Gridded observation dataset for Europe, E-OBS:\nhttps://cds.climate.copernicus.eu/datasets/insitu-gridded-observations-europe?tab=overview\n\n* High-resolution global climate models:\nhttps://highresmip.org/\n\n* Mountain tourism meteorological and snow indicators for Europe from 1950 to 2100 derived from reanalysis and climate projections: https://cds.climate.copernicus.eu/datasets/sis-tourism-snow-indicators?tab=overview\n\nSpecial mention to:\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)", "text_with_prefix": "EQC Quality Assessment: \"CORDEX temperature biases over the Alpine region for the ski and hydrological sectors.\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q05 | Category: Climate_Projections\nSection: CORDEX temperature biases over the Alpine region for the ski and hydrological sectors. > ℹ️ If you want to know more > Key resources\n---\nSome key resources and further reading were linked throughout this assessment.\n\n* Regional climate models - CORDEX: https://cds.climate.copernicus.eu/datasets/projections-cordex-domains-single-levels?tab=overview\n\n* Gridded observation dataset for Europe, E-OBS:\nhttps://cds.climate.copernicus.eu/datasets/insitu-gridded-observations-europe?tab=overview\n\n* High-resolution global climate models:\nhttps://highresmip.org/\n\n* Mountain tourism meteorological and snow indicators for Europe from 1950 to 2100 derived from reanalysis and climate projections: https://cds.climate.copernicus.eu/datasets/sis-tourism-snow-indicators?tab=overview\n\nSpecial mention to:\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)"} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q05__d63352d83568", "report_id": "climate_projections-cordex-domains-single-levels_validation_q05", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q05", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CORDEX temperature biases over the Alpine region for the ski and hydrological sectors. > ℹ️ If you want to know more > References", "title": "CORDEX temperature biases over the Alpine region for the ski and hydrological sectors.", "chunk_index": 10, "token_count": 1079, "text_raw": "[[1]](https://doi.org/10.1038/s41586-019-1822-y) Immerzeel, W.W., Lutz, A.F., Andrade, M. et al., 2020. Importance and vulnerability of the world’s water towers. Nature 577, 364–369. https://doi.org/10.1038/s41586-019-1822-y\n\n[[2]](https://doi.org/10.1175/BAMS-D-23-0082.1) Rudisill, W., Rhoades, A., Xu, Z., and Feldman, D. R.,2024. Are Atmospheric Models Too Cold in the Mountains? The State of Science and Insights from the SAIL Field Campaign. Bull. Amer. Meteor. Soc., 105, E1237–E1264, https://doi.org/10.1175/BAMS-D-23-0082.1\n\n[[3]](https://doi.org/10.5194/gmd-7-1297-2014) Kotlarski, S., Keuler, K., Christensen, O. B., Colette, A., Déqué, M., Gobiet, A., Goergen, K., Jacob, D., Lüthi, D., van Meijgaard, E., Nikulin, G., Schär, C., Teichmann, C., Vautard, R., Warrach-Sagi, K., and Wulfmeyer, V., 2014. Regional climate modeling on European scales: a joint standard evaluation of the EURO-CORDEX RCM ensemble, Geosci. Model Dev., 7, 1297–1333, https://doi.org/10.5194/gmd-7-1297-2014\n\n[[4]](https://doi.org/10.1007/s00382-016-3130-7) Winter, K.J.P.M., Kotlarski, S., Scherrer, S.C. et al., 2017. The Alpine snow-albedo feedback in regional climate models. Clim Dyn 48, 1109–1124. https://doi.org/10.1007/s00382-016-3130-7\n\n[[5]](https://doi.org/10.3390/atmos11010046) Matiu, M., Petitta, M., Notarnicola, C., and Zebisch, M., 2019. Evaluating Snow in EURO-CORDEX Regional Climate Models with Observations for the European Alps: Biases and Their Relationship to Orography, Temperature, and Precipitation Mismatches. Atmosphere, 11(1), 46. https://doi.org/10.3390/atmos11010046\n\n[[6]](https://doi.org/10.1029/2019JD032344) Vautard, R., Kadygrov, N., Iles, C., Boberg, F., Buonomo, E., Bülow, K., et al., 2021. Evaluation of the large EURO-CORDEX regional climate model ensemble. Journal of Geophysical Research: Atmospheres, 126, e2019JD032344. https://doi.org/10.1029/2019JD032344\n\n[[7]](https://doi.org/10.1002/joc.8331) Sy, S., Madonna, F., Serva, F., Diallo, I., and Quesada, B., 2024. Assessment of NA-CORDEX regional climate models, reanalysis and in situ gridded-observational data sets against the U.S. Climate Reference Network. International Journal of Climatology, 44(1), 305–327. https://doi.org/10.1002/joc.8331\n\n[[8]](https://doi.org/10.1002/joc.8222) Napoli, A., Parodi, A., von Hardenberg, J., and Pasquero, C. ,2023. Altitudinal dependence of projected changes in occurrence of extreme events in the Great Alpine Region. International Journal of Climatology, 43(12), 5813–5829. https://doi.org/10.1002/joc.8222\n\n[[9]](https://doi.org/10.1007/s00382-016-3327-9) Liu, C., Ikeda, K., Rasmussen, R. et al., 2017. Continental-scale convection-permitting modeling of the current and future climate of North America. Clim Dyn 49, 71–95. https://doi.org/10.1007/s00382-016-3327-9", "text_with_prefix": "EQC Quality Assessment: \"CORDEX temperature biases over the Alpine region for the ski and hydrological sectors.\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q05 | Category: Climate_Projections\nSection: CORDEX temperature biases over the Alpine region for the ski and hydrological sectors. > ℹ️ If you want to know more > References\n---\n[[1]](https://doi.org/10.1038/s41586-019-1822-y) Immerzeel, W.W., Lutz, A.F., Andrade, M. et al., 2020. Importance and vulnerability of the world’s water towers. Nature 577, 364–369. https://doi.org/10.1038/s41586-019-1822-y\n\n[[2]](https://doi.org/10.1175/BAMS-D-23-0082.1) Rudisill, W., Rhoades, A., Xu, Z., and Feldman, D. R.,2024. Are Atmospheric Models Too Cold in the Mountains? The State of Science and Insights from the SAIL Field Campaign. Bull. Amer. Meteor. Soc., 105, E1237–E1264, https://doi.org/10.1175/BAMS-D-23-0082.1\n\n[[3]](https://doi.org/10.5194/gmd-7-1297-2014) Kotlarski, S., Keuler, K., Christensen, O. B., Colette, A., Déqué, M., Gobiet, A., Goergen, K., Jacob, D., Lüthi, D., van Meijgaard, E., Nikulin, G., Schär, C., Teichmann, C., Vautard, R., Warrach-Sagi, K., and Wulfmeyer, V., 2014. Regional climate modeling on European scales: a joint standard evaluation of the EURO-CORDEX RCM ensemble, Geosci. Model Dev., 7, 1297–1333, https://doi.org/10.5194/gmd-7-1297-2014\n\n[[4]](https://doi.org/10.1007/s00382-016-3130-7) Winter, K.J.P.M., Kotlarski, S., Scherrer, S.C. et al., 2017. The Alpine snow-albedo feedback in regional climate models. Clim Dyn 48, 1109–1124. https://doi.org/10.1007/s00382-016-3130-7\n\n[[5]](https://doi.org/10.3390/atmos11010046) Matiu, M., Petitta, M., Notarnicola, C., and Zebisch, M., 2019. Evaluating Snow in EURO-CORDEX Regional Climate Models with Observations for the European Alps: Biases and Their Relationship to Orography, Temperature, and Precipitation Mismatches. Atmosphere, 11(1), 46. https://doi.org/10.3390/atmos11010046\n\n[[6]](https://doi.org/10.1029/2019JD032344) Vautard, R., Kadygrov, N., Iles, C., Boberg, F., Buonomo, E., Bülow, K., et al., 2021. Evaluation of the large EURO-CORDEX regional climate model ensemble. Journal of Geophysical Research: Atmospheres, 126, e2019JD032344. https://doi.org/10.1029/2019JD032344\n\n[[7]](https://doi.org/10.1002/joc.8331) Sy, S., Madonna, F., Serva, F., Diallo, I., and Quesada, B., 2024. Assessment of NA-CORDEX regional climate models, reanalysis and in situ gridded-observational data sets against the U.S. Climate Reference Network. International Journal of Climatology, 44(1), 305–327. https://doi.org/10.1002/joc.8331\n\n[[8]](https://doi.org/10.1002/joc.8222) Napoli, A., Parodi, A., von Hardenberg, J., and Pasquero, C. ,2023. Altitudinal dependence of projected changes in occurrence of extreme events in the Great Alpine Region. International Journal of Climatology, 43(12), 5813–5829. https://doi.org/10.1002/joc.8222\n\n[[9]](https://doi.org/10.1007/s00382-016-3327-9) Liu, C., Ikeda, K., Rasmussen, R. et al., 2017. Continental-scale convection-permitting modeling of the current and future climate of North America. Clim Dyn 49, 71–95. https://doi.org/10.1007/s00382-016-3327-9"} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q05__2aeb05f5d110", "report_id": "climate_projections-cordex-domains-single-levels_validation_q05", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q05", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CORDEX temperature biases over the Alpine region for the ski and hydrological sectors. > ℹ️ If you want to know more > References", "title": "CORDEX temperature biases over the Alpine region for the ski and hydrological sectors.", "chunk_index": 11, "token_count": 1064, "text_raw": "7. Continental-scale convection-permitting modeling of the current and future climate of North America. Clim Dyn 49, 71–95. https://doi.org/10.1007/s00382-016-3327-9\n\n[[10]](https://doi.org/10.1029/2022EF003408) van Tiel, M., Weiler, M., Freudiger, D., Moretti, G., Kohn, I., Gerlinger, K., and Stahl, K., 2023. Melting alpine water towers aggravate downstream low flows: A stress-test storyline approach. Earth's Future, 11, e2022EF003408. https://doi.org/10.1029/2022EF003408\n\n[[11]](https://doi.org/10.1016/j.envsci.2010.12.009) Beniston, M., Stoffel, M., and Hill, M., 2011. Impacts of climatic change on water and natural hazards in the Alps: Can current water governance cope with future challenges? Examples from the European “ACQWA” project, Environ. Sci. Policy, 14, 734–743. https://doi.org/10.1016/j.envsci.2010.12.009\n\n[[12]](https://doi.org/10.1029/2008JD010201) Haylock, M. R., Hofstra, N., Tank, A. M. G., Klok, E. J., Jones, P. D., and New, M., 2008. A European daily high‐resolution gridded data set of surface temperature and precipitation for 1950–2006. Journal of Geophysical Research, 113, D20119. https://doi.org/10.1029/2008JD010201\n\n[[13]](https://doi.org/10.1029/2017JD028200) Cornes, R. C., van der Schrier, G., van den Besselaar, E. J. M., and Jones, P. D., 2018. An ensemble version of the E-OBS temperature and precipitation data sets. Journal of Geophysical Research: Atmospheres, 123, 9391–9409. https://doi.org/10.1029/2017JD028200\n\n[[14]](https://doi.org/10.1007/s00382-024-07376-y) Matiu, M., Napoli, A., Kotlarski, S. et al., 2024. Elevation-dependent biases of raw and bias-adjusted EURO-CORDEX regional climate models in the European Alps. Clim Dyn 62, 9013–9030. https://doi.org/10.1007/s00382-024-07376-y\n\n[[15]](https://doi.org/10.1002/joc.3794) Isotta, F.A., Frei, C., Weilguni, V., Tadic Melita, P., Lassegues, P., Rudolf, B., Pavan, V., Cacciamani, C., Antolini, G., Ratto, S.M., Munari, M., Micheletti, S., Bonati, V., Lussana, C., Ronchi, C., Panettieri, E., Marigo, G., Vertacnik, G.. 2014. The climate of daily precipitation in the Alps: development and analysis of a high-resolution grid dataset from pan-Alpine rain-gauge data, International Journal of Climatology, 34, 1657-1675, https://doi.org/10.1002/joc.3794\n\n[[16]](https://doi.org/10.1002/qj.2813) Landelius, T., Dahlgren, P., Gollvik, S. et al., 2016. A high-resolution regional reanalysis for Europe. Part 2: 2D analysis of surface temperature, precipitation and wind. Q J R Meteorol Soc 142(698):2132–2142. https://doi.org/10.1002/qj.2813\n\n[[17]](https://doi.org/10.1002/2014WR015638) Weedon, G.P., Balsamo, G., Bellouin, N. et al., 2014. The WFDEI meteorological forcing data set: WATCH forcing data methodology applied to ERA-Interim reanalysis data. Water Resour Res 50(9):7505–7514. https://doi.org/10.1002/2014WR015638", "text_with_prefix": "EQC Quality Assessment: \"CORDEX temperature biases over the Alpine region for the ski and hydrological sectors.\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q05 | Category: Climate_Projections\nSection: CORDEX temperature biases over the Alpine region for the ski and hydrological sectors. > ℹ️ If you want to know more > References\n---\n7. Continental-scale convection-permitting modeling of the current and future climate of North America. Clim Dyn 49, 71–95. https://doi.org/10.1007/s00382-016-3327-9\n\n[[10]](https://doi.org/10.1029/2022EF003408) van Tiel, M., Weiler, M., Freudiger, D., Moretti, G., Kohn, I., Gerlinger, K., and Stahl, K., 2023. Melting alpine water towers aggravate downstream low flows: A stress-test storyline approach. Earth's Future, 11, e2022EF003408. https://doi.org/10.1029/2022EF003408\n\n[[11]](https://doi.org/10.1016/j.envsci.2010.12.009) Beniston, M., Stoffel, M., and Hill, M., 2011. Impacts of climatic change on water and natural hazards in the Alps: Can current water governance cope with future challenges? Examples from the European “ACQWA” project, Environ. Sci. Policy, 14, 734–743. https://doi.org/10.1016/j.envsci.2010.12.009\n\n[[12]](https://doi.org/10.1029/2008JD010201) Haylock, M. R., Hofstra, N., Tank, A. M. G., Klok, E. J., Jones, P. D., and New, M., 2008. A European daily high‐resolution gridded data set of surface temperature and precipitation for 1950–2006. Journal of Geophysical Research, 113, D20119. https://doi.org/10.1029/2008JD010201\n\n[[13]](https://doi.org/10.1029/2017JD028200) Cornes, R. C., van der Schrier, G., van den Besselaar, E. J. M., and Jones, P. D., 2018. An ensemble version of the E-OBS temperature and precipitation data sets. Journal of Geophysical Research: Atmospheres, 123, 9391–9409. https://doi.org/10.1029/2017JD028200\n\n[[14]](https://doi.org/10.1007/s00382-024-07376-y) Matiu, M., Napoli, A., Kotlarski, S. et al., 2024. Elevation-dependent biases of raw and bias-adjusted EURO-CORDEX regional climate models in the European Alps. Clim Dyn 62, 9013–9030. https://doi.org/10.1007/s00382-024-07376-y\n\n[[15]](https://doi.org/10.1002/joc.3794) Isotta, F.A., Frei, C., Weilguni, V., Tadic Melita, P., Lassegues, P., Rudolf, B., Pavan, V., Cacciamani, C., Antolini, G., Ratto, S.M., Munari, M., Micheletti, S., Bonati, V., Lussana, C., Ronchi, C., Panettieri, E., Marigo, G., Vertacnik, G.. 2014. The climate of daily precipitation in the Alps: development and analysis of a high-resolution grid dataset from pan-Alpine rain-gauge data, International Journal of Climatology, 34, 1657-1675, https://doi.org/10.1002/joc.3794\n\n[[16]](https://doi.org/10.1002/qj.2813) Landelius, T., Dahlgren, P., Gollvik, S. et al., 2016. A high-resolution regional reanalysis for Europe. Part 2: 2D analysis of surface temperature, precipitation and wind. Q J R Meteorol Soc 142(698):2132–2142. https://doi.org/10.1002/qj.2813\n\n[[17]](https://doi.org/10.1002/2014WR015638) Weedon, G.P., Balsamo, G., Bellouin, N. et al., 2014. The WFDEI meteorological forcing data set: WATCH forcing data methodology applied to ERA-Interim reanalysis data. Water Resour Res 50(9):7505–7514. https://doi.org/10.1002/2014WR015638"} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q05__21846fccae72", "report_id": "climate_projections-cordex-domains-single-levels_validation_q05", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q05", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CORDEX temperature biases over the Alpine region for the ski and hydrological sectors. > ℹ️ If you want to know more > References", "title": "CORDEX temperature biases over the Alpine region for the ski and hydrological sectors.", "chunk_index": 12, "token_count": 1095, "text_raw": "set: WATCH forcing data methodology applied to ERA-Interim reanalysis data. Water Resour Res 50(9):7505–7514. https://doi.org/10.1002/2014WR015638\n\n[[18]](https://doi.org/10.1007/s00382-023-06727-5) Blázquez, J., and Solman, S. A., 2023. Temperature and precipitation biases in CORDEX RCM simulations over South America: Possible origin and impacts on the regional climate change signal. Climate Dyn., 61, 2907–2920, https://doi.org/10.1007/s00382-023-06727-5\n\n[[19]](https://doi.org/10.1007/s00382-015-2529-x) García-Díez, M., Fernández, J., and Vautard, R., 2015. An RCM multi-physics ensemble over Europe: multi-variable evaluation to avoid error compensation. Clim Dyn 45, 3141–3156. https://doi.org/10.1007/s00382-015-2529-x\n\n[[20]](https://doi.org/10.1175/JHM-D-15-0241.1) Vionnet, V., Dombrowski-Etchevers, I., Lafaysse, M., Quéno, L., Seity, Y., and Bazile, E., 2016. Numerical weather forecasts at kilometer scale in the French Alps: evaluation and application for snowpack modeling, J. Hydrometeorol., 17, 2591–2614, https://doi.org/10.1175/JHM-D-15-0241.1\n\n[[21]](https://doi.org/10.3390/cli11070154) Shrestha, S., Zaramella, M., Callegari, M., Greifeneder, F., Borga, M., 2023. Scale Dependence of Errors in Snow Water Equivalent Simulations Using ERA5 Reanalysis over Alpine Basins. Climate, 11, 154. https://doi.org/10.3390/cli11070154\n\n[[22]](https://doi.org/10.1007/s00704-016-2031-3) Lhotka, O., Kyselý, J. and Farda, A., 2018. Climate change scenarios of heat waves in Central Europe and their uncertainties. Theor Appl Climatol 131, 1043–1054. https://doi.org/10.1007/s00704-016-2031-3\n\n[[23]](https://doi.org/10.1029/2020JD032824) Maraun, D., Truhetz, H., and Schaffer, A., 2021. Regional climate modelbiases, their dependence on synopticcirculation biases and the potentialfor bias adjustment: A process-oriented evaluation of the Austrianregional climate projections. Journalof Geophysical Research: Atmospheres,126, e2020JD032824. https://doi.org/10.1029/2020JD032824\n\n[[24]](https://doi.org/10.1007/s00382-024-07318-8) Bilbao-Barrenetxea, N., Santolaria-Otín, M., Teichmann, C. et al., 2024. Added value of EURO-CORDEX downscaling over the complex orography region of the Pyrenees. Clim Dyn 62, 7981–7996. https://doi.org/10.1007/s00382-024-07318-8\n\n[[25]](https://doi.org/10.5194/gmd-13-5485-2020) Demory, M.-E., Berthou, S., Fernández, J., Sørland, S. L., Brogli, R., Roberts, M. J., Beyerle, U., Seddon, J., Haarsma, R., Schär, C., Buonomo, E., Christensen, O. B., Ciarlo, J. M., Fealy, R., Nikulin, G., Peano, D., Putrasahan, D., Roberts, C. D., Senan, R., Steger, C., Teichmann, C., and Vautard, R., 2020. European daily precipitation according to EURO-CORDEX regional climate models (RCMs) and high-resolution global climate models (GCMs) from the High-Resolution Model Intercomparison Project (HighResMIP), Geosci. Model Dev., 13, 5485–5506, https://doi.org/10.5194/gmd-13-5485-2020", "text_with_prefix": "EQC Quality Assessment: \"CORDEX temperature biases over the Alpine region for the ski and hydrological sectors.\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q05 | Category: Climate_Projections\nSection: CORDEX temperature biases over the Alpine region for the ski and hydrological sectors. > ℹ️ If you want to know more > References\n---\nset: WATCH forcing data methodology applied to ERA-Interim reanalysis data. Water Resour Res 50(9):7505–7514. https://doi.org/10.1002/2014WR015638\n\n[[18]](https://doi.org/10.1007/s00382-023-06727-5) Blázquez, J., and Solman, S. A., 2023. Temperature and precipitation biases in CORDEX RCM simulations over South America: Possible origin and impacts on the regional climate change signal. Climate Dyn., 61, 2907–2920, https://doi.org/10.1007/s00382-023-06727-5\n\n[[19]](https://doi.org/10.1007/s00382-015-2529-x) García-Díez, M., Fernández, J., and Vautard, R., 2015. An RCM multi-physics ensemble over Europe: multi-variable evaluation to avoid error compensation. Clim Dyn 45, 3141–3156. https://doi.org/10.1007/s00382-015-2529-x\n\n[[20]](https://doi.org/10.1175/JHM-D-15-0241.1) Vionnet, V., Dombrowski-Etchevers, I., Lafaysse, M., Quéno, L., Seity, Y., and Bazile, E., 2016. Numerical weather forecasts at kilometer scale in the French Alps: evaluation and application for snowpack modeling, J. Hydrometeorol., 17, 2591–2614, https://doi.org/10.1175/JHM-D-15-0241.1\n\n[[21]](https://doi.org/10.3390/cli11070154) Shrestha, S., Zaramella, M., Callegari, M., Greifeneder, F., Borga, M., 2023. Scale Dependence of Errors in Snow Water Equivalent Simulations Using ERA5 Reanalysis over Alpine Basins. Climate, 11, 154. https://doi.org/10.3390/cli11070154\n\n[[22]](https://doi.org/10.1007/s00704-016-2031-3) Lhotka, O., Kyselý, J. and Farda, A., 2018. Climate change scenarios of heat waves in Central Europe and their uncertainties. Theor Appl Climatol 131, 1043–1054. https://doi.org/10.1007/s00704-016-2031-3\n\n[[23]](https://doi.org/10.1029/2020JD032824) Maraun, D., Truhetz, H., and Schaffer, A., 2021. Regional climate modelbiases, their dependence on synopticcirculation biases and the potentialfor bias adjustment: A process-oriented evaluation of the Austrianregional climate projections. Journalof Geophysical Research: Atmospheres,126, e2020JD032824. https://doi.org/10.1029/2020JD032824\n\n[[24]](https://doi.org/10.1007/s00382-024-07318-8) Bilbao-Barrenetxea, N., Santolaria-Otín, M., Teichmann, C. et al., 2024. Added value of EURO-CORDEX downscaling over the complex orography region of the Pyrenees. Clim Dyn 62, 7981–7996. https://doi.org/10.1007/s00382-024-07318-8\n\n[[25]](https://doi.org/10.5194/gmd-13-5485-2020) Demory, M.-E., Berthou, S., Fernández, J., Sørland, S. L., Brogli, R., Roberts, M. J., Beyerle, U., Seddon, J., Haarsma, R., Schär, C., Buonomo, E., Christensen, O. B., Ciarlo, J. M., Fealy, R., Nikulin, G., Peano, D., Putrasahan, D., Roberts, C. D., Senan, R., Steger, C., Teichmann, C., and Vautard, R., 2020. European daily precipitation according to EURO-CORDEX regional climate models (RCMs) and high-resolution global climate models (GCMs) from the High-Resolution Model Intercomparison Project (HighResMIP), Geosci. Model Dev., 13, 5485–5506, https://doi.org/10.5194/gmd-13-5485-2020"} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q05__f0fb5340d59b", "report_id": "climate_projections-cordex-domains-single-levels_validation_q05", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q05", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CORDEX temperature biases over the Alpine region for the ski and hydrological sectors. > ℹ️ If you want to know more > References", "title": "CORDEX temperature biases over the Alpine region for the ski and hydrological sectors.", "chunk_index": 13, "token_count": 1109, "text_raw": "High-Resolution Model Intercomparison Project (HighResMIP), Geosci. Model Dev., 13, 5485–5506, https://doi.org/10.5194/gmd-13-5485-2020\n\n[[26]](https://doi.org/10.5194/gmd-15-269-2022) Moreno-Chamarro, E., Caron, L. P., Loosveldt Tomas, S., Vegas-Regidor, J., Gutjahr, O., Moine, M. P., et al., 2022. Impact of increased resolution on long-standing biases in HighResMIP-PRIMAVERA climate models. Geoscientific Model Development, 15(1), 269–289. https://doi.org/10.5194/gmd-15-269-2022\n\n[[27]](https://doi.org/10.5194/gmd-18-1307-2025) Roberts, M. J., Reed, K. A., Bao, Q., Barsugli, J. J., Camargo, S. J., Caron, L. P., et al., 2025. High-Resolution Model Intercomparison Project phase 2 (HighResMIP2) towards CMIP7. Geoscientific Model Development, 18(4), 1307–1332. https://doi.org/10.5194/gmd-18-1307-2025\n\n[[28]](https://doi.org/10.1007/s00382-023-06769-9) Sangelantoni, L., Sobolowski, S., Lorenz, T., Hodnebrog, Cardoso, R. M., Soares, P. M. M., et al., 2024. Investigating the representation of heatwaves from an ensemble of km-scale regional climate simulations within CORDEX-FPS convection. Climate Dynamics. https://doi.org/10.1007/s00382-023-06769-9\n\n[[29]](https://doi.org/10.1029/2024GL111147) Sangelantoni, L., Sobolowski, S. P., Soares, P. M. M., Goergen, K., Cardoso, R. M., Adinolfi, M., et al., 2025. Heatwave future changes from an ensemble of km‐scale regional climate simulations within CORDEX‐FPS convection. Geophysical Research Letters, 52, e2024GL111147. https://doi.org/10.1029/2024GL111147\n\n[[30]](https://doi.org/10.1007/s00382-022-06593-7) Soares, P. M. M., Careto, J. A. M., Cardoso, R. M., Goergen, K., Katragkou, E., Sobolowski, S., et al., 2022. The added value of km-scale simulations to describe temperature over complex orography: the CORDEX FPS-Convection multi-model ensemble runs over the Alps. Climate Dynamics, (0123456789). https://doi.org/10.1007/s00382-022-06593-7\n\n[[31]](https://doi.org/10.5194/hess-25-273-2021) Schmith, T., Thejll, P., Berg, P., Boberg, F., Christensen, O. B., Christiansen, B., Christensen, J. H., Madsen, M. S., and Steger, C., 2021. Identifying robust bias adjustment methods for European extreme precipitation in a multi-model pseudo-reality setting, Hydrol. Earth Syst. Sci., 25, 273–290, https://doi.org/10.5194/hess-25-273-2021\n\n[[32]](https://doi.org/10.1002/2015JD023890) Gennaretti, F., L. Sangelantoni, and P. Grenier, 2015. Toward daily climate scenarios for Canadian Arctic coastal zones with more realistic temperature-precipitation interdependence, J. Geophys. Res. Atmos., 120, 11,862–11,877, https://doi.org/10.1002/2015JD023890\n\n[[33]](https://doi.org/10.1007/s00704-018-2406-8) Sangelantoni, L., Russo, A. and Gennaretti, F., 2019. Impact of bias correction and downscaling through quantile mapping on simulated climate change signal: a case study over Central Italy. Theor Appl Climatol 135, 725–740. https://doi.org/10.1007/s00704-018-2406-8", "text_with_prefix": "EQC Quality Assessment: \"CORDEX temperature biases over the Alpine region for the ski and hydrological sectors.\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q05 | Category: Climate_Projections\nSection: CORDEX temperature biases over the Alpine region for the ski and hydrological sectors. > ℹ️ If you want to know more > References\n---\nHigh-Resolution Model Intercomparison Project (HighResMIP), Geosci. Model Dev., 13, 5485–5506, https://doi.org/10.5194/gmd-13-5485-2020\n\n[[26]](https://doi.org/10.5194/gmd-15-269-2022) Moreno-Chamarro, E., Caron, L. P., Loosveldt Tomas, S., Vegas-Regidor, J., Gutjahr, O., Moine, M. P., et al., 2022. Impact of increased resolution on long-standing biases in HighResMIP-PRIMAVERA climate models. Geoscientific Model Development, 15(1), 269–289. https://doi.org/10.5194/gmd-15-269-2022\n\n[[27]](https://doi.org/10.5194/gmd-18-1307-2025) Roberts, M. J., Reed, K. A., Bao, Q., Barsugli, J. J., Camargo, S. J., Caron, L. P., et al., 2025. High-Resolution Model Intercomparison Project phase 2 (HighResMIP2) towards CMIP7. Geoscientific Model Development, 18(4), 1307–1332. https://doi.org/10.5194/gmd-18-1307-2025\n\n[[28]](https://doi.org/10.1007/s00382-023-06769-9) Sangelantoni, L., Sobolowski, S., Lorenz, T., Hodnebrog, Cardoso, R. M., Soares, P. M. M., et al., 2024. Investigating the representation of heatwaves from an ensemble of km-scale regional climate simulations within CORDEX-FPS convection. Climate Dynamics. https://doi.org/10.1007/s00382-023-06769-9\n\n[[29]](https://doi.org/10.1029/2024GL111147) Sangelantoni, L., Sobolowski, S. P., Soares, P. M. M., Goergen, K., Cardoso, R. M., Adinolfi, M., et al., 2025. Heatwave future changes from an ensemble of km‐scale regional climate simulations within CORDEX‐FPS convection. Geophysical Research Letters, 52, e2024GL111147. https://doi.org/10.1029/2024GL111147\n\n[[30]](https://doi.org/10.1007/s00382-022-06593-7) Soares, P. M. M., Careto, J. A. M., Cardoso, R. M., Goergen, K., Katragkou, E., Sobolowski, S., et al., 2022. The added value of km-scale simulations to describe temperature over complex orography: the CORDEX FPS-Convection multi-model ensemble runs over the Alps. Climate Dynamics, (0123456789). https://doi.org/10.1007/s00382-022-06593-7\n\n[[31]](https://doi.org/10.5194/hess-25-273-2021) Schmith, T., Thejll, P., Berg, P., Boberg, F., Christensen, O. B., Christiansen, B., Christensen, J. H., Madsen, M. S., and Steger, C., 2021. Identifying robust bias adjustment methods for European extreme precipitation in a multi-model pseudo-reality setting, Hydrol. Earth Syst. Sci., 25, 273–290, https://doi.org/10.5194/hess-25-273-2021\n\n[[32]](https://doi.org/10.1002/2015JD023890) Gennaretti, F., L. Sangelantoni, and P. Grenier, 2015. Toward daily climate scenarios for Canadian Arctic coastal zones with more realistic temperature-precipitation interdependence, J. Geophys. Res. Atmos., 120, 11,862–11,877, https://doi.org/10.1002/2015JD023890\n\n[[33]](https://doi.org/10.1007/s00704-018-2406-8) Sangelantoni, L., Russo, A. and Gennaretti, F., 2019. Impact of bias correction and downscaling through quantile mapping on simulated climate change signal: a case study over Central Italy. Theor Appl Climatol 135, 725–740. https://doi.org/10.1007/s00704-018-2406-8"} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q06__774e57f2d05b", "report_id": "climate_projections-cordex-domains-single-levels_validation_q06", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q06", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA", "title": "CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA", "chunk_index": 0, "token_count": 116, "text_raw": "Production date: 30-06-2025\n\nProduced by: CMCC foundation - Euro-Mediterranean Center on Climate Change. Albert Martinez Boti and Lorenzo Sangelantoni.", "text_with_prefix": "EQC Quality Assessment: \"CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q06 | Category: Climate_Projections\nSection: CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA\n---\nProduction date: 30-06-2025\n\nProduced by: CMCC foundation - Euro-Mediterranean Center on Climate Change. Albert Martinez Boti and Lorenzo Sangelantoni."} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q06__53a0e7adef98", "report_id": "climate_projections-cordex-domains-single-levels_validation_q06", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q06", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA > Quality assessment question", "title": "CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA", "chunk_index": 1, "token_count": 574, "text_raw": "* **Are the cold biases commonly found in Regional Climate Models over the Alps apparent when evaluated against CERRA?**\n* **Does applying a lapse-rate correction reduce these biases?**\n\nMountains, although covering a limited land area, are vital to the global water cycle, acting as key sources of freshwater for both ecological and human needs [[1]](https://doi.org/10.1038/s41586-019-1822-y). In many mountainous regions, temperature is a critical driver of hydrometeorological processes such as the ratio of rain to snow, the timing of snowmelt, and the frequency of low-flow conditions [[2]](https://doi.org/10.1175/BAMS-D-23-0082.1). These sensitivities make mountain regions key indicators of climate change. However, accurately representing temperature in complex terrain remains challenging for models due to factors like topographic variability, slope orientation, terrain shading, snow–albedo feedbacks, and vegetation heterogeneity [[2]](https://doi.org/10.1175/BAMS-D-23-0082.1)).\n\nA recent literature-based quality assessment, [CORDEX temperature biases over the Alpine region for the ski and hydrological sectors](https://ecmwf-projects.github.io/c3s2-eqc-quality-assessment/Climate_Projections/CORDEX/climate_projections-cordex-domains-single-levels_validation_q05.html), synthesised multiple studies and confirmed that Regional Climate Models (RCMs) exhibit persistent cold biases over the Alps despite their relatively high resolution and improved representation of physical processes in mountainous terrain. Given that bias estimates are sensitive to the choice of reference dataset -as highlighted in the quality assessment [[3]](https://doi.org/10.1007/s00382-024-07376-y)[[4]](https://doi.org/10.3390/cli11070154)- here we evaluate RCM biases against E-OBS and include CERRA, a high-resolution (5.5 km) regional reanalysis produced using the HARMONIE-ALADIN limited-area numerical weather prediction and data assimilation system. The analysis includes a broad ensemble of EURO-CORDEX models and examines the altitude dependency of biases by evaluating results across different elevation bands. Finally, we apply a lapse-rate correction to place all datasets at a common elevation and investigate whether this influences the estimated biases.", "text_with_prefix": "EQC Quality Assessment: \"CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q06 | Category: Climate_Projections\nSection: CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA > Quality assessment question\n---\n* **Are the cold biases commonly found in Regional Climate Models over the Alps apparent when evaluated against CERRA?**\n* **Does applying a lapse-rate correction reduce these biases?**\n\nMountains, although covering a limited land area, are vital to the global water cycle, acting as key sources of freshwater for both ecological and human needs [[1]](https://doi.org/10.1038/s41586-019-1822-y). In many mountainous regions, temperature is a critical driver of hydrometeorological processes such as the ratio of rain to snow, the timing of snowmelt, and the frequency of low-flow conditions [[2]](https://doi.org/10.1175/BAMS-D-23-0082.1). These sensitivities make mountain regions key indicators of climate change. However, accurately representing temperature in complex terrain remains challenging for models due to factors like topographic variability, slope orientation, terrain shading, snow–albedo feedbacks, and vegetation heterogeneity [[2]](https://doi.org/10.1175/BAMS-D-23-0082.1)).\n\nA recent literature-based quality assessment, [CORDEX temperature biases over the Alpine region for the ski and hydrological sectors](https://ecmwf-projects.github.io/c3s2-eqc-quality-assessment/Climate_Projections/CORDEX/climate_projections-cordex-domains-single-levels_validation_q05.html), synthesised multiple studies and confirmed that Regional Climate Models (RCMs) exhibit persistent cold biases over the Alps despite their relatively high resolution and improved representation of physical processes in mountainous terrain. Given that bias estimates are sensitive to the choice of reference dataset -as highlighted in the quality assessment [[3]](https://doi.org/10.1007/s00382-024-07376-y)[[4]](https://doi.org/10.3390/cli11070154)- here we evaluate RCM biases against E-OBS and include CERRA, a high-resolution (5.5 km) regional reanalysis produced using the HARMONIE-ALADIN limited-area numerical weather prediction and data assimilation system. The analysis includes a broad ensemble of EURO-CORDEX models and examines the altitude dependency of biases by evaluating results across different elevation bands. Finally, we apply a lapse-rate correction to place all datasets at a common elevation and investigate whether this influences the estimated biases."} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q06__ee9ecd08ea33", "report_id": "climate_projections-cordex-domains-single-levels_validation_q06", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q06", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA > Quality assessment statement", "title": "CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA", "chunk_index": 2, "token_count": 496, "text_raw": "These are the key outcomes of this assessment\n\n* CORDEX simulations exhibit biases relative to both reference datasets (E-OBS and CERRA).\n\n* RCMs’ systematic cold biases above 1000–1500 m, reported in the literature, are evident when evaluated against E-OBS.\n\n* CORDEX biases relative to CERRA are weaker and show no systematic increase with altitude, unlike the case with E-OBS.\n\n* Large inter-model spread—especially for winter cold extremes above 1500 m—highlights significant uncertainty in representing conditions critical for alpine risk assessments.\n\n* Lapse-rate correction might slightly reduce biases by ensuring no elevation differences exist between datasets, but it does not fundamentally change the shape of vertical bias profiles or address the underlying model or observational issues driving the biases.\n\n* The findings of this Alpine-focused assessment highlight the need for caution when interpreting future temperature projections from RCMs. A central concern is the potential for present-climate biases to persist or propagate into future warming scenarios—an aspect that remains largely underexplored.\n\n```\n\nattachment:0eeb016f-678f-44d4-b821-4d782590b4f2.png\n---\nwidth: 900px\nalt: Visual abstract \n---\nElevation profiles of biases relative to E-OBS. Differences compared to the E-OBS vertical profile are shown for each EURO-CORDEX model individually, the ensemble median, the spread (calculated as the standard deviation), and CERRA. Profiles are plotted separately for winter (DJF) and summer (JJA), and for the 5th, 50th, and 95th percentiles. The ensemble median is represented by a solid black line, while individual EURO-CORDEX models appear as grey dashed lines. The model spread, calculated as the standard deviation across models, is shown as a grey shaded area. CERRA is depicted in orange. No lapse-rate correction is applied in this case, and data are grouped by the elevation of each model or reference dataset.\n```", "text_with_prefix": "EQC Quality Assessment: \"CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q06 | Category: Climate_Projections\nSection: CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA > Quality assessment statement\n---\nThese are the key outcomes of this assessment\n\n* CORDEX simulations exhibit biases relative to both reference datasets (E-OBS and CERRA).\n\n* RCMs’ systematic cold biases above 1000–1500 m, reported in the literature, are evident when evaluated against E-OBS.\n\n* CORDEX biases relative to CERRA are weaker and show no systematic increase with altitude, unlike the case with E-OBS.\n\n* Large inter-model spread—especially for winter cold extremes above 1500 m—highlights significant uncertainty in representing conditions critical for alpine risk assessments.\n\n* Lapse-rate correction might slightly reduce biases by ensuring no elevation differences exist between datasets, but it does not fundamentally change the shape of vertical bias profiles or address the underlying model or observational issues driving the biases.\n\n* The findings of this Alpine-focused assessment highlight the need for caution when interpreting future temperature projections from RCMs. A central concern is the potential for present-climate biases to persist or propagate into future warming scenarios—an aspect that remains largely underexplored.\n\n```\n\nattachment:0eeb016f-678f-44d4-b821-4d782590b4f2.png\n---\nwidth: 900px\nalt: Visual abstract \n---\nElevation profiles of biases relative to E-OBS. Differences compared to the E-OBS vertical profile are shown for each EURO-CORDEX model individually, the ensemble median, the spread (calculated as the standard deviation), and CERRA. Profiles are plotted separately for winter (DJF) and summer (JJA), and for the 5th, 50th, and 95th percentiles. The ensemble median is represented by a solid black line, while individual EURO-CORDEX models appear as grey dashed lines. The model spread, calculated as the standard deviation across models, is shown as a grey shaded area. CERRA is depicted in orange. No lapse-rate correction is applied in this case, and data are grouped by the elevation of each model or reference dataset.\n```"} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q06__959b86497454", "report_id": "climate_projections-cordex-domains-single-levels_validation_q06", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q06", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA > Methodology", "title": "CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA", "chunk_index": 3, "token_count": 1071, "text_raw": "A subset of 39 models from [EURO-CORDEX](https://cds.climate.copernicus.eu/datasets/projections-cordex-domains-single-levels?tab=quality_assurance_tab) is compared to [E-OBSv30.0e](https://cds.climate.copernicus.eu/datasets/insitu-gridded-observations-europe?tab=overview) to assess the elevation dependency of temperature biases. Following the approach of [[3]](https://doi.org/10.1007/s00382-024-07376-y), we evaluate differences between the RCMs and E-OBS in representing the 5th, 50th, and 95th percentiles of temperature at various altitude ranges, for both summer and winter, across the Greater Alpine Region (GAR; 4–19°E, 43–49°N, 0–3500 m a.s.l.; [[5]](https://doi.org/10.1007/s00704-018-2506-5)[[6]](https://doi.org/10.1002/joc.1377)). The extent of the GAR domain can be seen in *Figure .2 - GAR* reproduced from Haslinger et al., 2019 [[5]](https://doi.org/10.1007/s00704-018-2506-5) under the [CC BY 4.0](https://creativecommons.org/licenses/by/4.0/).\n\nTo assess consistency across reference datasets, CERRA—a high-resolution (5.5 km) regional reanalysis produced using the HARMONIE-ALADIN limited-area numerical weather prediction and data assimilation system—is also compared to E-OBS and the RCMs.\n\nAll datasets are regridded to a 0.11° grid corresponding to a selected RCM (user-defined). There is no clear consensus in the literature on the optimal regridding approach: regridding to a finer resolution can be useful for certain applications but risks introducing artificial spatial detail not supported by the underlying data, whereas regridding to a coarser grid preserves the original physical consistency. In this work, we have chosen to regrid to the coarser resolution, following the approach in [[7]](https://doi.org/10.1002/joc.5249).\n\nWhile regridding enables consistent spatial comparison, it does not inherently align the elevations of different datasets, which may still differ due to variations in underlying orography. To address this, a lapse-rate correction is applied to adjust all datasets to the same target elevation. CERRA's elevation (regridded to 0.11°) is used as the target, as CERRA originally has the highest spatial resolution. A spatially and temporally uniform lapse rate of 0.0065 °C m⁻¹ is applied to all regridded datasets—including E-OBS and the RCMs—following a procedure similar to [[7]](https://doi.org/10.1002/joc.5249). Differences in temperatures before and after this correction are evaluated to determine whether the adjustment produces substantial changes.\n\nThe analysis and results follow the next outline:\n\n**[](climate_projections-cordex-domains-single-levels_validation_q06:section-1)**\n * [](climate_projections-cordex-domains-single-levels_validation_q06:section-1.1)\n * [](climate_projections-cordex-domains-single-levels_validation_q06:section-1.2)\n * [](climate_projections-cordex-domains-single-levels_validation_q06:section-1.3)\n * [](climate_projections-cordex-domains-single-levels_validation_q06:section-1.4)\n * [](climate_projections-cordex-domains-single-levels_validation_q06:section-1.5)\n * [](climate_projections-cordex-domains-single-levels_validation_q06:section-1.6)\n * [](climate_projections-cordex-domains-single-levels_validation_q06:section-1.7)\n\n**[](climate_projections-cordex-domains-single-levels_validation_q06:section-2)**\n * [](climate_projections-cordex-domains-single-levels_validation_q06:section-2.1)\n * [](climate_projections-cordex-domains-single-levels_validation_q06:section-2.2)\n * [](climate_projections-cordex-domains-single-levels_validation_q06:section-2.3)\n * [](climate_projections-cordex-domains-single-levels_validation_q06:section-2.4)\n * [](climate_projections-cordex-domains-single-levels_validation_q06:section-2.5)", "text_with_prefix": "EQC Quality Assessment: \"CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q06 | Category: Climate_Projections\nSection: CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA > Methodology\n---\nA subset of 39 models from [EURO-CORDEX](https://cds.climate.copernicus.eu/datasets/projections-cordex-domains-single-levels?tab=quality_assurance_tab) is compared to [E-OBSv30.0e](https://cds.climate.copernicus.eu/datasets/insitu-gridded-observations-europe?tab=overview) to assess the elevation dependency of temperature biases. Following the approach of [[3]](https://doi.org/10.1007/s00382-024-07376-y), we evaluate differences between the RCMs and E-OBS in representing the 5th, 50th, and 95th percentiles of temperature at various altitude ranges, for both summer and winter, across the Greater Alpine Region (GAR; 4–19°E, 43–49°N, 0–3500 m a.s.l.; [[5]](https://doi.org/10.1007/s00704-018-2506-5)[[6]](https://doi.org/10.1002/joc.1377)). The extent of the GAR domain can be seen in *Figure .2 - GAR* reproduced from Haslinger et al., 2019 [[5]](https://doi.org/10.1007/s00704-018-2506-5) under the [CC BY 4.0](https://creativecommons.org/licenses/by/4.0/).\n\nTo assess consistency across reference datasets, CERRA—a high-resolution (5.5 km) regional reanalysis produced using the HARMONIE-ALADIN limited-area numerical weather prediction and data assimilation system—is also compared to E-OBS and the RCMs.\n\nAll datasets are regridded to a 0.11° grid corresponding to a selected RCM (user-defined). There is no clear consensus in the literature on the optimal regridding approach: regridding to a finer resolution can be useful for certain applications but risks introducing artificial spatial detail not supported by the underlying data, whereas regridding to a coarser grid preserves the original physical consistency. In this work, we have chosen to regrid to the coarser resolution, following the approach in [[7]](https://doi.org/10.1002/joc.5249).\n\nWhile regridding enables consistent spatial comparison, it does not inherently align the elevations of different datasets, which may still differ due to variations in underlying orography. To address this, a lapse-rate correction is applied to adjust all datasets to the same target elevation. CERRA's elevation (regridded to 0.11°) is used as the target, as CERRA originally has the highest spatial resolution. A spatially and temporally uniform lapse rate of 0.0065 °C m⁻¹ is applied to all regridded datasets—including E-OBS and the RCMs—following a procedure similar to [[7]](https://doi.org/10.1002/joc.5249). Differences in temperatures before and after this correction are evaluated to determine whether the adjustment produces substantial changes.\n\nThe analysis and results follow the next outline:\n\n**[](climate_projections-cordex-domains-single-levels_validation_q06:section-1)**\n * [](climate_projections-cordex-domains-single-levels_validation_q06:section-1.1)\n * [](climate_projections-cordex-domains-single-levels_validation_q06:section-1.2)\n * [](climate_projections-cordex-domains-single-levels_validation_q06:section-1.3)\n * [](climate_projections-cordex-domains-single-levels_validation_q06:section-1.4)\n * [](climate_projections-cordex-domains-single-levels_validation_q06:section-1.5)\n * [](climate_projections-cordex-domains-single-levels_validation_q06:section-1.6)\n * [](climate_projections-cordex-domains-single-levels_validation_q06:section-1.7)\n\n**[](climate_projections-cordex-domains-single-levels_validation_q06:section-2)**\n * [](climate_projections-cordex-domains-single-levels_validation_q06:section-2.1)\n * [](climate_projections-cordex-domains-single-levels_validation_q06:section-2.2)\n * [](climate_projections-cordex-domains-single-levels_validation_q06:section-2.3)\n * [](climate_projections-cordex-domains-single-levels_validation_q06:section-2.4)\n * [](climate_projections-cordex-domains-single-levels_validation_q06:section-2.5)"} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q06__e70bc8f08a54", "report_id": "climate_projections-cordex-domains-single-levels_validation_q06", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q06", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA > Methodology", "title": "CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA", "chunk_index": 4, "token_count": 450, "text_raw": ".3)\n * [](climate_projections-cordex-domains-single-levels_validation_q06:section-2.4)\n * [](climate_projections-cordex-domains-single-levels_validation_q06:section-2.5)\n\n**[](climate_projections-cordex-domains-single-levels_validation_q06:section-3)**\n * [](climate_projections-cordex-domains-single-levels_validation_q06:section-3.1)\n * [](climate_projections-cordex-domains-single-levels_validation_q06:section-3.2)\n * [](climate_projections-cordex-domains-single-levels_validation_q06:section-3.3)\n * [](climate_projections-cordex-domains-single-levels_validation_q06:section-3.4)\n * [](climate_projections-cordex-domains-single-levels_validation_q06:section-3.5)\n * [](climate_projections-cordex-domains-single-levels_validation_q06:section-3.6)\n\nattachment:5fd0ab4f-7024-4560-8446-80e6664efd46.png\n---\nwidth: 900px\nalt: figure 2\n---\nMap of Central and Southern Europe. The broken line indicates the boundaries of the Greater Alpine Region; the solid line represents a generalized outline of the 1000 m a.s.l. isoline of the Alps which should help in locating the mountainous areas of the domain in the following figures. Image reproduced from *Figure 1* in Haslinger et al., 2019 [[5]](https://doi.org/10.1007/s00704-018-2506-5), under the [CC BY 4.0](https://creativecommons.org/licenses/by/4.0/)\n\n```", "text_with_prefix": "EQC Quality Assessment: \"CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q06 | Category: Climate_Projections\nSection: CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA > Methodology\n---\n.3)\n * [](climate_projections-cordex-domains-single-levels_validation_q06:section-2.4)\n * [](climate_projections-cordex-domains-single-levels_validation_q06:section-2.5)\n\n**[](climate_projections-cordex-domains-single-levels_validation_q06:section-3)**\n * [](climate_projections-cordex-domains-single-levels_validation_q06:section-3.1)\n * [](climate_projections-cordex-domains-single-levels_validation_q06:section-3.2)\n * [](climate_projections-cordex-domains-single-levels_validation_q06:section-3.3)\n * [](climate_projections-cordex-domains-single-levels_validation_q06:section-3.4)\n * [](climate_projections-cordex-domains-single-levels_validation_q06:section-3.5)\n * [](climate_projections-cordex-domains-single-levels_validation_q06:section-3.6)\n\nattachment:5fd0ab4f-7024-4560-8446-80e6664efd46.png\n---\nwidth: 900px\nalt: figure 2\n---\nMap of Central and Southern Europe. The broken line indicates the boundaries of the Greater Alpine Region; the solid line represents a generalized outline of the 1000 m a.s.l. isoline of the Alps which should help in locating the mountainous areas of the domain in the following figures. Image reproduced from *Figure 1* in Haslinger et al., 2019 [[5]](https://doi.org/10.1007/s00704-018-2506-5), under the [CC BY 4.0](https://creativecommons.org/licenses/by/4.0/)\n\n```"} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q06__bc61828f91ba", "report_id": "climate_projections-cordex-domains-single-levels_validation_q06", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q06", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA > Analysis and results > 1. Parameters, requests and functions definition > 1.2. Define Parameters", "title": "CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA", "chunk_index": 5, "token_count": 505, "text_raw": "In the \"Define Parameters\" section, various customisable options for the notebook are specified:\n\n- The start and end years of the historical period can be specified using the `year_start` and `year_stop` parameters. The default period, 1986–2005, is selected to maximise dataset availability while avoiding the need to include future projection data.\n- The `chunk` selection allows the user to define if dividing into chunks when downloading the data on their local machine. Although it does not significantly affect the analysis, it is recommended to keep the default value for optimal performance.\n- `variable` and `collection_id` are not customisable for this assessment and are set to 'temperture' and 'CORDEX'. Expert users can use this notebook as a guide to create similar analyses for other variables or model sets.\n- `area` defines the geographical domain over which the analysis will be performed. Note that this notebook is designed for the EURO-CORDEX domain, so the selected area must lie within its boundaries. The Greater Alpine Region (GAR; 4–19°E, 43–49°N, 0–3500 m a.s.l.; [[5]](https://doi.org/10.1007/s00704-018-2506-5)[[6]](https://doi.org/10.1002/joc.1377)) is choosen for this assessment.\n- `rcm_model_regrid` and `gcm_driven_model_regrid` specify the EURO-CORDEX regional climate model and the driving global climate model, respectively, whose grid will be used as the target for remapping. Note that both models must be included in the matrix of RCM-GCM combinations defined by the `models_cordex` dictionary in Section 1.3.\n\nChoose CORDEX or CMIP6\nArea to show (Alpine Area)\n\n(climate_projections-cordex-domains-single-levels_validation_q06:section-1.3)=", "text_with_prefix": "EQC Quality Assessment: \"CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q06 | Category: Climate_Projections\nSection: CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA > Analysis and results > 1. Parameters, requests and functions definition > 1.2. Define Parameters\n---\nIn the \"Define Parameters\" section, various customisable options for the notebook are specified:\n\n- The start and end years of the historical period can be specified using the `year_start` and `year_stop` parameters. The default period, 1986–2005, is selected to maximise dataset availability while avoiding the need to include future projection data.\n- The `chunk` selection allows the user to define if dividing into chunks when downloading the data on their local machine. Although it does not significantly affect the analysis, it is recommended to keep the default value for optimal performance.\n- `variable` and `collection_id` are not customisable for this assessment and are set to 'temperture' and 'CORDEX'. Expert users can use this notebook as a guide to create similar analyses for other variables or model sets.\n- `area` defines the geographical domain over which the analysis will be performed. Note that this notebook is designed for the EURO-CORDEX domain, so the selected area must lie within its boundaries. The Greater Alpine Region (GAR; 4–19°E, 43–49°N, 0–3500 m a.s.l.; [[5]](https://doi.org/10.1007/s00704-018-2506-5)[[6]](https://doi.org/10.1002/joc.1377)) is choosen for this assessment.\n- `rcm_model_regrid` and `gcm_driven_model_regrid` specify the EURO-CORDEX regional climate model and the driving global climate model, respectively, whose grid will be used as the target for remapping. Note that both models must be included in the matrix of RCM-GCM combinations defined by the `models_cordex` dictionary in Section 1.3.\n\nChoose CORDEX or CMIP6\nArea to show (Alpine Area)\n\n(climate_projections-cordex-domains-single-levels_validation_q06:section-1.3)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q06__fce047407f7f", "report_id": "climate_projections-cordex-domains-single-levels_validation_q06", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q06", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA > Analysis and results > 1. Parameters, requests and functions definition > 1.3. Define models", "title": "CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA", "chunk_index": 6, "token_count": 187, "text_raw": "The following climate analyses are conducted using the EURO-CORDEX models available at the time of download, limited to those that include the orography variable in the Climate Data Store ([CDS](https://cds.climate.copernicus.eu/datasets/projections-cordex-domains-single-levels?tab=overview)).\n\n(climate_projections-cordex-domains-single-levels_validation_q06:section-1.4)=", "text_with_prefix": "EQC Quality Assessment: \"CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q06 | Category: Climate_Projections\nSection: CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA > Analysis and results > 1. Parameters, requests and functions definition > 1.3. Define models\n---\nThe following climate analyses are conducted using the EURO-CORDEX models available at the time of download, limited to those that include the orography variable in the Climate Data Store ([CDS](https://cds.climate.copernicus.eu/datasets/projections-cordex-domains-single-levels?tab=overview)).\n\n(climate_projections-cordex-domains-single-levels_validation_q06:section-1.4)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q06__f60ed35e1833", "report_id": "climate_projections-cordex-domains-single-levels_validation_q06", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q06", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA > Analysis and results > 1. Parameters, requests and functions definition > 1.4. Define CERRA request", "title": "CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA", "chunk_index": 7, "token_count": 166, "text_raw": "Within this notebook, CERRA is also used to stress the sensitivity of results to the choice of reference dataset. In this section, we define the necessary parameters for the cds-api data request for CERRA.\n\n(climate_projections-cordex-domains-single-levels_validation_q06:section-1.5)=", "text_with_prefix": "EQC Quality Assessment: \"CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q06 | Category: Climate_Projections\nSection: CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA > Analysis and results > 1. Parameters, requests and functions definition > 1.4. Define CERRA request\n---\nWithin this notebook, CERRA is also used to stress the sensitivity of results to the choice of reference dataset. In this section, we define the necessary parameters for the cds-api data request for CERRA.\n\n(climate_projections-cordex-domains-single-levels_validation_q06:section-1.5)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q06__ddae4e6fb4d6", "report_id": "climate_projections-cordex-domains-single-levels_validation_q06", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q06", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA > Analysis and results > 1. Parameters, requests and functions definition > 1.5. Define E-OBS request", "title": "CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA", "chunk_index": 8, "token_count": 156, "text_raw": "Within this notebook, E-OBS serves as the reference product. In this section, we set the required parameters for the cds-api data-request of E-OBS.\n\n(climate_projections-cordex-domains-single-levels_validation_q06:section-1.6)=", "text_with_prefix": "EQC Quality Assessment: \"CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q06 | Category: Climate_Projections\nSection: CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA > Analysis and results > 1. Parameters, requests and functions definition > 1.5. Define E-OBS request\n---\nWithin this notebook, E-OBS serves as the reference product. In this section, we set the required parameters for the cds-api data-request of E-OBS.\n\n(climate_projections-cordex-domains-single-levels_validation_q06:section-1.6)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q06__4e3a27841995", "report_id": "climate_projections-cordex-domains-single-levels_validation_q06", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q06", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA > Analysis and results > 1. Parameters, requests and functions definition > 1.6. Define model requests", "title": "CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA", "chunk_index": 9, "token_count": 142, "text_raw": "In this section we set the required parameters for the cds-api data-request of EURO-CORDEX models.\n\n(climate_projections-cordex-domains-single-levels_validation_q06:section-1.7)=", "text_with_prefix": "EQC Quality Assessment: \"CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q06 | Category: Climate_Projections\nSection: CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA > Analysis and results > 1. Parameters, requests and functions definition > 1.6. Define model requests\n---\nIn this section we set the required parameters for the cds-api data-request of EURO-CORDEX models.\n\n(climate_projections-cordex-domains-single-levels_validation_q06:section-1.7)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q06__7950435489ee", "report_id": "climate_projections-cordex-domains-single-levels_validation_q06", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q06", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA > Analysis and results > 1. Parameters, requests and functions definition > 1.7. Functions to cache", "title": "CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA", "chunk_index": 10, "token_count": 346, "text_raw": "In this section, functions that will be executed in the caching phase are defined. Caching is the process of storing copies of files in a temporary storage location, so that they can be accessed more quickly. This process also checks if the user has already downloaded a file, avoiding redundant downloads.\n\nDescription of the main functions:\n\n- **`compute_percentiles`**: This function resamples data to daily frequency if needed (e.g., for hourly CERRA data), performs regridding to the target grid, selects the subregion of interest, and calculates the 5th, 50th, and 95th percentiles for summer and winter.\n- **`orog_regrid`**: This function regrids the orography data from the models, E-OBS and CERRA onto the selected RCM grid (defined by the `model_regrid` parameter), enabling a subsequent point-by-point lapse-rate correction.\n\nEnsure time is datetime type\nFilter by years\nFilter by season\nSelect a subdomain\nOriginal bounds for conservative interpolation\nCompute percentiles\nSelect a subdomain\nOriginal bounds for conservative interpolation\n\n(climate_projections-cordex-domains-single-levels_validation_q06:section-2)=", "text_with_prefix": "EQC Quality Assessment: \"CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q06 | Category: Climate_Projections\nSection: CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA > Analysis and results > 1. Parameters, requests and functions definition > 1.7. Functions to cache\n---\nIn this section, functions that will be executed in the caching phase are defined. Caching is the process of storing copies of files in a temporary storage location, so that they can be accessed more quickly. This process also checks if the user has already downloaded a file, avoiding redundant downloads.\n\nDescription of the main functions:\n\n- **`compute_percentiles`**: This function resamples data to daily frequency if needed (e.g., for hourly CERRA data), performs regridding to the target grid, selects the subregion of interest, and calculates the 5th, 50th, and 95th percentiles for summer and winter.\n- **`orog_regrid`**: This function regrids the orography data from the models, E-OBS and CERRA onto the selected RCM grid (defined by the `model_regrid` parameter), enabling a subsequent point-by-point lapse-rate correction.\n\nEnsure time is datetime type\nFilter by years\nFilter by season\nSelect a subdomain\nOriginal bounds for conservative interpolation\nCompute percentiles\nSelect a subdomain\nOriginal bounds for conservative interpolation\n\n(climate_projections-cordex-domains-single-levels_validation_q06:section-2)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q06__b4d623981b98", "report_id": "climate_projections-cordex-domains-single-levels_validation_q06", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q06", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA > Analysis and results > 2. Downloading and processing > 2.1. Download and transform the regridding model", "title": "CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA", "chunk_index": 11, "token_count": 382, "text_raw": "In this section, the `download.download_and_transform` function from the `c3s_eqc_automatic_quality_control` package is employed to download daily data from the selected EURO-CORDEX regridding model, select the subregion of interest, and calculate the 5th, 50th, and 95th percentiles for summer and winter over the period 1986–2005, caching the results to avoid redundant downloads and processing.\n\nThe regridding model here refers to the model whose grid is used as the target grid for remapping the other models, as well as CERRA5 and E-OBS. This ensures all datasets share a common grid, facilitating direct comparison at each grid cell. Within this notebook, the regridding model is set to `\"ichec_ec_earth_smhi_rca4\"` but it can be changed by modifying the `gcm_driven_model_regrid` and `rcm_model_regrid` parameters in Section 1.2. Note that both models must be included in the matrix of RCM-GCM combinations defined by the `models_cordex` dictionary in Section 1.3. It is important to highlight that the choice of the target grid can impact the analysis depending on the specific application.\n\n(climate_projections-cordex-domains-single-levels_validation_q06:section-2.2)=", "text_with_prefix": "EQC Quality Assessment: \"CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q06 | Category: Climate_Projections\nSection: CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA > Analysis and results > 2. Downloading and processing > 2.1. Download and transform the regridding model\n---\nIn this section, the `download.download_and_transform` function from the `c3s_eqc_automatic_quality_control` package is employed to download daily data from the selected EURO-CORDEX regridding model, select the subregion of interest, and calculate the 5th, 50th, and 95th percentiles for summer and winter over the period 1986–2005, caching the results to avoid redundant downloads and processing.\n\nThe regridding model here refers to the model whose grid is used as the target grid for remapping the other models, as well as CERRA5 and E-OBS. This ensures all datasets share a common grid, facilitating direct comparison at each grid cell. Within this notebook, the regridding model is set to `\"ichec_ec_earth_smhi_rca4\"` but it can be changed by modifying the `gcm_driven_model_regrid` and `rcm_model_regrid` parameters in Section 1.2. Note that both models must be included in the matrix of RCM-GCM combinations defined by the `models_cordex` dictionary in Section 1.3. It is important to highlight that the choice of the target grid can impact the analysis depending on the specific application.\n\n(climate_projections-cordex-domains-single-levels_validation_q06:section-2.2)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q06__6e3aebf4dde2", "report_id": "climate_projections-cordex-domains-single-levels_validation_q06", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q06", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA > Analysis and results > 2. Downloading and processing > 2.2. Download and transform CERRA", "title": "CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA", "chunk_index": 12, "token_count": 259, "text_raw": "In this section, the `download.download_and_transform` function from the `c3s_eqc_automatic_quality_control` package is used to download sub-daily data from CERRA, resample it to daily frequency, regrid to the target grid, select the subregion of interest, and compute the 5th, 50th, and 95th percentiles for summer and winter over the period 1986–2005. The results are cached to avoid redundant downloads and processing. Elevation data is also retrieved and regridded to the target grid (defined by the model_regrid parameter), as it is required to perform the lapse-rate correction.\n\n(climate_projections-cordex-domains-single-levels_validation_q06:section-2.3)=", "text_with_prefix": "EQC Quality Assessment: \"CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q06 | Category: Climate_Projections\nSection: CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA > Analysis and results > 2. Downloading and processing > 2.2. Download and transform CERRA\n---\nIn this section, the `download.download_and_transform` function from the `c3s_eqc_automatic_quality_control` package is used to download sub-daily data from CERRA, resample it to daily frequency, regrid to the target grid, select the subregion of interest, and compute the 5th, 50th, and 95th percentiles for summer and winter over the period 1986–2005. The results are cached to avoid redundant downloads and processing. Elevation data is also retrieved and regridded to the target grid (defined by the model_regrid parameter), as it is required to perform the lapse-rate correction.\n\n(climate_projections-cordex-domains-single-levels_validation_q06:section-2.3)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q06__f89e954569d4", "report_id": "climate_projections-cordex-domains-single-levels_validation_q06", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q06", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA > Analysis and results > 2. Downloading and processing > 2.3. Download and transform E-OBS", "title": "CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA", "chunk_index": 13, "token_count": 250, "text_raw": "In this section, the `download.download_and_transform` function from the `c3s_eqc_automatic_quality_control` package is used to download daily data from E-OBS, regrid to the target grid, select the subregion of interest, and compute the 5th, 50th, and 95th percentiles for summer and winter over the period 1986–2005. The results are cached to avoid redundant downloads and processing. Elevation data is also retrieved and regridded to the target grid (defined by the model_regrid parameter), as it is required to perform the lapse-rate correction.\n\n(climate_projections-cordex-domains-single-levels_validation_q06:section-2.4)=", "text_with_prefix": "EQC Quality Assessment: \"CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q06 | Category: Climate_Projections\nSection: CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA > Analysis and results > 2. Downloading and processing > 2.3. Download and transform E-OBS\n---\nIn this section, the `download.download_and_transform` function from the `c3s_eqc_automatic_quality_control` package is used to download daily data from E-OBS, regrid to the target grid, select the subregion of interest, and compute the 5th, 50th, and 95th percentiles for summer and winter over the period 1986–2005. The results are cached to avoid redundant downloads and processing. Elevation data is also retrieved and regridded to the target grid (defined by the model_regrid parameter), as it is required to perform the lapse-rate correction.\n\n(climate_projections-cordex-domains-single-levels_validation_q06:section-2.4)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q06__38d3ad15d134", "report_id": "climate_projections-cordex-domains-single-levels_validation_q06", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q06", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA > Analysis and results > 2. Downloading and processing > 2.4. Download and transform models", "title": "CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA", "chunk_index": 14, "token_count": 935, "text_raw": "In this section, the `download.download_and_transform` function from the `c3s_eqc_automatic_quality_control` package is used to download daily data from the selected EURO-CORDEX models, regrid to the target grid, select the subregion of interest, and compute the 5th, 50th, and 95th percentiles for summer and winter over the period 1986–2005. The results are cached to avoid redundant downloads and processing. Elevation data is also retrieved and regridded to the target grid (defined by the model_regrid parameter), as it is required to perform the lapse-rate correction.\n\n```text\nmodel='cccma_canesm2_gerics_remo2015'\nmodel='cnrm_cerfacs_cm5_clmcom_eth_cosmo_crclim'\nmodel='cnrm_cerfacs_cm5_cnrm_aladin63'\nmodel='cnrm_cerfacs_cm5_dmi_hirham5'\nmodel='cnrm_cerfacs_cm5_gerics_remo2015'\nmodel='cnrm_cerfacs_cm5_knmi_racmo22e'\nmodel='cnrm_cerfacs_cm5_mohc_hadrem3_ga7_05'\nmodel='ichec_ec_earth_clmcom_eth_cosmo_crclim'\nmodel='ichec_ec_earth_dmi_hirham5'\nmodel='ichec_ec_earth_knmi_racmo22e'\nmodel='ichec_ec_earth_smhi_rca4'\nmodel='ipsl_cm5a_mr_dmi_hirham5'\nmodel='ipsl_cm5a_mr_gerics_remo2015'\nmodel='ipsl_cm5a_mr_knmi_racmo22e'\nmodel='ipsl_cm5a_mr_smhi_rca4'\nmodel='miroc_miroc5_clmcom_clm_cclm4_8_17'\nmodel='miroc_miroc5_gerics_remo2015'\nmodel='mohc_hadgem2_es_clmcom_clm_cclm4_8_17'\nmodel='mohc_hadgem2_es_cnrm_aladin63'\nmodel='mohc_hadgem2_es_dmi_hirham5'\nmodel='mohc_hadgem2_es_gerics_remo2015'\nmodel='mohc_hadgem2_es_knmi_racmo22e'\nmodel='mohc_hadgem2_es_mohc_hadrem3_ga7_05'\nmodel='mohc_hadgem2_es_smhi_rca4'\nmodel='mpi_m_mpi_esm_lr_clmcom_clm_cclm4_8_17'\nmodel='mpi_m_mpi_esm_lr_clmcom_eth_cosmo_crclim'\nmodel='mpi_m_mpi_esm_lr_cnrm_aladin63'\nmodel='mpi_m_mpi_esm_lr_dmi_hirham5'\nmodel='mpi_m_mpi_esm_lr_knmi_racmo22e'\nmodel='mpi_m_mpi_esm_lr_mohc_hadrem3_ga7_05'\nmodel='mpi_m_mpi_esm_lr_mpi_csc_remo2009'\nmodel='mpi_m_mpi_esm_lr_smhi_rca4'\nmodel='mpi_m_mpi_esm_lr_uhoh_wrf361h'\nmodel='ncc_noresm1_m_cnrm_aladin63'\nmodel='ncc_noresm1_m_dmi_hirham5'\nmodel='ncc_noresm1_m_gerics_remo2015'\nmodel='ncc_noresm1_m_knmi_racmo22e'\nmodel='ncc_noresm1_m_mohc_hadrem3_ga7_05'\nmodel='ncc_noresm1_m_smhi_rca4'\n```\n\n(climate_projections-cordex-domains-single-levels_validation_q06:section-2.5)=", "text_with_prefix": "EQC Quality Assessment: \"CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q06 | Category: Climate_Projections\nSection: CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA > Analysis and results > 2. Downloading and processing > 2.4. Download and transform models\n---\nIn this section, the `download.download_and_transform` function from the `c3s_eqc_automatic_quality_control` package is used to download daily data from the selected EURO-CORDEX models, regrid to the target grid, select the subregion of interest, and compute the 5th, 50th, and 95th percentiles for summer and winter over the period 1986–2005. The results are cached to avoid redundant downloads and processing. Elevation data is also retrieved and regridded to the target grid (defined by the model_regrid parameter), as it is required to perform the lapse-rate correction.\n\n```text\nmodel='cccma_canesm2_gerics_remo2015'\nmodel='cnrm_cerfacs_cm5_clmcom_eth_cosmo_crclim'\nmodel='cnrm_cerfacs_cm5_cnrm_aladin63'\nmodel='cnrm_cerfacs_cm5_dmi_hirham5'\nmodel='cnrm_cerfacs_cm5_gerics_remo2015'\nmodel='cnrm_cerfacs_cm5_knmi_racmo22e'\nmodel='cnrm_cerfacs_cm5_mohc_hadrem3_ga7_05'\nmodel='ichec_ec_earth_clmcom_eth_cosmo_crclim'\nmodel='ichec_ec_earth_dmi_hirham5'\nmodel='ichec_ec_earth_knmi_racmo22e'\nmodel='ichec_ec_earth_smhi_rca4'\nmodel='ipsl_cm5a_mr_dmi_hirham5'\nmodel='ipsl_cm5a_mr_gerics_remo2015'\nmodel='ipsl_cm5a_mr_knmi_racmo22e'\nmodel='ipsl_cm5a_mr_smhi_rca4'\nmodel='miroc_miroc5_clmcom_clm_cclm4_8_17'\nmodel='miroc_miroc5_gerics_remo2015'\nmodel='mohc_hadgem2_es_clmcom_clm_cclm4_8_17'\nmodel='mohc_hadgem2_es_cnrm_aladin63'\nmodel='mohc_hadgem2_es_dmi_hirham5'\nmodel='mohc_hadgem2_es_gerics_remo2015'\nmodel='mohc_hadgem2_es_knmi_racmo22e'\nmodel='mohc_hadgem2_es_mohc_hadrem3_ga7_05'\nmodel='mohc_hadgem2_es_smhi_rca4'\nmodel='mpi_m_mpi_esm_lr_clmcom_clm_cclm4_8_17'\nmodel='mpi_m_mpi_esm_lr_clmcom_eth_cosmo_crclim'\nmodel='mpi_m_mpi_esm_lr_cnrm_aladin63'\nmodel='mpi_m_mpi_esm_lr_dmi_hirham5'\nmodel='mpi_m_mpi_esm_lr_knmi_racmo22e'\nmodel='mpi_m_mpi_esm_lr_mohc_hadrem3_ga7_05'\nmodel='mpi_m_mpi_esm_lr_mpi_csc_remo2009'\nmodel='mpi_m_mpi_esm_lr_smhi_rca4'\nmodel='mpi_m_mpi_esm_lr_uhoh_wrf361h'\nmodel='ncc_noresm1_m_cnrm_aladin63'\nmodel='ncc_noresm1_m_dmi_hirham5'\nmodel='ncc_noresm1_m_gerics_remo2015'\nmodel='ncc_noresm1_m_knmi_racmo22e'\nmodel='ncc_noresm1_m_mohc_hadrem3_ga7_05'\nmodel='ncc_noresm1_m_smhi_rca4'\n```\n\n(climate_projections-cordex-domains-single-levels_validation_q06:section-2.5)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q06__a9ad9c38b39f", "report_id": "climate_projections-cordex-domains-single-levels_validation_q06", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q06", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA > Analysis and results > 2. Downloading and processing > 2.5. Apply land-sea mask", "title": "CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA", "chunk_index": 15, "token_count": 194, "text_raw": "This section applies a land–sea mask to all models, as well as to CERRA and E-OBS, to ensure consistency in the land-only analysis.\n\nRename elevation variable in ds_orog_eobs\n\n```text\n100%|██████████| 1/1 [00:00<00:00, 20.27it/s]\n```\n\n(climate_projections-cordex-domains-single-levels_validation_q06:section-3)=", "text_with_prefix": "EQC Quality Assessment: \"CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q06 | Category: Climate_Projections\nSection: CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA > Analysis and results > 2. Downloading and processing > 2.5. Apply land-sea mask\n---\nThis section applies a land–sea mask to all models, as well as to CERRA and E-OBS, to ensure consistency in the land-only analysis.\n\nRename elevation variable in ds_orog_eobs\n\n```text\n100%|██████████| 1/1 [00:00<00:00, 20.27it/s]\n```\n\n(climate_projections-cordex-domains-single-levels_validation_q06:section-3)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q06__37f48f0e0414", "report_id": "climate_projections-cordex-domains-single-levels_validation_q06", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q06", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA > Analysis and results > 3. Plot and describe results", "title": "CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA", "chunk_index": 16, "token_count": 323, "text_raw": "This section will display the following results:\n\n- **Elevation profiles of biases relative to E-OBS**: Differences compared to the E-OBS vertical profile are shown for each EURO-CORDEX model individually, the ensemble median, the spread (calculated as the standard deviation), and CERRA. Profiles are plotted separately for winter (DJF) and summer (JJA), and for the 5th, 50th, and 95th percentiles.\n\n- **Elevation profiles of biases after applying lapse-rate correction**: As above, this includes each EURO-CORDEX model, the ensemble median, the spread, and CERRA, for both seasons and selected percentiles, but with all datasets corrected to CERRA’s elevation (regridded to 0.11°) using a uniform lapse rate.\n\n- **Elevation profiles of the impact of lapse-rate correction**: This shows the difference between the corrected and uncorrected profiles for E-OBS, individual EURO-CORDEX models, the ensemble median, and the spread.\n\n(climate_projections-cordex-domains-single-levels_validation_q06:section-3.1)=", "text_with_prefix": "EQC Quality Assessment: \"CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q06 | Category: Climate_Projections\nSection: CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA > Analysis and results > 3. Plot and describe results\n---\nThis section will display the following results:\n\n- **Elevation profiles of biases relative to E-OBS**: Differences compared to the E-OBS vertical profile are shown for each EURO-CORDEX model individually, the ensemble median, the spread (calculated as the standard deviation), and CERRA. Profiles are plotted separately for winter (DJF) and summer (JJA), and for the 5th, 50th, and 95th percentiles.\n\n- **Elevation profiles of biases after applying lapse-rate correction**: As above, this includes each EURO-CORDEX model, the ensemble median, the spread, and CERRA, for both seasons and selected percentiles, but with all datasets corrected to CERRA’s elevation (regridded to 0.11°) using a uniform lapse rate.\n\n- **Elevation profiles of the impact of lapse-rate correction**: This shows the difference between the corrected and uncorrected profiles for E-OBS, individual EURO-CORDEX models, the ensemble median, and the spread.\n\n(climate_projections-cordex-domains-single-levels_validation_q06:section-3.1)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q06__01e9f44326b6", "report_id": "climate_projections-cordex-domains-single-levels_validation_q06", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q06", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA > Analysis and results > 3. Plot and describe results > 3.1. Define plotting functions", "title": "CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA", "chunk_index": 17, "token_count": 536, "text_raw": "The functions **`compute_elevation_profile`**, **`plot_differences_vs_eobs`**, **`plot_lapse_rate_impact`** and **`apply_lapse_rate_correction`** are used to generate and visualise the elevation-dependent biases of temperature. Specifically:\n\n- **`compute_elevation_profile`** calculates the vertical (elevation-based) profiles of the 5th, 50th, and 95th percentiles for each dataset (EURO-CORDEX models, CERRA, and E-OBS), separately for winter (DJF) and summer (JJA).\n\n- **`apply_lapse_rate_correction`** adjusts temperatures to account for differences in elevation by applying a uniform lapse rate (−0.0065 °C m⁻¹), bringing the regridded models and E-OBS to the same elevation as CERRA (regridded to 0.11°).\n\n- **`plot_differences_vs_eobs`** plots the differences (biases) between the elevation profiles of the EURO-CORDEX models and CERRA, relative to that of E-OBS, for each percentile and season.\n\n- **`plot_lapse_rate_impact`** highlights the change introduced by the lapse-rate correction, showing the difference between corrected and uncorrected profiles for each dataset.\n\nIn the plots, the ensemble median is represented by a solid black line, while individual EURO-CORDEX models appear as grey dashed lines. The model spread, calculated as the standard deviation across models, is shown as a grey shaded area. CERRA is depicted in orange. Finally, E-OBS is shown in blue, but only in the plots that assess the impact of the lapse-rate correction.\n\nCompute E-OBS and CERRA profiles\nStore model differences\nConvert to DataFrame for ensemble stats\nPlot ensemble median\nPlot ensemble spread\nAdd gridlines\nShared legend\n--- Compute difference: corrected - original ---\nPlot individual model differences\nEnsemble statistics\nPlot ensemble median + spread\n\n(climate_projections-cordex-domains-single-levels_validation_q06:section-3.2)=", "text_with_prefix": "EQC Quality Assessment: \"CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q06 | Category: Climate_Projections\nSection: CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA > Analysis and results > 3. Plot and describe results > 3.1. Define plotting functions\n---\nThe functions **`compute_elevation_profile`**, **`plot_differences_vs_eobs`**, **`plot_lapse_rate_impact`** and **`apply_lapse_rate_correction`** are used to generate and visualise the elevation-dependent biases of temperature. Specifically:\n\n- **`compute_elevation_profile`** calculates the vertical (elevation-based) profiles of the 5th, 50th, and 95th percentiles for each dataset (EURO-CORDEX models, CERRA, and E-OBS), separately for winter (DJF) and summer (JJA).\n\n- **`apply_lapse_rate_correction`** adjusts temperatures to account for differences in elevation by applying a uniform lapse rate (−0.0065 °C m⁻¹), bringing the regridded models and E-OBS to the same elevation as CERRA (regridded to 0.11°).\n\n- **`plot_differences_vs_eobs`** plots the differences (biases) between the elevation profiles of the EURO-CORDEX models and CERRA, relative to that of E-OBS, for each percentile and season.\n\n- **`plot_lapse_rate_impact`** highlights the change introduced by the lapse-rate correction, showing the difference between corrected and uncorrected profiles for each dataset.\n\nIn the plots, the ensemble median is represented by a solid black line, while individual EURO-CORDEX models appear as grey dashed lines. The model spread, calculated as the standard deviation across models, is shown as a grey shaded area. CERRA is depicted in orange. Finally, E-OBS is shown in blue, but only in the plots that assess the impact of the lapse-rate correction.\n\nCompute E-OBS and CERRA profiles\nStore model differences\nConvert to DataFrame for ensemble stats\nPlot ensemble median\nPlot ensemble spread\nAdd gridlines\nShared legend\n--- Compute difference: corrected - original ---\nPlot individual model differences\nEnsemble statistics\nPlot ensemble median + spread\n\n(climate_projections-cordex-domains-single-levels_validation_q06:section-3.2)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q06__2e8999ee310b", "report_id": "climate_projections-cordex-domains-single-levels_validation_q06", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q06", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA > Analysis and results > 3. Plot and describe results > 3.2.Elevation profiles of biases relative to E-OBS", "title": "CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA", "chunk_index": 18, "token_count": 412, "text_raw": "Differences compared to the E-OBS vertical profile are shown for each EURO-CORDEX model individually, the ensemble median, the spread (calculated as the standard deviation), and CERRA. Profiles are plotted separately for winter (DJF) and summer (JJA), and for the 5th, 50th, and 95th percentiles.\n\nNote that no lapse-rate correction is applied in this case, and data are grouped by the elevation of each model or reference dataset.\n\n
\n
\n

Fig 3. \n Elevation profiles of biases relative to E-OBS. Differences compared to the E-OBS vertical profile are shown for each EURO-CORDEX model individually, the ensemble median, the spread (calculated as the standard deviation), and CERRA. Profiles are plotted separately for winter (DJF) and summer (JJA), and for the 5th, 50th, and 95th percentiles. The ensemble median is represented by a solid black line, while individual EURO-CORDEX models appear as grey dashed lines. The model spread, calculated as the standard deviation across models, is shown as a grey shaded area. CERRA is depicted in orange. No lapse-rate correction is applied in this case, and data are grouped by the elevation of each model or reference dataset.\n\n

\n\n(climate_projections-cordex-domains-single-levels_validation_q06:section-3.3)=", "text_with_prefix": "EQC Quality Assessment: \"CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q06 | Category: Climate_Projections\nSection: CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA > Analysis and results > 3. Plot and describe results > 3.2.Elevation profiles of biases relative to E-OBS\n---\nDifferences compared to the E-OBS vertical profile are shown for each EURO-CORDEX model individually, the ensemble median, the spread (calculated as the standard deviation), and CERRA. Profiles are plotted separately for winter (DJF) and summer (JJA), and for the 5th, 50th, and 95th percentiles.\n\nNote that no lapse-rate correction is applied in this case, and data are grouped by the elevation of each model or reference dataset.\n\n
\n
\n

Fig 3. \n Elevation profiles of biases relative to E-OBS. Differences compared to the E-OBS vertical profile are shown for each EURO-CORDEX model individually, the ensemble median, the spread (calculated as the standard deviation), and CERRA. Profiles are plotted separately for winter (DJF) and summer (JJA), and for the 5th, 50th, and 95th percentiles. The ensemble median is represented by a solid black line, while individual EURO-CORDEX models appear as grey dashed lines. The model spread, calculated as the standard deviation across models, is shown as a grey shaded area. CERRA is depicted in orange. No lapse-rate correction is applied in this case, and data are grouped by the elevation of each model or reference dataset.\n\n

\n\n(climate_projections-cordex-domains-single-levels_validation_q06:section-3.3)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q06__777139f02a00", "report_id": "climate_projections-cordex-domains-single-levels_validation_q06", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q06", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA > Analysis and results > 3. Plot and describe results > 3.3. Elevation profiles of biases after applying lapse-rate correction", "title": "CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA", "chunk_index": 19, "token_count": 634, "text_raw": "Differences compared to the E-OBS vertical profile are shown for each EURO-CORDEX model individually, the ensemble median, the spread (calculated as the standard deviation), and CERRA. Profiles are plotted separately for winter (DJF) and summer (JJA), and for the 5th, 50th, and 95th percentiles.\n\nNote that in this case, each dataset is adjusted to a common target elevation—the CERRA orography regridded to 0.11°—using a uniform lapse-rate correction (−0.0065 °C m⁻¹) before computing the differences. For example, if a grid point in a model is 200 m lower than the corresponding point in the target CERRA, the temperature at that point is corrected by 200 m × −0.0065 °C m⁻¹ = −1.3 °C.\n\nApplying lapse-rate correction to align model elevations with CERRA's elevation\nApplying lapse-rate correction to align E-OBS elevation with CERRA's elevation\n\n
\n
\n

Fig 4. \n Elevation profiles of biases after applying lapse-rate correction. Differences compared to the E-OBS vertical profile are shown for each EURO-CORDEX model individually, the ensemble median, the spread (calculated as the standard deviation), and CERRA. Profiles are plotted separately for winter (DJF) and summer (JJA), and for the 5th, 50th, and 95th percentiles. The ensemble median is represented by a solid black line, while individual EURO-CORDEX models appear as grey dashed lines. The model spread, calculated as the standard deviation across models, is shown as a grey shaded area. CERRA is depicted in orange. Each dataset is adjusted to a common target elevation—the CERRA orography regridded to 0.11°—using a uniform lapse-rate correction (−0.0065 °C m⁻¹) before computing the differences. For example, if a grid point in a model is 200 m lower than the corresponding point in the target CERRA, the temperature at that point is corrected by 200 m × −0.0065 °C m⁻¹ = −1.3 °C.\n\n

\n\n(climate_projections-cordex-domains-single-levels_validation_q06:section-3.4)=", "text_with_prefix": "EQC Quality Assessment: \"CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q06 | Category: Climate_Projections\nSection: CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA > Analysis and results > 3. Plot and describe results > 3.3. Elevation profiles of biases after applying lapse-rate correction\n---\nDifferences compared to the E-OBS vertical profile are shown for each EURO-CORDEX model individually, the ensemble median, the spread (calculated as the standard deviation), and CERRA. Profiles are plotted separately for winter (DJF) and summer (JJA), and for the 5th, 50th, and 95th percentiles.\n\nNote that in this case, each dataset is adjusted to a common target elevation—the CERRA orography regridded to 0.11°—using a uniform lapse-rate correction (−0.0065 °C m⁻¹) before computing the differences. For example, if a grid point in a model is 200 m lower than the corresponding point in the target CERRA, the temperature at that point is corrected by 200 m × −0.0065 °C m⁻¹ = −1.3 °C.\n\nApplying lapse-rate correction to align model elevations with CERRA's elevation\nApplying lapse-rate correction to align E-OBS elevation with CERRA's elevation\n\n
\n
\n

Fig 4. \n Elevation profiles of biases after applying lapse-rate correction. Differences compared to the E-OBS vertical profile are shown for each EURO-CORDEX model individually, the ensemble median, the spread (calculated as the standard deviation), and CERRA. Profiles are plotted separately for winter (DJF) and summer (JJA), and for the 5th, 50th, and 95th percentiles. The ensemble median is represented by a solid black line, while individual EURO-CORDEX models appear as grey dashed lines. The model spread, calculated as the standard deviation across models, is shown as a grey shaded area. CERRA is depicted in orange. Each dataset is adjusted to a common target elevation—the CERRA orography regridded to 0.11°—using a uniform lapse-rate correction (−0.0065 °C m⁻¹) before computing the differences. For example, if a grid point in a model is 200 m lower than the corresponding point in the target CERRA, the temperature at that point is corrected by 200 m × −0.0065 °C m⁻¹ = −1.3 °C.\n\n

\n\n(climate_projections-cordex-domains-single-levels_validation_q06:section-3.4)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q06__fd324f2ffb84", "report_id": "climate_projections-cordex-domains-single-levels_validation_q06", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q06", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA > Analysis and results > 3. Plot and describe results > 3.4. Elevation profiles of the impact of lapse-rate correction", "title": "CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA", "chunk_index": 20, "token_count": 523, "text_raw": "This section highlights the change introduced by the lapse-rate correction. The temperature at each grid point of every model has been previously adjusted using a uniform lapse rate of −0.0065 °C m⁻¹ to align all datasets to the same target elevation, defined by the regridded CERRA orography (0.11° resolution). For example, if a grid point in a model is 200 m lower than the corresponding point in the target CERRA, the temperature at that point is corrected by 200 m × −0.0065 °C m⁻¹ = −1.3 °C. The plot below shows the vertical profiles of these differences, illustrating how the lapse-rate correction affects temperatures across elevation bands. In other words, it shows the difference between assuming that all models and reference datasets share the same elevation after regridding (as sometimes assumed in the literature) and applying a correction to account for residual elevation differences.\n\n
\n
\n

Fig 5. \n Elevation profiles of the impact of lapse-rate correction. The plot shows differences between lapse-rate corrected and uncorrected temperature profiles for E-OBS and EURO-CORDEX models. The correction adjusts temperatures to a common target elevation (that of regridded CERRA at 0.11°) using a uniform lapse rate (−0.0065 °C m⁻¹), ensuring consistent comparison across datasets. Negative values indicate that regridded model or E-OBS elevations are lower than CERRA. E-OBS is shown as a solid blue line, the ensemble median as a solid black line, individual models as grey dashed lines and the inter-model spread (standard deviation) as a grey shaded area. CERRA is not displayed since it defines the elevation target.\n\n

\n\n(climate_projections-cordex-domains-single-levels_validation_q06:section-3.5)=", "text_with_prefix": "EQC Quality Assessment: \"CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q06 | Category: Climate_Projections\nSection: CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA > Analysis and results > 3. Plot and describe results > 3.4. Elevation profiles of the impact of lapse-rate correction\n---\nThis section highlights the change introduced by the lapse-rate correction. The temperature at each grid point of every model has been previously adjusted using a uniform lapse rate of −0.0065 °C m⁻¹ to align all datasets to the same target elevation, defined by the regridded CERRA orography (0.11° resolution). For example, if a grid point in a model is 200 m lower than the corresponding point in the target CERRA, the temperature at that point is corrected by 200 m × −0.0065 °C m⁻¹ = −1.3 °C. The plot below shows the vertical profiles of these differences, illustrating how the lapse-rate correction affects temperatures across elevation bands. In other words, it shows the difference between assuming that all models and reference datasets share the same elevation after regridding (as sometimes assumed in the literature) and applying a correction to account for residual elevation differences.\n\n
\n
\n

Fig 5. \n Elevation profiles of the impact of lapse-rate correction. The plot shows differences between lapse-rate corrected and uncorrected temperature profiles for E-OBS and EURO-CORDEX models. The correction adjusts temperatures to a common target elevation (that of regridded CERRA at 0.11°) using a uniform lapse rate (−0.0065 °C m⁻¹), ensuring consistent comparison across datasets. Negative values indicate that regridded model or E-OBS elevations are lower than CERRA. E-OBS is shown as a solid blue line, the ensemble median as a solid black line, individual models as grey dashed lines and the inter-model spread (standard deviation) as a grey shaded area. CERRA is not displayed since it defines the elevation target.\n\n

\n\n(climate_projections-cordex-domains-single-levels_validation_q06:section-3.5)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q06__3ca994f28292", "report_id": "climate_projections-cordex-domains-single-levels_validation_q06", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q06", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA > Analysis and results > 3. Plot and describe results > 3.5. Results summary", "title": "CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA", "chunk_index": 21, "token_count": 455, "text_raw": "- Regional Climate Models (RCMs) show biases relative to both reference datasets (E-OBS and CERRA).\n\n- Systematic cold biases are observed in RCMs above 1000–1500 m during winter across the 5th, 50th, and 95th percentiles when evaluated against E-OBS, consistent with findings from [[3]](https://doi.org/10.1007/s00382-024-07376-y).\n\n- CORDEX biases relative to CERRA are smaller in magnitude and do not increase systematically with altitude, unlike the case with E-OBS.\n\n- Biases relative to E-OBS generally increase with elevation (becoming more negative), particularly for the 5th percentile in winter, where the ensemble median reaches its largest negative value.\n\n- Even in summer, models exhibit negative biases at higher elevations relative to E-OBS, notably for the 50th and 95th percentiles.\n\n- The greatest inter-model spread occurs for the 5th percentile in winter.\n\n- CERRA shows large cold differences relative to E-OBS; for the 50th and 95th percentiles in winter, it is colder than the CORDEX ensemble median above 2000–2500 m.\n\n- Applying a lapse-rate correction to ensure all datasets are compared at the same elevation does not substantially alter the shape of the vertical bias profiles.\n\n- The lapse-rate correction introduces small (generally <1 °C) negative differences that increase with elevation, reflecting that regridded model and E-OBS elevations are typically lower than those of regridded CERRA.\n\n(climate_projections-cordex-domains-single-levels_validation_q06:section-3.6)=", "text_with_prefix": "EQC Quality Assessment: \"CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q06 | Category: Climate_Projections\nSection: CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA > Analysis and results > 3. Plot and describe results > 3.5. Results summary\n---\n- Regional Climate Models (RCMs) show biases relative to both reference datasets (E-OBS and CERRA).\n\n- Systematic cold biases are observed in RCMs above 1000–1500 m during winter across the 5th, 50th, and 95th percentiles when evaluated against E-OBS, consistent with findings from [[3]](https://doi.org/10.1007/s00382-024-07376-y).\n\n- CORDEX biases relative to CERRA are smaller in magnitude and do not increase systematically with altitude, unlike the case with E-OBS.\n\n- Biases relative to E-OBS generally increase with elevation (becoming more negative), particularly for the 5th percentile in winter, where the ensemble median reaches its largest negative value.\n\n- Even in summer, models exhibit negative biases at higher elevations relative to E-OBS, notably for the 50th and 95th percentiles.\n\n- The greatest inter-model spread occurs for the 5th percentile in winter.\n\n- CERRA shows large cold differences relative to E-OBS; for the 50th and 95th percentiles in winter, it is colder than the CORDEX ensemble median above 2000–2500 m.\n\n- Applying a lapse-rate correction to ensure all datasets are compared at the same elevation does not substantially alter the shape of the vertical bias profiles.\n\n- The lapse-rate correction introduces small (generally <1 °C) negative differences that increase with elevation, reflecting that regridded model and E-OBS elevations are typically lower than those of regridded CERRA.\n\n(climate_projections-cordex-domains-single-levels_validation_q06:section-3.6)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q06__df6ba520b23e", "report_id": "climate_projections-cordex-domains-single-levels_validation_q06", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q06", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA > Analysis and results > 3. Plot and describe results > 3.6. Implications for the users", "title": "CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA", "chunk_index": 22, "token_count": 697, "text_raw": "- Elevation-dependent biases in RCM outputs are systematic and significant above 1000–1500 m, particularly in winter, when evaluated against E-OBS, and affect all percentiles. Users working with climate data in mountainous regions should be aware that cold biases tend to intensify with elevation, especially for cold extremes (5th percentile), potentially influencing impact assessments and adaptation planning.\n\n- Bias magnitude and inter-model spread vary by season and percentile, with the greatest variability observed in winter cold tails (5th percentile). This highlights substantial uncertainty for extreme cold conditions at high elevations, which should be explicitly considered in risk and vulnerability evaluations.\n\n- CERRA exhibits strong cold differences when compared to E-OBS. This has two key implications: (1) caution is needed when using CERRA as a reference or input dataset for mountain climate analyses, and (2) despite its high resolution, the underlying model may not adequately represent complex high-altitude dynamics and processes.\n\n- A critical caveat lies in the inherent uncertainty and sparsity of observational data at high elevations, particularly in E-OBS. These limitations—stemming from sparse station coverage, complex topography, and measurement challenges—can cause some of the apparent model biases to actually reflect deficiencies in the reference data.\n\n- Interpolation accuracy generally declines with reduced station density, performs worse for spatially variable parameters (e.g. precipitation), and deteriorates further in areas of complex terrain such as mountainous regions [[8]](https://doi.org/10.1029/2009JD011799). Moreover, in some cases, E-OBS tends to report warmer values than national station datasets across the 5th, 50th, and 95th percentiles [[3]](https://doi.org/10.1007/s00382-024-07376-y), and has been shown to overestimate temperatures in the distribution tails, even over moderately complex orography [[9]](https://doi.org/10.1029/2010JD014123).\n\n- Lapse-rate correction may slightly reduce biases by adjusting for elevation differences between datasets but does not fundamentally change the shape of vertical bias profiles. While useful for harmonising datasets, it is not a solution to the underlying model or observational issues driving the biases.\n\n- Where feasible, users may benefit from complementing gridded analyses with station-based evaluations corrected for elevation. While the CDS does not currently provide access to raw station-level data, external sources may be used.\n\n- Altogether, the combination of systematic model biases, observational uncertainties, and limited high-elevation data availability underscores the need for caution when interpreting temperature extremes and climate signals in mountainous regions.\n\n- Bias-adjustment methods do not ensure that corrected future projections will more accurately reflect reality, as the underlying biases may change under evolving climate conditions [[10]](https://doi.org/10.5194/hess-25-273-2021).", "text_with_prefix": "EQC Quality Assessment: \"CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q06 | Category: Climate_Projections\nSection: CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA > Analysis and results > 3. Plot and describe results > 3.6. Implications for the users\n---\n- Elevation-dependent biases in RCM outputs are systematic and significant above 1000–1500 m, particularly in winter, when evaluated against E-OBS, and affect all percentiles. Users working with climate data in mountainous regions should be aware that cold biases tend to intensify with elevation, especially for cold extremes (5th percentile), potentially influencing impact assessments and adaptation planning.\n\n- Bias magnitude and inter-model spread vary by season and percentile, with the greatest variability observed in winter cold tails (5th percentile). This highlights substantial uncertainty for extreme cold conditions at high elevations, which should be explicitly considered in risk and vulnerability evaluations.\n\n- CERRA exhibits strong cold differences when compared to E-OBS. This has two key implications: (1) caution is needed when using CERRA as a reference or input dataset for mountain climate analyses, and (2) despite its high resolution, the underlying model may not adequately represent complex high-altitude dynamics and processes.\n\n- A critical caveat lies in the inherent uncertainty and sparsity of observational data at high elevations, particularly in E-OBS. These limitations—stemming from sparse station coverage, complex topography, and measurement challenges—can cause some of the apparent model biases to actually reflect deficiencies in the reference data.\n\n- Interpolation accuracy generally declines with reduced station density, performs worse for spatially variable parameters (e.g. precipitation), and deteriorates further in areas of complex terrain such as mountainous regions [[8]](https://doi.org/10.1029/2009JD011799). Moreover, in some cases, E-OBS tends to report warmer values than national station datasets across the 5th, 50th, and 95th percentiles [[3]](https://doi.org/10.1007/s00382-024-07376-y), and has been shown to overestimate temperatures in the distribution tails, even over moderately complex orography [[9]](https://doi.org/10.1029/2010JD014123).\n\n- Lapse-rate correction may slightly reduce biases by adjusting for elevation differences between datasets but does not fundamentally change the shape of vertical bias profiles. While useful for harmonising datasets, it is not a solution to the underlying model or observational issues driving the biases.\n\n- Where feasible, users may benefit from complementing gridded analyses with station-based evaluations corrected for elevation. While the CDS does not currently provide access to raw station-level data, external sources may be used.\n\n- Altogether, the combination of systematic model biases, observational uncertainties, and limited high-elevation data availability underscores the need for caution when interpreting temperature extremes and climate signals in mountainous regions.\n\n- Bias-adjustment methods do not ensure that corrected future projections will more accurately reflect reality, as the underlying biases may change under evolving climate conditions [[10]](https://doi.org/10.5194/hess-25-273-2021)."} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q06__4821a7ef8d63", "report_id": "climate_projections-cordex-domains-single-levels_validation_q06", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q06", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA > ℹ️ If you want to know more > Key resources", "title": "CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA", "chunk_index": 23, "token_count": 360, "text_raw": "Some key resources and further reading were linked throughout this assessment.\n\n* Regional climate models - CORDEX: https://cds.climate.copernicus.eu/datasets/projections-cordex-domains-single-levels?tab=overview\n* Gridded observation dataset for Europe, E-OBS: https://cds.climate.copernicus.eu/datasets/insitu-gridded-observations-europe?tab=overview\n* CERRA sub-daily regional reanalysis data for Europe on single levels from 1984 to present: https://cds.climate.copernicus.eu/datasets/reanalysis-cerra-single-levels?tab=overview \n* High-resolution global climate models: https://highresmip.org/\n* Mountain tourism meteorological and snow indicators for Europe from 1950 to 2100 derived from reanalysis and climate projections: https://cds.climate.copernicus.eu/datasets/sis-tourism-snow-indicators?tab=overview\n\nCode libraries used:\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)", "text_with_prefix": "EQC Quality Assessment: \"CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q06 | Category: Climate_Projections\nSection: CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA > ℹ️ If you want to know more > Key resources\n---\nSome key resources and further reading were linked throughout this assessment.\n\n* Regional climate models - CORDEX: https://cds.climate.copernicus.eu/datasets/projections-cordex-domains-single-levels?tab=overview\n* Gridded observation dataset for Europe, E-OBS: https://cds.climate.copernicus.eu/datasets/insitu-gridded-observations-europe?tab=overview\n* CERRA sub-daily regional reanalysis data for Europe on single levels from 1984 to present: https://cds.climate.copernicus.eu/datasets/reanalysis-cerra-single-levels?tab=overview \n* High-resolution global climate models: https://highresmip.org/\n* Mountain tourism meteorological and snow indicators for Europe from 1950 to 2100 derived from reanalysis and climate projections: https://cds.climate.copernicus.eu/datasets/sis-tourism-snow-indicators?tab=overview\n\nCode libraries used:\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)"} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q06__9608b2a6aebf", "report_id": "climate_projections-cordex-domains-single-levels_validation_q06", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q06", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA > ℹ️ If you want to know more > References", "title": "CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA", "chunk_index": 24, "token_count": 1029, "text_raw": "[[1]](https://doi.org/10.1038/s41586-019-1822-y) Immerzeel, W.W., Lutz, A.F., Andrade, M. et al., 2020. Importance and vulnerability of the world’s water towers. Nature 577, 364–369. https://doi.org/10.1038/s41586-019-1822-y\n\n[[2]](https://doi.org/10.1175/BAMS-D-23-0082.1) Rudisill, W., Rhoades, A., Xu, Z., and Feldman, D. R.,2024. Are Atmospheric Models Too Cold in the Mountains? The State of Science and Insights from the SAIL Field Campaign. Bull. Amer. Meteor. Soc., 105, E1237–E1264, https://doi.org/10.1175/BAMS-D-23-0082.1\n\n[[3]](https://doi.org/10.1007/s00382-024-07376-y) Matiu, M., Napoli, A., Kotlarski, S. et al., 2024. Elevation-dependent biases of raw and bias-adjusted EURO-CORDEX regional climate models in the European Alps. Clim Dyn 62, 9013–9030. https://doi.org/10.1007/s00382-024-07376-y\n\n[[4]](https://doi.org/10.3390/cli11070154) Shrestha, S., Zaramella, M., Callegari, M., Greifeneder, F., Borga, M., 2023. Scale Dependence of Errors in Snow Water Equivalent Simulations Using ERA5 Reanalysis over Alpine Basins. Climate, 11, 154. https://doi.org/10.3390/cli11070154\n\n[[5]](https://doi.org/10.1007/s00704-018-2506-5) Haslinger, K., Holawe, F., and Blöschl, G., 2019. Spatial characteristics of precipitation shortfalls in the Greater Alpine Region—a data-based analysis from observations. Theor Appl Climatol 136, 717–731. https://doi.org/10.1007/s00704-018-2506-5\n\n[[6]](https://doi.org/10.1002/joc.1377) Auer, I., Böhm, R., Jurkovic, A., Lipa, W., Orlik, A., Potzmann, R., Schöner, W., Ungersböck, M., Matulla, C., Briffa, K., Jones, P., Efthymiadis, D., Brunetti, M., Nanni, T., Maugeri, M., Mercalli, L., Mestre, O., Moisselin, J.-M., Begert, M., Müller-Westermeier, G., Kveton, V., Bochnicek, O., Stastny, P., Lapin, M., Szalai, S., Szentimrey, T., Cegnar, T., Dolinar, M., Gajic-Capka, M., Zaninovic, K., Majstorovic, Z. and Nieplova, E., 2007. HISTALP—historical instrumental climatological surface time series of the Greater Alpine Region. Int. J. Climatol., 27: 17-46. https://doi.org/10.1002/joc.1377\n\n[[7]](https://doi.org/10.1002/joc.5249) Kotlarski, S., Szabó, P., Herrera S., et al., 2019. Observational uncertainty and regional climate model evaluation: A pan-European perspective. Int J Climatol. 39: 3730–3749. https://doi.org/10.1002/joc.5249\n\n[[8]](https://doi.org/10.1029/2009JD011799) Hofstra, N., Haylock, M., New, M., and Jones, P. D., 2009. Testing E-OBS European high-resolution gridded data set of daily precipitation and surface temperature, J. Geophys. Res., 114, D21101. https://doi.org/10.1029/2009JD011799", "text_with_prefix": "EQC Quality Assessment: \"CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q06 | Category: Climate_Projections\nSection: CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA > ℹ️ If you want to know more > References\n---\n[[1]](https://doi.org/10.1038/s41586-019-1822-y) Immerzeel, W.W., Lutz, A.F., Andrade, M. et al., 2020. Importance and vulnerability of the world’s water towers. Nature 577, 364–369. https://doi.org/10.1038/s41586-019-1822-y\n\n[[2]](https://doi.org/10.1175/BAMS-D-23-0082.1) Rudisill, W., Rhoades, A., Xu, Z., and Feldman, D. R.,2024. Are Atmospheric Models Too Cold in the Mountains? The State of Science and Insights from the SAIL Field Campaign. Bull. Amer. Meteor. Soc., 105, E1237–E1264, https://doi.org/10.1175/BAMS-D-23-0082.1\n\n[[3]](https://doi.org/10.1007/s00382-024-07376-y) Matiu, M., Napoli, A., Kotlarski, S. et al., 2024. Elevation-dependent biases of raw and bias-adjusted EURO-CORDEX regional climate models in the European Alps. Clim Dyn 62, 9013–9030. https://doi.org/10.1007/s00382-024-07376-y\n\n[[4]](https://doi.org/10.3390/cli11070154) Shrestha, S., Zaramella, M., Callegari, M., Greifeneder, F., Borga, M., 2023. Scale Dependence of Errors in Snow Water Equivalent Simulations Using ERA5 Reanalysis over Alpine Basins. Climate, 11, 154. https://doi.org/10.3390/cli11070154\n\n[[5]](https://doi.org/10.1007/s00704-018-2506-5) Haslinger, K., Holawe, F., and Blöschl, G., 2019. Spatial characteristics of precipitation shortfalls in the Greater Alpine Region—a data-based analysis from observations. Theor Appl Climatol 136, 717–731. https://doi.org/10.1007/s00704-018-2506-5\n\n[[6]](https://doi.org/10.1002/joc.1377) Auer, I., Böhm, R., Jurkovic, A., Lipa, W., Orlik, A., Potzmann, R., Schöner, W., Ungersböck, M., Matulla, C., Briffa, K., Jones, P., Efthymiadis, D., Brunetti, M., Nanni, T., Maugeri, M., Mercalli, L., Mestre, O., Moisselin, J.-M., Begert, M., Müller-Westermeier, G., Kveton, V., Bochnicek, O., Stastny, P., Lapin, M., Szalai, S., Szentimrey, T., Cegnar, T., Dolinar, M., Gajic-Capka, M., Zaninovic, K., Majstorovic, Z. and Nieplova, E., 2007. HISTALP—historical instrumental climatological surface time series of the Greater Alpine Region. Int. J. Climatol., 27: 17-46. https://doi.org/10.1002/joc.1377\n\n[[7]](https://doi.org/10.1002/joc.5249) Kotlarski, S., Szabó, P., Herrera S., et al., 2019. Observational uncertainty and regional climate model evaluation: A pan-European perspective. Int J Climatol. 39: 3730–3749. https://doi.org/10.1002/joc.5249\n\n[[8]](https://doi.org/10.1029/2009JD011799) Hofstra, N., Haylock, M., New, M., and Jones, P. D., 2009. Testing E-OBS European high-resolution gridded data set of daily precipitation and surface temperature, J. Geophys. Res., 114, D21101. https://doi.org/10.1029/2009JD011799"} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q06__587e82e084b1", "report_id": "climate_projections-cordex-domains-single-levels_validation_q06", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q06", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA > ℹ️ If you want to know more > References", "title": "CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA", "chunk_index": 25, "token_count": 364, "text_raw": "E-OBS European high-resolution gridded data set of daily precipitation and surface temperature, J. Geophys. Res., 114, D21101. https://doi.org/10.1029/2009JD011799\n\n[[9]](https://doi.org/10.1029/2010JD014123) Kyselý, J., and Plavcová, E., 2010. A critical remark on the applicability of E-OBS European gridded temperature data set for validating control climate simulations, J. Geophys. Res., 115, D23118. https://doi.org/10.1029/2010JD014123\n\n[[10]](https://doi.org/10.5194/hess-25-273-2021) Schmith, T., Thejll, P., Berg, P., Boberg, F., Christensen, O. B., Christiansen, B., Christensen, J. H., Madsen, M. S., and Steger, C., 2021. Identifying robust bias adjustment methods for European extreme precipitation in a multi-model pseudo-reality setting, Hydrol. Earth Syst. Sci., 25, 273–290, https://doi.org/10.5194/hess-25-273-2021", "text_with_prefix": "EQC Quality Assessment: \"CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q06 | Category: Climate_Projections\nSection: CORDEX temperature biases over the Alpine region — assessment using E-OBS and CERRA > ℹ️ If you want to know more > References\n---\nE-OBS European high-resolution gridded data set of daily precipitation and surface temperature, J. Geophys. Res., 114, D21101. https://doi.org/10.1029/2009JD011799\n\n[[9]](https://doi.org/10.1029/2010JD014123) Kyselý, J., and Plavcová, E., 2010. A critical remark on the applicability of E-OBS European gridded temperature data set for validating control climate simulations, J. Geophys. Res., 115, D23118. https://doi.org/10.1029/2010JD014123\n\n[[10]](https://doi.org/10.5194/hess-25-273-2021) Schmith, T., Thejll, P., Berg, P., Boberg, F., Christensen, O. B., Christiansen, B., Christensen, J. H., Madsen, M. S., and Steger, C., 2021. Identifying robust bias adjustment methods for European extreme precipitation in a multi-model pseudo-reality setting, Hydrol. Earth Syst. Sci., 25, 273–290, https://doi.org/10.5194/hess-25-273-2021"} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q07__dabdd24a4686", "report_id": "climate_projections-cordex-domains-single-levels_validation_q07", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q07", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CORDEX biases in the SPEI6 drought index over the Mediterranean region", "title": "CORDEX biases in the SPEI6 drought index over the Mediterranean region", "chunk_index": 0, "token_count": 108, "text_raw": "Production date: 17-03-2026\n\nProduced by: CMCC foundation - Euro-Mediterranean Center on Climate Change. Albert Martinez Boti and Lorenzo Sangelantoni.", "text_with_prefix": "EQC Quality Assessment: \"CORDEX biases in the SPEI6 drought index over the Mediterranean region\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q07 | Category: Climate_Projections\nSection: CORDEX biases in the SPEI6 drought index over the Mediterranean region\n---\nProduction date: 17-03-2026\n\nProduced by: CMCC foundation - Euro-Mediterranean Center on Climate Change. Albert Martinez Boti and Lorenzo Sangelantoni."} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q07__41dfadf3d451", "report_id": "climate_projections-cordex-domains-single-levels_validation_q07", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q07", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CORDEX biases in the SPEI6 drought index over the Mediterranean region > Quality assessment question", "title": "CORDEX biases in the SPEI6 drought index over the Mediterranean region", "chunk_index": 1, "token_count": 618, "text_raw": "* **How well do CORDEX models simulate drought compared to signals derived from ERA5?**\n\nThe Standardized Precipitation-Evapotranspiration Index (SPEI) [[1]](https://doi.org/10.1175/2009JCLI2909.1) is a widely used drought indicator that integrates precipitation and potential evapotranspiration, providing a more complete measure of water deficits than precipitation alone. Its multiscalar formulation allows the evaluation of different types of drought [[2]](https://digitalcommons.unl.edu/droughtfacpub/69/), from short-term agricultural or soil moisture deficits to long-term hydrological or ecological events. Being standardised, SPEI is dimensionless, facilitating comparisons across regions and climates. By capturing variations in frequency, duration, and severity, it effectively reflects the complex nature of drought, making it a robust and physically meaningful proxy for drought assessment in the context of climate change [[3]](https://doi.org/10.1175/2012EI000434.1).\n\nUnderstanding SPEI is particularly relevant for the water management sector. By capturing multiple types of drought SPEI serves as a versatile indicator for various users. Climate projections using SPEI enable planners to anticipate future water deficits, guide allocation, and design effective strategies for reservoir management, irrigation, and drought preparedness, helping to reduce socio-economic and ecological impacts [[4]](https://doi.org/10.3390/eesp2025035029)[[5]](https://doi.org/10.1002/joc.7302).\n\nThis assessment evaluates SPEI6 (six-month accumulation) across a subset of [CORDEX](https://cds.climate.copernicus.eu/datasets/projections-cordex-domains-single-levels?tab=overview) models available through the Copernicus Climate Data Store (CDS). Model outputs are compared against SPEI6 derived from [ERA5](https://cds.climate.copernicus.eu/datasets/reanalysis-era5-complete?tab=doc) reanalysis, which serves as the reference product. We specifically evaluate whether changes in the 1991–2010 period relative to the 1971–1990 baseline are accurately represented by CORDEX models and the trend across the whole 1971-2010 period for the Mediterranean domain.\n\nThis notebook is the counterpart to the assessment performed for CMIP6 ([CMIP6 biases in the SPEI6 drought index over the Mediterranean region](https://ecmwf-projects.github.io/c3s2-eqc-quality-assessment/Climate_Projections/CMIP6/climate_projections-cmip6_validation_q13.html))", "text_with_prefix": "EQC Quality Assessment: \"CORDEX biases in the SPEI6 drought index over the Mediterranean region\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q07 | Category: Climate_Projections\nSection: CORDEX biases in the SPEI6 drought index over the Mediterranean region > Quality assessment question\n---\n* **How well do CORDEX models simulate drought compared to signals derived from ERA5?**\n\nThe Standardized Precipitation-Evapotranspiration Index (SPEI) [[1]](https://doi.org/10.1175/2009JCLI2909.1) is a widely used drought indicator that integrates precipitation and potential evapotranspiration, providing a more complete measure of water deficits than precipitation alone. Its multiscalar formulation allows the evaluation of different types of drought [[2]](https://digitalcommons.unl.edu/droughtfacpub/69/), from short-term agricultural or soil moisture deficits to long-term hydrological or ecological events. Being standardised, SPEI is dimensionless, facilitating comparisons across regions and climates. By capturing variations in frequency, duration, and severity, it effectively reflects the complex nature of drought, making it a robust and physically meaningful proxy for drought assessment in the context of climate change [[3]](https://doi.org/10.1175/2012EI000434.1).\n\nUnderstanding SPEI is particularly relevant for the water management sector. By capturing multiple types of drought SPEI serves as a versatile indicator for various users. Climate projections using SPEI enable planners to anticipate future water deficits, guide allocation, and design effective strategies for reservoir management, irrigation, and drought preparedness, helping to reduce socio-economic and ecological impacts [[4]](https://doi.org/10.3390/eesp2025035029)[[5]](https://doi.org/10.1002/joc.7302).\n\nThis assessment evaluates SPEI6 (six-month accumulation) across a subset of [CORDEX](https://cds.climate.copernicus.eu/datasets/projections-cordex-domains-single-levels?tab=overview) models available through the Copernicus Climate Data Store (CDS). Model outputs are compared against SPEI6 derived from [ERA5](https://cds.climate.copernicus.eu/datasets/reanalysis-era5-complete?tab=doc) reanalysis, which serves as the reference product. We specifically evaluate whether changes in the 1991–2010 period relative to the 1971–1990 baseline are accurately represented by CORDEX models and the trend across the whole 1971-2010 period for the Mediterranean domain.\n\nThis notebook is the counterpart to the assessment performed for CMIP6 ([CMIP6 biases in the SPEI6 drought index over the Mediterranean region](https://ecmwf-projects.github.io/c3s2-eqc-quality-assessment/Climate_Projections/CMIP6/climate_projections-cmip6_validation_q13.html))"} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q07__8fe39ef6392c", "report_id": "climate_projections-cordex-domains-single-levels_validation_q07", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q07", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CORDEX biases in the SPEI6 drought index over the Mediterranean region > Quality assessment statement", "title": "CORDEX biases in the SPEI6 drought index over the Mediterranean region", "chunk_index": 2, "token_count": 423, "text_raw": "These are the key outcomes of this assessment\n\n* The CORDEX ensemble median for the subset of models considered in this assessment does not generally capture the general intensification of drought over the Mediterranean observed in ERA5, showing SPEI6 changes close to zero for most regions.\n\n* Individual models exhibit highly diverse spatial patterns, with some closely resembling ERA5, underscoring the importance of taking into account the whole ensemble (and not only the ensemble mean) for planning purposes.\n\n* Despite these biases, the selected subset of CORDEX models may still be useful for exploring SPEI-based future water deficits and guiding drought preparedness, as long as limitations are considered.\n\n* Regions with high inter-model spread, such as North Africa, Turkey, and the eastern Mediterranean, may be associated with future higher uncertainty, suggesting that adaptation strategies in these areas should consider a wider range of possible futures.\n\n```\n\nattachment:591afcf9-a410-4d53-ac0f-713bd18acac5.png \n---\nwidth: 1500px\nalt: Visual abstract \n---\nSpatial patterns of SPEI6 change over the Mediterranean region. Panels show (a) ERA5 reference, (b) CORDEX ensemble median (median of the selected subset of models at each grid cell), (c) ensemble median bias relative to ERA5, and (d) ensemble spread (standard deviation across the selected models). SPEI6 is standardised over the reference period 1971–2010. SPEI6 change is calculated at each grid point as the difference between the monthly climatologies of the target period (1991–2010) and the baseline period (1971–1990), and then averaged to obtain the ANNUAL mean.\n```", "text_with_prefix": "EQC Quality Assessment: \"CORDEX biases in the SPEI6 drought index over the Mediterranean region\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q07 | Category: Climate_Projections\nSection: CORDEX biases in the SPEI6 drought index over the Mediterranean region > Quality assessment statement\n---\nThese are the key outcomes of this assessment\n\n* The CORDEX ensemble median for the subset of models considered in this assessment does not generally capture the general intensification of drought over the Mediterranean observed in ERA5, showing SPEI6 changes close to zero for most regions.\n\n* Individual models exhibit highly diverse spatial patterns, with some closely resembling ERA5, underscoring the importance of taking into account the whole ensemble (and not only the ensemble mean) for planning purposes.\n\n* Despite these biases, the selected subset of CORDEX models may still be useful for exploring SPEI-based future water deficits and guiding drought preparedness, as long as limitations are considered.\n\n* Regions with high inter-model spread, such as North Africa, Turkey, and the eastern Mediterranean, may be associated with future higher uncertainty, suggesting that adaptation strategies in these areas should consider a wider range of possible futures.\n\n```\n\nattachment:591afcf9-a410-4d53-ac0f-713bd18acac5.png \n---\nwidth: 1500px\nalt: Visual abstract \n---\nSpatial patterns of SPEI6 change over the Mediterranean region. Panels show (a) ERA5 reference, (b) CORDEX ensemble median (median of the selected subset of models at each grid cell), (c) ensemble median bias relative to ERA5, and (d) ensemble spread (standard deviation across the selected models). SPEI6 is standardised over the reference period 1971–2010. SPEI6 change is calculated at each grid point as the difference between the monthly climatologies of the target period (1991–2010) and the baseline period (1971–1990), and then averaged to obtain the ANNUAL mean.\n```"} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q07__a517b465f711", "report_id": "climate_projections-cordex-domains-single-levels_validation_q07", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q07", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CORDEX biases in the SPEI6 drought index over the Mediterranean region > Methodology", "title": "CORDEX biases in the SPEI6 drought index over the Mediterranean region", "chunk_index": 3, "token_count": 1064, "text_raw": "The Standardized Precipitation-Evapotranspiration Index accumulated to six months (SPEI6) has been calculated for this assessment — through the use of [xclim](https://xclim.readthedocs.io/en/stable/) — as a proxy to evaluate drought conditions for a subset of [CORDEX](https://cds.climate.copernicus.eu/datasets/projections-cordex-domains-single-levels?tab=overview) models, which are evaluated against SPEI6 derived from [ERA5](https://cds.climate.copernicus.eu/datasets/reanalysis-era5-complete?tab=doc) reanalysis. The study covers the Mediterranean domain, following IPCC-AR6 regional definitions [[6]](https://doi.org/10.5194/essd-12-2959-2020), and results include spatial maps of SPEI6 changes for ERA5, CORDEX, and the biases of these changes. Additionally, time series showing the temporal evolution of drought across the domain are calculated, comparing trends from the models with ERA5 over the historical period considered within this assessment (1971–2010).\n\nPotential evapotranspiration (PET) is calculated with the Hargreaves method [[7]](https://doi.org/10.13031/2013.26773), which requires only maximum and minimum temperature (and optionally mean temperature), providing a computationally efficient yet robust estimate (e.g., [[8]](https://doi.org/10.1002/joc.3887)). The accumulated series of precipitation minus PET is then standardised for the reference period, defined by the [C3S Atlas](https://confluence.ecmwf.int/display/CKB/Gridded+data+underpinning+the+Copernicus+Interactive+Climate+Atlas%3A+Description+of+the+datasets+and+variables#GriddeddataunderpinningtheCopernicusInteractiveClimateAtlas:Descriptionofthedatasetsandvariables) (1971–2010).\n\nFive key concepts are defined:\n\n- **Reference period:** The period used to standardise SPEI values, here 1971–2010, following the [C3S Atlas](https://confluence.ecmwf.int/display/CKB/Gridded+data+underpinning+the+Copernicus+Interactive+Climate+Atlas%3A+Description+of+the+datasets+and+variables#GriddeddataunderpinningtheCopernicusInteractiveClimateAtlas:Descriptionofthedatasetsandvariables). \n- **Baseline period:** The period against which changes are calculated, here 1971–1990. \n- **Target period:** The period for which changes are evaluated, here 1991–2010. \n- **Historical period:** The entire period considered in this assessment, here 1971–2010. \n- **SPEI6 change:** Differences in SPEI6 between the target period and the baseline, representing anomalies in drought conditions.\n\nThe analysis and results follow the next outline:\n\n**[](climate_projections-cordex_validation_q07:section-1)**\n * [](climate_projections-cordex_validation_q07:section-1.1)\n * [](climate_projections-cordex_validation_q07:section-1.2)\n * [](climate_projections-cordex_validation_q07:section-1.3)\n * [](climate_projections-cordex_validation_q07:section-1.4)\n * [](climate_projections-cordex_validation_q07:section-1.5)\n * [](climate_projections-cordex_validation_q07:section-1.6)\n\n**[](climate_projections-cordex_validation_q07:section-2)**\n * [](climate_projections-cordex_validation_q07:section-2.1)\n * [](climate_projections-cordex_validation_q07:section-2.2)\n * [](climate_projections-cordex_validation_q07:section-2.3)\n\n**[](climate_projections-cordex_validation_q07:section-3)**\n * [](climate_projections-cordex_validation_q07:section-3.1)\n * [](climate_projections-cordex_validation_q07:section-3.2)\n * [](climate_projections-cordex_validation_q07:section-3.3)\n * [](climate_projections-cordex_validation_q07:section-3.4)\n * [](climate_projections-cordex_validation_q07:section-3.5)\n * [](climate_projections-cordex_validation_q07:section-3.6)\n * [](climate_projections-cordex_validation_q07:section-3.7)\n * [](climate_projections-cordex_validation_q07:section-3.8)", "text_with_prefix": "EQC Quality Assessment: \"CORDEX biases in the SPEI6 drought index over the Mediterranean region\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q07 | Category: Climate_Projections\nSection: CORDEX biases in the SPEI6 drought index over the Mediterranean region > Methodology\n---\nThe Standardized Precipitation-Evapotranspiration Index accumulated to six months (SPEI6) has been calculated for this assessment — through the use of [xclim](https://xclim.readthedocs.io/en/stable/) — as a proxy to evaluate drought conditions for a subset of [CORDEX](https://cds.climate.copernicus.eu/datasets/projections-cordex-domains-single-levels?tab=overview) models, which are evaluated against SPEI6 derived from [ERA5](https://cds.climate.copernicus.eu/datasets/reanalysis-era5-complete?tab=doc) reanalysis. The study covers the Mediterranean domain, following IPCC-AR6 regional definitions [[6]](https://doi.org/10.5194/essd-12-2959-2020), and results include spatial maps of SPEI6 changes for ERA5, CORDEX, and the biases of these changes. Additionally, time series showing the temporal evolution of drought across the domain are calculated, comparing trends from the models with ERA5 over the historical period considered within this assessment (1971–2010).\n\nPotential evapotranspiration (PET) is calculated with the Hargreaves method [[7]](https://doi.org/10.13031/2013.26773), which requires only maximum and minimum temperature (and optionally mean temperature), providing a computationally efficient yet robust estimate (e.g., [[8]](https://doi.org/10.1002/joc.3887)). The accumulated series of precipitation minus PET is then standardised for the reference period, defined by the [C3S Atlas](https://confluence.ecmwf.int/display/CKB/Gridded+data+underpinning+the+Copernicus+Interactive+Climate+Atlas%3A+Description+of+the+datasets+and+variables#GriddeddataunderpinningtheCopernicusInteractiveClimateAtlas:Descriptionofthedatasetsandvariables) (1971–2010).\n\nFive key concepts are defined:\n\n- **Reference period:** The period used to standardise SPEI values, here 1971–2010, following the [C3S Atlas](https://confluence.ecmwf.int/display/CKB/Gridded+data+underpinning+the+Copernicus+Interactive+Climate+Atlas%3A+Description+of+the+datasets+and+variables#GriddeddataunderpinningtheCopernicusInteractiveClimateAtlas:Descriptionofthedatasetsandvariables). \n- **Baseline period:** The period against which changes are calculated, here 1971–1990. \n- **Target period:** The period for which changes are evaluated, here 1991–2010. \n- **Historical period:** The entire period considered in this assessment, here 1971–2010. \n- **SPEI6 change:** Differences in SPEI6 between the target period and the baseline, representing anomalies in drought conditions.\n\nThe analysis and results follow the next outline:\n\n**[](climate_projections-cordex_validation_q07:section-1)**\n * [](climate_projections-cordex_validation_q07:section-1.1)\n * [](climate_projections-cordex_validation_q07:section-1.2)\n * [](climate_projections-cordex_validation_q07:section-1.3)\n * [](climate_projections-cordex_validation_q07:section-1.4)\n * [](climate_projections-cordex_validation_q07:section-1.5)\n * [](climate_projections-cordex_validation_q07:section-1.6)\n\n**[](climate_projections-cordex_validation_q07:section-2)**\n * [](climate_projections-cordex_validation_q07:section-2.1)\n * [](climate_projections-cordex_validation_q07:section-2.2)\n * [](climate_projections-cordex_validation_q07:section-2.3)\n\n**[](climate_projections-cordex_validation_q07:section-3)**\n * [](climate_projections-cordex_validation_q07:section-3.1)\n * [](climate_projections-cordex_validation_q07:section-3.2)\n * [](climate_projections-cordex_validation_q07:section-3.3)\n * [](climate_projections-cordex_validation_q07:section-3.4)\n * [](climate_projections-cordex_validation_q07:section-3.5)\n * [](climate_projections-cordex_validation_q07:section-3.6)\n * [](climate_projections-cordex_validation_q07:section-3.7)\n * [](climate_projections-cordex_validation_q07:section-3.8)"} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q07__cd74024b2c0b", "report_id": "climate_projections-cordex-domains-single-levels_validation_q07", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q07", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CORDEX biases in the SPEI6 drought index over the Mediterranean region > Methodology", "title": "CORDEX biases in the SPEI6 drought index over the Mediterranean region", "chunk_index": 4, "token_count": 313, "text_raw": "-cordex_validation_q07:section-3.6)\n * [](climate_projections-cordex_validation_q07:section-3.7)\n * [](climate_projections-cordex_validation_q07:section-3.8)\n\n\n
\nNOTE ON THE METHODOLOGY:
\nThis notebook follows the same methodology as the assessment performed for CMIP6 \n(\nCMIP6 biases in the SPEI6 drought index over the Mediterranean region). \nThe historical period considered here is slightly shorter (1971–2010) compared to 1961–2020 in the CMIP6 assessment. This choice helps to avoid memory overload, ensures efficient execution of the notebook, guarantees availability of CORDEX models for the historical period, and avoids including many years of scenario data (CORDEX historical runs only extend until 2005).\n
", "text_with_prefix": "EQC Quality Assessment: \"CORDEX biases in the SPEI6 drought index over the Mediterranean region\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q07 | Category: Climate_Projections\nSection: CORDEX biases in the SPEI6 drought index over the Mediterranean region > Methodology\n---\n-cordex_validation_q07:section-3.6)\n * [](climate_projections-cordex_validation_q07:section-3.7)\n * [](climate_projections-cordex_validation_q07:section-3.8)\n\n\n
\nNOTE ON THE METHODOLOGY:
\nThis notebook follows the same methodology as the assessment performed for CMIP6 \n(\nCMIP6 biases in the SPEI6 drought index over the Mediterranean region). \nThe historical period considered here is slightly shorter (1971–2010) compared to 1961–2020 in the CMIP6 assessment. This choice helps to avoid memory overload, ensures efficient execution of the notebook, guarantees availability of CORDEX models for the historical period, and avoids including many years of scenario data (CORDEX historical runs only extend until 2005).\n
"} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q07__ed8103acbabd", "report_id": "climate_projections-cordex-domains-single-levels_validation_q07", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q07", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CORDEX biases in the SPEI6 drought index over the Mediterranean region > Analysis and results > 1. Parameters, requests and functions definition > 1.2. Define Parameters", "title": "CORDEX biases in the SPEI6 drought index over the Mediterranean region", "chunk_index": 5, "token_count": 585, "text_raw": "In the \"Define Parameters\" section, various customisable options for the notebook are specified:\n\n- The historical period can be adjusted by modifying `historical_slice`; the default range is 1971-2010. \n- The reference period can be adjusted by modifying `reference_slice`; the default range is 1971–2010, following the [C3S Atlas](https://confluence.ecmwf.int/display/CKB/Gridded+data+underpinning+the+Copernicus+Interactive+Climate+Atlas%3A+Description+of+the+datasets+and+variables#GriddeddataunderpinningtheCopernicusInteractiveClimateAtlas:Descriptionofthedatasetsandvariables).\n- The baseline period can be adjusted by modifying `baseline_slice`; the default range is 1971-1990.\n- The target period can be adjusted by modifying `target_slice`; the default range is 1991-2010.\n- The `interpolation_method` parameter allows selecting the interpolation method when regridding is performed.\n- `collection_id` is not customisable for this assessment and is set to 'CORDEX'.\n- The `area` parameter specifies the geographical domain of interest. By default, the Mediterranean domain is used, following IPCC-AR6 regional definitions [[6]](https://doi.org/10.5194/essd-12-2959-2020)\n- The `time_agg` allows selecting the time aggregation for the analysis. Options include: annual, DJF, MAM, JJA, SON, Jan, Feb, Mar, Apr, May, Jun, Jul, Aug, Sep, Oct, Nov, Dec.\n- SPEI accumulation window (`window`): Defines the accumulation period in months for the SPEI calculation. Common options include 3, 6, or 12 months, depending on whether short-term or long-term droughts are of interest. In this assessment, a six-month accumulation is used.\n- The `chunks` selection allows the user to define if dividing into chunks when downloading the data on their local machine. Although it does not significantly affect the analysis, it is recommended to keep the default value for optimal performance.\n\nPeriods definition\nInterpolation method\nCollection id\nDefine region for analysis\ntime aggregation\nSPEI accumulation\nChunks for download\n\n(climate_projections-cordex_validation_q07:section-1.3)=", "text_with_prefix": "EQC Quality Assessment: \"CORDEX biases in the SPEI6 drought index over the Mediterranean region\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q07 | Category: Climate_Projections\nSection: CORDEX biases in the SPEI6 drought index over the Mediterranean region > Analysis and results > 1. Parameters, requests and functions definition > 1.2. Define Parameters\n---\nIn the \"Define Parameters\" section, various customisable options for the notebook are specified:\n\n- The historical period can be adjusted by modifying `historical_slice`; the default range is 1971-2010. \n- The reference period can be adjusted by modifying `reference_slice`; the default range is 1971–2010, following the [C3S Atlas](https://confluence.ecmwf.int/display/CKB/Gridded+data+underpinning+the+Copernicus+Interactive+Climate+Atlas%3A+Description+of+the+datasets+and+variables#GriddeddataunderpinningtheCopernicusInteractiveClimateAtlas:Descriptionofthedatasetsandvariables).\n- The baseline period can be adjusted by modifying `baseline_slice`; the default range is 1971-1990.\n- The target period can be adjusted by modifying `target_slice`; the default range is 1991-2010.\n- The `interpolation_method` parameter allows selecting the interpolation method when regridding is performed.\n- `collection_id` is not customisable for this assessment and is set to 'CORDEX'.\n- The `area` parameter specifies the geographical domain of interest. By default, the Mediterranean domain is used, following IPCC-AR6 regional definitions [[6]](https://doi.org/10.5194/essd-12-2959-2020)\n- The `time_agg` allows selecting the time aggregation for the analysis. Options include: annual, DJF, MAM, JJA, SON, Jan, Feb, Mar, Apr, May, Jun, Jul, Aug, Sep, Oct, Nov, Dec.\n- SPEI accumulation window (`window`): Defines the accumulation period in months for the SPEI calculation. Common options include 3, 6, or 12 months, depending on whether short-term or long-term droughts are of interest. In this assessment, a six-month accumulation is used.\n- The `chunks` selection allows the user to define if dividing into chunks when downloading the data on their local machine. Although it does not significantly affect the analysis, it is recommended to keep the default value for optimal performance.\n\nPeriods definition\nInterpolation method\nCollection id\nDefine region for analysis\ntime aggregation\nSPEI accumulation\nChunks for download\n\n(climate_projections-cordex_validation_q07:section-1.3)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q07__1b13e4812aa6", "report_id": "climate_projections-cordex-domains-single-levels_validation_q07", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q07", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CORDEX biases in the SPEI6 drought index over the Mediterranean region > Analysis and results > 1. Parameters, requests and functions definition > 1.3. Define models", "title": "CORDEX biases in the SPEI6 drought index over the Mediterranean region", "chunk_index": 6, "token_count": 292, "text_raw": "The following climate analyses are performed using a subset of CORDEX models that provide both the historical and RCP8.5 experiments, as well as maximum, minimum and mean temperature and precipitation, within the Climate Data Store ([CDS](https://cds.climate.copernicus.eu/datasets/projections-cordex-domains-single-levels?tab=overview)). All Global Climate Models (GCMs) driving the selected CORDEX simulations were included, and the selection ensures a representative coverage of the available Regional Climate Models (RCMs). A total of 17 models were chosen to maintain consistency with the number used in the quality assessment [CMIP6 biases in the SPEI6 drought index over the Mediterranean region](https://ecmwf-projects.github.io/c3s2-eqc-quality-assessment/Climate_Projections/CMIP6/climate_projections-cmip6_validation_q13.html).\n\n(climate_projections-cordex_validation_q07:section-1.4)=", "text_with_prefix": "EQC Quality Assessment: \"CORDEX biases in the SPEI6 drought index over the Mediterranean region\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q07 | Category: Climate_Projections\nSection: CORDEX biases in the SPEI6 drought index over the Mediterranean region > Analysis and results > 1. Parameters, requests and functions definition > 1.3. Define models\n---\nThe following climate analyses are performed using a subset of CORDEX models that provide both the historical and RCP8.5 experiments, as well as maximum, minimum and mean temperature and precipitation, within the Climate Data Store ([CDS](https://cds.climate.copernicus.eu/datasets/projections-cordex-domains-single-levels?tab=overview)). All Global Climate Models (GCMs) driving the selected CORDEX simulations were included, and the selection ensures a representative coverage of the available Regional Climate Models (RCMs). A total of 17 models were chosen to maintain consistency with the number used in the quality assessment [CMIP6 biases in the SPEI6 drought index over the Mediterranean region](https://ecmwf-projects.github.io/c3s2-eqc-quality-assessment/Climate_Projections/CMIP6/climate_projections-cmip6_validation_q13.html).\n\n(climate_projections-cordex_validation_q07:section-1.4)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q07__ecfc0e56bc0b", "report_id": "climate_projections-cordex-domains-single-levels_validation_q07", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q07", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CORDEX biases in the SPEI6 drought index over the Mediterranean region > Analysis and results > 1. Parameters, requests and functions definition > 1.4. Define ERA5 request", "title": "CORDEX biases in the SPEI6 drought index over the Mediterranean region", "chunk_index": 7, "token_count": 140, "text_raw": "Within this notebook, ERA5 serves as the reference product. In this section, we set the required parameters for the cds-api data-request of ERA5.\n\n(climate_projections-cordex_validation_q07:section-1.5)=", "text_with_prefix": "EQC Quality Assessment: \"CORDEX biases in the SPEI6 drought index over the Mediterranean region\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q07 | Category: Climate_Projections\nSection: CORDEX biases in the SPEI6 drought index over the Mediterranean region > Analysis and results > 1. Parameters, requests and functions definition > 1.4. Define ERA5 request\n---\nWithin this notebook, ERA5 serves as the reference product. In this section, we set the required parameters for the cds-api data-request of ERA5.\n\n(climate_projections-cordex_validation_q07:section-1.5)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q07__6125689a6169", "report_id": "climate_projections-cordex-domains-single-levels_validation_q07", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q07", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CORDEX biases in the SPEI6 drought index over the Mediterranean region > Analysis and results > 1. Parameters, requests and functions definition > 1.5. Define model requests", "title": "CORDEX biases in the SPEI6 drought index over the Mediterranean region", "chunk_index": 8, "token_count": 164, "text_raw": "In this section we set the required parameters for the cds-api data-request.\n\n\"experiment\": \"historical\",\nHistorical (until 2005)\nOnly historical\nHistorical part (until 2005)\nFuture part (RCP8.5 from 2006 onwards)\n\n(climate_projections-cordex_validation_q07:section-1.6)=", "text_with_prefix": "EQC Quality Assessment: \"CORDEX biases in the SPEI6 drought index over the Mediterranean region\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q07 | Category: Climate_Projections\nSection: CORDEX biases in the SPEI6 drought index over the Mediterranean region > Analysis and results > 1. Parameters, requests and functions definition > 1.5. Define model requests\n---\nIn this section we set the required parameters for the cds-api data-request.\n\n\"experiment\": \"historical\",\nHistorical (until 2005)\nOnly historical\nHistorical part (until 2005)\nFuture part (RCP8.5 from 2006 onwards)\n\n(climate_projections-cordex_validation_q07:section-1.6)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q07__da1f0011a97b", "report_id": "climate_projections-cordex-domains-single-levels_validation_q07", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q07", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CORDEX biases in the SPEI6 drought index over the Mediterranean region > Analysis and results > 1. Parameters, requests and functions definition > 1.6. Functions to cache", "title": "CORDEX biases in the SPEI6 drought index over the Mediterranean region", "chunk_index": 9, "token_count": 330, "text_raw": "In this section, functions that will be executed in the caching phase are defined. Caching is the process of storing copies of files in a temporary storage location, so that they can be accessed more quickly. This process also checks if the user has already downloaded a file, avoiding redundant downloads.\n\nDescription of the main functions:\n\n- **`compute_spei_cordex`**: Uses the xclim package to calculate the Standardized Precipitation-Evapotranspiration Index (SPEI). The `reference_slice` argument specifies the reference period used for standardization, while the `window` argument defines the number of months over which precipitation and evapotranspiration are accumulated (6 months in this assessment).\n\n- **`get_mask_below_03`**: Generates a mask identifying grid points where precipitation is below 0.3 mm/day, allowing filtering of extremely dry regions.\n\nOriginal bounds for conservative interpolation\nRechunk spatial dimensions to a single chunk for xESMF / apply_ufunc\nEnsure variables are numeric for xESMF\nApply correct daily reductions\nWrap into dataset\nApply correct daily reductions\n\n(climate_projections-cordex_validation_q07:section-2)=", "text_with_prefix": "EQC Quality Assessment: \"CORDEX biases in the SPEI6 drought index over the Mediterranean region\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q07 | Category: Climate_Projections\nSection: CORDEX biases in the SPEI6 drought index over the Mediterranean region > Analysis and results > 1. Parameters, requests and functions definition > 1.6. Functions to cache\n---\nIn this section, functions that will be executed in the caching phase are defined. Caching is the process of storing copies of files in a temporary storage location, so that they can be accessed more quickly. This process also checks if the user has already downloaded a file, avoiding redundant downloads.\n\nDescription of the main functions:\n\n- **`compute_spei_cordex`**: Uses the xclim package to calculate the Standardized Precipitation-Evapotranspiration Index (SPEI). The `reference_slice` argument specifies the reference period used for standardization, while the `window` argument defines the number of months over which precipitation and evapotranspiration are accumulated (6 months in this assessment).\n\n- **`get_mask_below_03`**: Generates a mask identifying grid points where precipitation is below 0.3 mm/day, allowing filtering of extremely dry regions.\n\nOriginal bounds for conservative interpolation\nRechunk spatial dimensions to a single chunk for xESMF / apply_ufunc\nEnsure variables are numeric for xESMF\nApply correct daily reductions\nWrap into dataset\nApply correct daily reductions\n\n(climate_projections-cordex_validation_q07:section-2)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q07__d1c7c1134d12", "report_id": "climate_projections-cordex-domains-single-levels_validation_q07", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q07", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CORDEX biases in the SPEI6 drought index over the Mediterranean region > Analysis and results > 2. Downloading and processing > 2.1. Download and transform ERA5", "title": "CORDEX biases in the SPEI6 drought index over the Mediterranean region", "chunk_index": 10, "token_count": 190, "text_raw": "In this section, the `download.download_and_transform` function from the 'c3s_eqc_automatic_quality_control' package is used to retrieve ERA5 hourly reference data, resample it to daily frequency, compute daily potential evapotranspiration, calculate the daily water balance, and derive the SPEI (SPEI6 for this assessment). Results are cached to avoid redundant downloads and repeated processing.\n\n(climate_projections-cordex_validation_q07:section-2.2)=", "text_with_prefix": "EQC Quality Assessment: \"CORDEX biases in the SPEI6 drought index over the Mediterranean region\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q07 | Category: Climate_Projections\nSection: CORDEX biases in the SPEI6 drought index over the Mediterranean region > Analysis and results > 2. Downloading and processing > 2.1. Download and transform ERA5\n---\nIn this section, the `download.download_and_transform` function from the 'c3s_eqc_automatic_quality_control' package is used to retrieve ERA5 hourly reference data, resample it to daily frequency, compute daily potential evapotranspiration, calculate the daily water balance, and derive the SPEI (SPEI6 for this assessment). Results are cached to avoid redundant downloads and repeated processing.\n\n(climate_projections-cordex_validation_q07:section-2.2)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q07__43a0927a7c2d", "report_id": "climate_projections-cordex-domains-single-levels_validation_q07", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q07", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CORDEX biases in the SPEI6 drought index over the Mediterranean region > Analysis and results > 2. Downloading and processing > 2.2. Download and transform models", "title": "CORDEX biases in the SPEI6 drought index over the Mediterranean region", "chunk_index": 11, "token_count": 536, "text_raw": "In this section, the `download.download_and_transform` function from the 'c3s_eqc_automatic_quality_control' package is used to download daily data from CORDEX models, compute daily potential evapotranspiration, calculate the daily water balance, and derive the SPEI (SPEI6 for this assessment). SPEI6 is computed on each model's native grid and also on the ERA5 grid to enable bias assessment. When regridding is required, it is performed for each essential climate variable before the index calculation to preserve standardisation. Results are cached to avoid redundant downloads and repeated processing.\n\nOriginal model\nInterpolated model\n\n```text\nmodel='cccma_canesm2_clmcom_clm_cclm4_8_17'\nmodel='cccma_canesm2_gerics_remo2015'\nmodel='cnrm_cerfacs_cm5_ipsl_wrf381p'\nmodel='cnrm_cerfacs_cm5_smhi_rca4'\nmodel='ichec_ec_earth_dmi_hirham5'\nmodel='ichec_ec_earth_knmi_racmo22e'\nmodel='ipsl_cm5a_mr_ipsl_wrf381p'\nmodel='ipsl_cm5a_mr_knmi_racmo22e'\nmodel='miroc_miroc5_clmcom_clm_cclm4_8_17'\nmodel='miroc_miroc5_gerics_remo2015'\nmodel='mohc_hadgem2_es_cnrm_aladin63'\nmodel='mohc_hadgem2_es_mohc_hadrem3_ga7_05'\nmodel='mpi_m_mpi_esm_lr_mpi_csc_remo2009'\nmodel='mpi_m_mpi_esm_lr_cnrm_aladin63'\nmodel='ncc_noresm1_m_dmi_hirham5'\nmodel='ncc_noresm1_m_mohc_hadrem3_ga7_05'\nmodel='ncc_noresm1_m_smhi_rca4'\n```\n\n(climate_projections-cordex_validation_q07:section-2.3)=", "text_with_prefix": "EQC Quality Assessment: \"CORDEX biases in the SPEI6 drought index over the Mediterranean region\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q07 | Category: Climate_Projections\nSection: CORDEX biases in the SPEI6 drought index over the Mediterranean region > Analysis and results > 2. Downloading and processing > 2.2. Download and transform models\n---\nIn this section, the `download.download_and_transform` function from the 'c3s_eqc_automatic_quality_control' package is used to download daily data from CORDEX models, compute daily potential evapotranspiration, calculate the daily water balance, and derive the SPEI (SPEI6 for this assessment). SPEI6 is computed on each model's native grid and also on the ERA5 grid to enable bias assessment. When regridding is required, it is performed for each essential climate variable before the index calculation to preserve standardisation. Results are cached to avoid redundant downloads and repeated processing.\n\nOriginal model\nInterpolated model\n\n```text\nmodel='cccma_canesm2_clmcom_clm_cclm4_8_17'\nmodel='cccma_canesm2_gerics_remo2015'\nmodel='cnrm_cerfacs_cm5_ipsl_wrf381p'\nmodel='cnrm_cerfacs_cm5_smhi_rca4'\nmodel='ichec_ec_earth_dmi_hirham5'\nmodel='ichec_ec_earth_knmi_racmo22e'\nmodel='ipsl_cm5a_mr_ipsl_wrf381p'\nmodel='ipsl_cm5a_mr_knmi_racmo22e'\nmodel='miroc_miroc5_clmcom_clm_cclm4_8_17'\nmodel='miroc_miroc5_gerics_remo2015'\nmodel='mohc_hadgem2_es_cnrm_aladin63'\nmodel='mohc_hadgem2_es_mohc_hadrem3_ga7_05'\nmodel='mpi_m_mpi_esm_lr_mpi_csc_remo2009'\nmodel='mpi_m_mpi_esm_lr_cnrm_aladin63'\nmodel='ncc_noresm1_m_dmi_hirham5'\nmodel='ncc_noresm1_m_mohc_hadrem3_ga7_05'\nmodel='ncc_noresm1_m_smhi_rca4'\n```\n\n(climate_projections-cordex_validation_q07:section-2.3)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q07__900c031b3e67", "report_id": "climate_projections-cordex-domains-single-levels_validation_q07", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q07", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CORDEX biases in the SPEI6 drought index over the Mediterranean region > Analysis and results > 2. Downloading and processing > 2.3. Apply land-sea mask and change some attributes", "title": "CORDEX biases in the SPEI6 drought index over the Mediterranean region", "chunk_index": 12, "token_count": 436, "text_raw": "This section changes some attributes and applies a land–sea mask to all models, as well as to ERA5. A bare-earth (barrean) mask is also applied to ensure meaningful results in arid regions, as in the [C3S atlas](https://atlas.climate.copernicus.eu/atlas). In particular, we follow the methodology available in their [Github repository](https://github.com/ecmwf-projects/c3s-atlas/blob/main/auxiliar/barrean_mask.ipynb). While their implementation of the mask also considers factors such as snow depth and vegetation cover, here we only apply the filter based on annual mean precipitation for 1971–2010 being less than 0.3 mm day⁻¹. This simplification is justified because, in the region under consideration, the bare-earth mask obtained with the full set of filters closely matches the one derived from the precipitation threshold alone.\n\n**Note:** `ds_interpolated` contains data from the models regridded to the ERA5 grid. `model_datasets` contain the same data on the original grid of each model. Regridding is performed for every essential climate variable prior to the index calculation in order to avoid compromising standardisation.\n\nDownload and prepare lsm\nEnsure CF-compliant attributes for model datasets\nApply land-sea mask to each model\nApply precipitation mask to ERA5 and interpolated\nApply PR mask to each model\n\n```text\n100%|██████████| 1/1 [00:00<00:00, 22.34it/s]\n```\n\n(climate_projections-cordex_validation_q07:section-3)=", "text_with_prefix": "EQC Quality Assessment: \"CORDEX biases in the SPEI6 drought index over the Mediterranean region\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q07 | Category: Climate_Projections\nSection: CORDEX biases in the SPEI6 drought index over the Mediterranean region > Analysis and results > 2. Downloading and processing > 2.3. Apply land-sea mask and change some attributes\n---\nThis section changes some attributes and applies a land–sea mask to all models, as well as to ERA5. A bare-earth (barrean) mask is also applied to ensure meaningful results in arid regions, as in the [C3S atlas](https://atlas.climate.copernicus.eu/atlas). In particular, we follow the methodology available in their [Github repository](https://github.com/ecmwf-projects/c3s-atlas/blob/main/auxiliar/barrean_mask.ipynb). While their implementation of the mask also considers factors such as snow depth and vegetation cover, here we only apply the filter based on annual mean precipitation for 1971–2010 being less than 0.3 mm day⁻¹. This simplification is justified because, in the region under consideration, the bare-earth mask obtained with the full set of filters closely matches the one derived from the precipitation threshold alone.\n\n**Note:** `ds_interpolated` contains data from the models regridded to the ERA5 grid. `model_datasets` contain the same data on the original grid of each model. Regridding is performed for every essential climate variable prior to the index calculation in order to avoid compromising standardisation.\n\nDownload and prepare lsm\nEnsure CF-compliant attributes for model datasets\nApply land-sea mask to each model\nApply precipitation mask to ERA5 and interpolated\nApply PR mask to each model\n\n```text\n100%|██████████| 1/1 [00:00<00:00, 22.34it/s]\n```\n\n(climate_projections-cordex_validation_q07:section-3)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q07__9770559b7bd4", "report_id": "climate_projections-cordex-domains-single-levels_validation_q07", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q07", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CORDEX biases in the SPEI6 drought index over the Mediterranean region > Analysis and results > 3. Plot and describe results", "title": "CORDEX biases in the SPEI6 drought index over the Mediterranean region", "chunk_index": 13, "token_count": 260, "text_raw": "This section will display the following results:\n\n- SPEI6 change maps comparing ERA5 and the ensemble median (defined as the median of the change values for the selected subset of models at each grid cell). The layout includes ERA5, the ensemble median, the ensemble median bias, and the ensemble spread (calculated as the standard deviation of the change across the selected subset of models). \n- SPEI6 change maps for each individual model. \n- Bias maps of the SPEI6 change for each model. \n- Time series of SPEI6 change (averaged over the Mediterranean region).\n\n**Note:** SPEI6 change is calculated at each grid point as the arithmetic difference between climatologies in the target period (1991–2010) and the baseline period (1971–1990).\n\n(climate_projections-cordex_validation_q07:section-3.1)=", "text_with_prefix": "EQC Quality Assessment: \"CORDEX biases in the SPEI6 drought index over the Mediterranean region\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q07 | Category: Climate_Projections\nSection: CORDEX biases in the SPEI6 drought index over the Mediterranean region > Analysis and results > 3. Plot and describe results\n---\nThis section will display the following results:\n\n- SPEI6 change maps comparing ERA5 and the ensemble median (defined as the median of the change values for the selected subset of models at each grid cell). The layout includes ERA5, the ensemble median, the ensemble median bias, and the ensemble spread (calculated as the standard deviation of the change across the selected subset of models). \n- SPEI6 change maps for each individual model. \n- Bias maps of the SPEI6 change for each model. \n- Time series of SPEI6 change (averaged over the Mediterranean region).\n\n**Note:** SPEI6 change is calculated at each grid point as the arithmetic difference between climatologies in the target period (1991–2010) and the baseline period (1971–1990).\n\n(climate_projections-cordex_validation_q07:section-3.1)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q07__92684b81ad2f", "report_id": "climate_projections-cordex-domains-single-levels_validation_q07", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q07", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CORDEX biases in the SPEI6 drought index over the Mediterranean region > Analysis and results > 3. Plot and describe results > 3.1. Define plotting functions", "title": "CORDEX biases in the SPEI6 drought index over the Mediterranean region", "chunk_index": 14, "token_count": 506, "text_raw": "The functions presented here are used to calculate SPEI6 changes and generate corresponding layout plots. Three types of layout can be displayed, depending on the plotting function:\n\n1. Reference and ensemble summary: Includes the ERA5 reference product, the ensemble median, the bias of the ensemble median, and the ensemble spread. This is generated using `plot_ensemble()`.\n\n2. Individual models: Displays all models individually using `plot_models()`.\n\n3. Model biases: Displays the bias of each model relative to the reference, also using `plot_models()`.\n\n**Calculation of SPEI6 change:**\n\n- `compute_spei_change()` calculates SPEI6 change at each grid point as the arithmetic difference between monthly climatologies of the target period (1991–2010) and the baseline period (1971–1990). It then aggregates the change according to the user-specified `time_agg` in Section 1.2.\n\n- `compute_change_4timeseries()` computes SPEI6 anomalies relative to the baseline climatology for each month of the timeseries and aggregates them according to the `time_agg` parameter. This provides monthly, seasonal, or annual-level information depending on the user’s selection.\n\nMapping months to names\nMapping seasons to months\nSelect baseline and target periods\nMonthly climatologies\nAggregate according to user choice\nMapping months to names\nMapping seasons to months\nSelect baseline period\nMonthly climatology for the baseline\nCompute monthly anomalies relative to baseline climatology\nPreserve attributes\nAggregate anomalies\nGroup by year\nSelect months for the season, then group by year\nSelect the month, keep time as is\n--- Determine colour parameters ---\n--- Helper function to split GCM and RCM ---\n--- Plot each model ---\n--- Hide empty axes ---\n--- Shared extent and colourbar ---\nDefault behaviour\nCreate figure and axes\n--- ERA5 plot ---\n--- Ensemble Median ---\n--- Ensemble Bias (Median - ERA5) ---\n--- Ensemble Standard Deviation ---\n\n(climate_projections-cordex_validation_q07:section-3.2)=", "text_with_prefix": "EQC Quality Assessment: \"CORDEX biases in the SPEI6 drought index over the Mediterranean region\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q07 | Category: Climate_Projections\nSection: CORDEX biases in the SPEI6 drought index over the Mediterranean region > Analysis and results > 3. Plot and describe results > 3.1. Define plotting functions\n---\nThe functions presented here are used to calculate SPEI6 changes and generate corresponding layout plots. Three types of layout can be displayed, depending on the plotting function:\n\n1. Reference and ensemble summary: Includes the ERA5 reference product, the ensemble median, the bias of the ensemble median, and the ensemble spread. This is generated using `plot_ensemble()`.\n\n2. Individual models: Displays all models individually using `plot_models()`.\n\n3. Model biases: Displays the bias of each model relative to the reference, also using `plot_models()`.\n\n**Calculation of SPEI6 change:**\n\n- `compute_spei_change()` calculates SPEI6 change at each grid point as the arithmetic difference between monthly climatologies of the target period (1991–2010) and the baseline period (1971–1990). It then aggregates the change according to the user-specified `time_agg` in Section 1.2.\n\n- `compute_change_4timeseries()` computes SPEI6 anomalies relative to the baseline climatology for each month of the timeseries and aggregates them according to the `time_agg` parameter. This provides monthly, seasonal, or annual-level information depending on the user’s selection.\n\nMapping months to names\nMapping seasons to months\nSelect baseline and target periods\nMonthly climatologies\nAggregate according to user choice\nMapping months to names\nMapping seasons to months\nSelect baseline period\nMonthly climatology for the baseline\nCompute monthly anomalies relative to baseline climatology\nPreserve attributes\nAggregate anomalies\nGroup by year\nSelect months for the season, then group by year\nSelect the month, keep time as is\n--- Determine colour parameters ---\n--- Helper function to split GCM and RCM ---\n--- Plot each model ---\n--- Hide empty axes ---\n--- Shared extent and colourbar ---\nDefault behaviour\nCreate figure and axes\n--- ERA5 plot ---\n--- Ensemble Median ---\n--- Ensemble Bias (Median - ERA5) ---\n--- Ensemble Standard Deviation ---\n\n(climate_projections-cordex_validation_q07:section-3.2)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q07__69f5ad558a7d", "report_id": "climate_projections-cordex-domains-single-levels_validation_q07", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q07", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CORDEX biases in the SPEI6 drought index over the Mediterranean region > Analysis and results > 3. Plot and describe results > 3.2. Plot ensemble maps", "title": "CORDEX biases in the SPEI6 drought index over the Mediterranean region", "chunk_index": 15, "token_count": 466, "text_raw": "In this section, we invoke the `plot_ensemble()` function to visualise the SPEI6 change for the following: (a) the reference ERA5 product, (b) the ensemble median (defined as the median of the change values for the selected subset of models at each grid cell), (c) the bias of the ensemble median, and (d) the ensemble spread (calculated as the standard deviation of the change across the selected subset of models).\n\n**ANNUAL aggregation**\n\nChange default colorbars\nShared colorbar for ERA5 and ensemble median\nGet spei change\n\n
\n
\n

Fig 1. \n Spatial patterns of SPEI6 change over the Mediterranean region. Panels show (a) ERA5 reference, (b) CORDEX ensemble median (median of the selected subset of models at each grid cell), (c) ensemble median bias relative to ERA5, and (d) ensemble spread (standard deviation across the selected models). SPEI6 is standardised over the reference period 1971–2010. SPEI6 change is calculated at each grid point as the difference between the monthly climatologies of the target period (1991–2010) and the baseline period (1971–1990), and then averaged to obtain the ANNUAL mean.\n\n

\n\n\n
\nNOTE:
\nThe regridding artefacts visible in panel (d) result from the use of a conservative regridding method, which is commonly recommended when dealing with precipitation. They typically arise from large values in specific models and differences among the models’ native grids.\n\n(climate_projections-cordex_validation_q07:section-3.3)=", "text_with_prefix": "EQC Quality Assessment: \"CORDEX biases in the SPEI6 drought index over the Mediterranean region\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q07 | Category: Climate_Projections\nSection: CORDEX biases in the SPEI6 drought index over the Mediterranean region > Analysis and results > 3. Plot and describe results > 3.2. Plot ensemble maps\n---\nIn this section, we invoke the `plot_ensemble()` function to visualise the SPEI6 change for the following: (a) the reference ERA5 product, (b) the ensemble median (defined as the median of the change values for the selected subset of models at each grid cell), (c) the bias of the ensemble median, and (d) the ensemble spread (calculated as the standard deviation of the change across the selected subset of models).\n\n**ANNUAL aggregation**\n\nChange default colorbars\nShared colorbar for ERA5 and ensemble median\nGet spei change\n\n
\n
\n

Fig 1. \n Spatial patterns of SPEI6 change over the Mediterranean region. Panels show (a) ERA5 reference, (b) CORDEX ensemble median (median of the selected subset of models at each grid cell), (c) ensemble median bias relative to ERA5, and (d) ensemble spread (standard deviation across the selected models). SPEI6 is standardised over the reference period 1971–2010. SPEI6 change is calculated at each grid point as the difference between the monthly climatologies of the target period (1991–2010) and the baseline period (1971–1990), and then averaged to obtain the ANNUAL mean.\n\n

\n\n\n
\nNOTE:
\nThe regridding artefacts visible in panel (d) result from the use of a conservative regridding method, which is commonly recommended when dealing with precipitation. They typically arise from large values in specific models and differences among the models’ native grids.\n\n(climate_projections-cordex_validation_q07:section-3.3)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q07__650f220eba02", "report_id": "climate_projections-cordex-domains-single-levels_validation_q07", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q07", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CORDEX biases in the SPEI6 drought index over the Mediterranean region > Analysis and results > 3. Plot and describe results > 3.3. Plot ensemble maps - Seasonal aggregations", "title": "CORDEX biases in the SPEI6 drought index over the Mediterranean region", "chunk_index": 16, "token_count": 839, "text_raw": "Same as 3.2 section but for: DJF, MAM, JJA and SON\n\n**WINTER (DJF)**\n\nGet spei change\n\n
\n
\n

Fig 2. \n Spatial patterns of SPEI6 change over the Mediterranean region. Panels show (a) ERA5 reference, (b) CORDEX ensemble median (median of the selected subset of models at each grid cell), (c) ensemble median bias relative to ERA5, and (d) ensemble spread (standard deviation across the selected models). SPEI6 is standardised over the reference period 1971–2010. SPEI6 change is calculated at each grid point as the difference between the monthly climatologies of the target period (1991–2010) and the baseline period (1971–1990), and then averaged across the winter to obtain the DJF mean.\n\n

\n\n**SPRING (MAM)**\n\nGet spei change\n\n
\n
\n

Fig 3. \n Spatial patterns of SPEI6 change over the Mediterranean region. Panels show (a) ERA5 reference, (b) CORDEX ensemble median (median of the selected subset of models at each grid cell), (c) ensemble median bias relative to ERA5, and (d) ensemble spread (standard deviation across the selected models). SPEI6 is standardised over the reference period 1971–2010. SPEI6 change is calculated at each grid point as the difference between the monthly climatologies of the target period (1991–2010) and the baseline period (1971–1990), and then averaged across the spring to obtain the MAM mean.\n\n

\n\n**SUMMER (JJA)**\n\nGet spei change\n\n
\n
\n

Fig 4. \n Spatial patterns of SPEI6 change over the Mediterranean region. Panels show (a) ERA5 reference, (b) CORDEX ensemble median (median of the selected subset of models at each grid cell), (c) ensemble median bias relative to ERA5, and (d) ensemble spread (standard deviation across the selected models). SPEI6 is standardised over the reference period 1971–2010. SPEI6 change is calculated at each grid point as the difference between the monthly climatologies of the target period (1991–2010) and the baseline period (1971–1990), and then averaged across the summer to obtain the JJA mean.\n\n

\n\n**AUTUMN (SON)**\n\nGet spei change\n\n
\n
\n

Fig 5. \n Spatial patterns of SPEI6 change over the Mediterranean region. Panels show (a) ERA5 reference, (b) CORDEX ensemble median (median of the selected subset of models at each grid cell), (c) ensemble median bias relative to ERA5, and (d) ensemble spread (standard deviation across the selected models). SPEI6 is standardised over the reference period 1971–2010. SPEI6 change is calculated at each grid point as the difference between the monthly climatologies of the target period (1991–2010) and the baseline period (1971–1990), and then averaged across the autumn to obtain the SON mean.\n\n

\n\n(climate_projections-cordex_validation_q07:section-3.4)=", "text_with_prefix": "EQC Quality Assessment: \"CORDEX biases in the SPEI6 drought index over the Mediterranean region\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q07 | Category: Climate_Projections\nSection: CORDEX biases in the SPEI6 drought index over the Mediterranean region > Analysis and results > 3. Plot and describe results > 3.3. Plot ensemble maps - Seasonal aggregations\n---\nSame as 3.2 section but for: DJF, MAM, JJA and SON\n\n**WINTER (DJF)**\n\nGet spei change\n\n
\n
\n

Fig 2. \n Spatial patterns of SPEI6 change over the Mediterranean region. Panels show (a) ERA5 reference, (b) CORDEX ensemble median (median of the selected subset of models at each grid cell), (c) ensemble median bias relative to ERA5, and (d) ensemble spread (standard deviation across the selected models). SPEI6 is standardised over the reference period 1971–2010. SPEI6 change is calculated at each grid point as the difference between the monthly climatologies of the target period (1991–2010) and the baseline period (1971–1990), and then averaged across the winter to obtain the DJF mean.\n\n

\n\n**SPRING (MAM)**\n\nGet spei change\n\n
\n
\n

Fig 3. \n Spatial patterns of SPEI6 change over the Mediterranean region. Panels show (a) ERA5 reference, (b) CORDEX ensemble median (median of the selected subset of models at each grid cell), (c) ensemble median bias relative to ERA5, and (d) ensemble spread (standard deviation across the selected models). SPEI6 is standardised over the reference period 1971–2010. SPEI6 change is calculated at each grid point as the difference between the monthly climatologies of the target period (1991–2010) and the baseline period (1971–1990), and then averaged across the spring to obtain the MAM mean.\n\n

\n\n**SUMMER (JJA)**\n\nGet spei change\n\n
\n
\n

Fig 4. \n Spatial patterns of SPEI6 change over the Mediterranean region. Panels show (a) ERA5 reference, (b) CORDEX ensemble median (median of the selected subset of models at each grid cell), (c) ensemble median bias relative to ERA5, and (d) ensemble spread (standard deviation across the selected models). SPEI6 is standardised over the reference period 1971–2010. SPEI6 change is calculated at each grid point as the difference between the monthly climatologies of the target period (1991–2010) and the baseline period (1971–1990), and then averaged across the summer to obtain the JJA mean.\n\n

\n\n**AUTUMN (SON)**\n\nGet spei change\n\n
\n
\n

Fig 5. \n Spatial patterns of SPEI6 change over the Mediterranean region. Panels show (a) ERA5 reference, (b) CORDEX ensemble median (median of the selected subset of models at each grid cell), (c) ensemble median bias relative to ERA5, and (d) ensemble spread (standard deviation across the selected models). SPEI6 is standardised over the reference period 1971–2010. SPEI6 change is calculated at each grid point as the difference between the monthly climatologies of the target period (1991–2010) and the baseline period (1971–1990), and then averaged across the autumn to obtain the SON mean.\n\n

\n\n(climate_projections-cordex_validation_q07:section-3.4)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q07__1254fd9d498b", "report_id": "climate_projections-cordex-domains-single-levels_validation_q07", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q07", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CORDEX biases in the SPEI6 drought index over the Mediterranean region > Analysis and results > 3. Plot and describe results > 3.4. Plot model maps", "title": "CORDEX biases in the SPEI6 drought index over the Mediterranean region", "chunk_index": 17, "token_count": 275, "text_raw": "In this section, we invoke the `plot_models()` function to visualise the SPEI6 change of every model individually. Note that the model data used in this section maintains its original grid. Only the annual mean is shown here.\n\nReduce horizontal spacing between columns\n\n
\n
\n

Fig 6.\n Spatial patterns of SPEI6 change for each individual CORDEX model over the Mediterranean region. SPEI6 is standardised over the reference period 1971–2010. For each model, SPEI6 change is calculated at each grid point as the difference between the monthly climatologies of the target period (1991–2010) and the baseline period (1971–1990), and then averaged to obtain the annual mean. \n

\n\n(climate_projections-cordex_validation_q07:section-3.5)=", "text_with_prefix": "EQC Quality Assessment: \"CORDEX biases in the SPEI6 drought index over the Mediterranean region\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q07 | Category: Climate_Projections\nSection: CORDEX biases in the SPEI6 drought index over the Mediterranean region > Analysis and results > 3. Plot and describe results > 3.4. Plot model maps\n---\nIn this section, we invoke the `plot_models()` function to visualise the SPEI6 change of every model individually. Note that the model data used in this section maintains its original grid. Only the annual mean is shown here.\n\nReduce horizontal spacing between columns\n\n
\n
\n

Fig 6.\n Spatial patterns of SPEI6 change for each individual CORDEX model over the Mediterranean region. SPEI6 is standardised over the reference period 1971–2010. For each model, SPEI6 change is calculated at each grid point as the difference between the monthly climatologies of the target period (1991–2010) and the baseline period (1971–1990), and then averaged to obtain the annual mean. \n

\n\n(climate_projections-cordex_validation_q07:section-3.5)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q07__06eb0ba74752", "report_id": "climate_projections-cordex-domains-single-levels_validation_q07", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q07", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CORDEX biases in the SPEI6 drought index over the Mediterranean region > Analysis and results > 3. Plot and describe results > 3.5. Plot bias maps", "title": "CORDEX biases in the SPEI6 drought index over the Mediterranean region", "chunk_index": 18, "token_count": 309, "text_raw": "In this section, we invoke the `plot_models()` function to visualise the bias of the SPEI6 change for every model individually. Note that the model data used in this section has previously been interpolated to the ERA5 grid. Regridding is performed for each essential climate variable before the index calculation to preserve standardisation. Only the annual mean is shown here.\n\nPlot using the shared bias parameters\n\n
\n
\n

Fig 7.\n Bias of SPEI6 change for each individual CORDEX model relative to ERA5 over the Mediterranean region. SPEI6 is standardised over the reference period 1971–2010. For each model and ERA5, SPEI6 change is calculated at each grid point as the difference between the monthly climatologies of the target period (1991–2010) and the baseline period (1971–1990), and then averaged to obtain the annual mean. \n

\n\n(climate_projections-cordex_validation_q07:section-3.6)=", "text_with_prefix": "EQC Quality Assessment: \"CORDEX biases in the SPEI6 drought index over the Mediterranean region\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q07 | Category: Climate_Projections\nSection: CORDEX biases in the SPEI6 drought index over the Mediterranean region > Analysis and results > 3. Plot and describe results > 3.5. Plot bias maps\n---\nIn this section, we invoke the `plot_models()` function to visualise the bias of the SPEI6 change for every model individually. Note that the model data used in this section has previously been interpolated to the ERA5 grid. Regridding is performed for each essential climate variable before the index calculation to preserve standardisation. Only the annual mean is shown here.\n\nPlot using the shared bias parameters\n\n
\n
\n

Fig 7.\n Bias of SPEI6 change for each individual CORDEX model relative to ERA5 over the Mediterranean region. SPEI6 is standardised over the reference period 1971–2010. For each model and ERA5, SPEI6 change is calculated at each grid point as the difference between the monthly climatologies of the target period (1991–2010) and the baseline period (1971–1990), and then averaged to obtain the annual mean. \n

\n\n(climate_projections-cordex_validation_q07:section-3.6)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q07__8fe10f02bdf1", "report_id": "climate_projections-cordex-domains-single-levels_validation_q07", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q07", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CORDEX biases in the SPEI6 drought index over the Mediterranean region > Analysis and results > 3. Plot and describe results > 3.6. Timeseries", "title": "CORDEX biases in the SPEI6 drought index over the Mediterranean region", "chunk_index": 19, "token_count": 490, "text_raw": "This section shows the timeseries of SPEI6 anomalies relative to the baseline climatology. We first use `compute_change_4timeseries()`, which calculates SPEI6 anomalies for each month of the timeseries (relative to the baseline monthly climatologies) and aggregates them according to the `time_agg` parameter. Spatially averaged values are then obtained. Light blue shading highlights the baseline period, while light orange shading marks the target period. The analysis compares the CORDEX ensemble median, ERA5, and individual models, displaying trends for both ERA5 and the ensemble median. Additionally, it shows the ensemble spread and the slope values of the trends for ERA5 and the CORDEX ensemble median. Only the annual aggregation is shown here.\n\nDefine colors\nCutout region\nCompute the change/anomaly for ERA5\nRegionalise and compute spatially weighted mean\nCompute change/anomaly for each model and spatially aggregate\nEnsemble statistics\nInitialize plot\nShade baseline period (light blue)\nShade target period (light orange)\n--- Create custom legend entries with full opacity ---\nPlot individual model time series\nAdd legend entry for individual models\nPlot CORDEX median\nFill ±1 std (spread)\nCompute trends\nLabels and grid\nFormat x-axis as years\nAnnotate trends\n\n
\n
\n

Fig 8.\n Timeseries of SPEI6 anomalies over the Mediterranean region. Monthly anomalies are calculated relative to the monthly baseline climatologies, aggregated according to the specified timescale, and then spatially averaged over the region. Light blue shading indicates the baseline period, and light orange shading indicates the target period. The figure compares ERA5, the CORDEX ensemble median, and individual models, showing trends for ERA5 and the ensemble median, as well as the ensemble spread and slope values of the trends.\n

\n\n(climate_projections-cordex_validation_q07:section-3.7)=", "text_with_prefix": "EQC Quality Assessment: \"CORDEX biases in the SPEI6 drought index over the Mediterranean region\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q07 | Category: Climate_Projections\nSection: CORDEX biases in the SPEI6 drought index over the Mediterranean region > Analysis and results > 3. Plot and describe results > 3.6. Timeseries\n---\nThis section shows the timeseries of SPEI6 anomalies relative to the baseline climatology. We first use `compute_change_4timeseries()`, which calculates SPEI6 anomalies for each month of the timeseries (relative to the baseline monthly climatologies) and aggregates them according to the `time_agg` parameter. Spatially averaged values are then obtained. Light blue shading highlights the baseline period, while light orange shading marks the target period. The analysis compares the CORDEX ensemble median, ERA5, and individual models, displaying trends for both ERA5 and the ensemble median. Additionally, it shows the ensemble spread and the slope values of the trends for ERA5 and the CORDEX ensemble median. Only the annual aggregation is shown here.\n\nDefine colors\nCutout region\nCompute the change/anomaly for ERA5\nRegionalise and compute spatially weighted mean\nCompute change/anomaly for each model and spatially aggregate\nEnsemble statistics\nInitialize plot\nShade baseline period (light blue)\nShade target period (light orange)\n--- Create custom legend entries with full opacity ---\nPlot individual model time series\nAdd legend entry for individual models\nPlot CORDEX median\nFill ±1 std (spread)\nCompute trends\nLabels and grid\nFormat x-axis as years\nAnnotate trends\n\n
\n
\n

Fig 8.\n Timeseries of SPEI6 anomalies over the Mediterranean region. Monthly anomalies are calculated relative to the monthly baseline climatologies, aggregated according to the specified timescale, and then spatially averaged over the region. Light blue shading indicates the baseline period, and light orange shading indicates the target period. The figure compares ERA5, the CORDEX ensemble median, and individual models, showing trends for ERA5 and the ensemble median, as well as the ensemble spread and slope values of the trends.\n

\n\n(climate_projections-cordex_validation_q07:section-3.7)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q07__1561360d773e", "report_id": "climate_projections-cordex-domains-single-levels_validation_q07", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q07", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CORDEX biases in the SPEI6 drought index over the Mediterranean region > Analysis and results > 3. Plot and describe results > 3.7. Results summary", "title": "CORDEX biases in the SPEI6 drought index over the Mediterranean region", "chunk_index": 20, "token_count": 740, "text_raw": "- ERA5 indicates a general intensification of drought conditions over the Mediterranean region when comparing the baseline (1971–1990) and target (1991–2010) periods.\n\n- The CORDEX ensemble median (obtained from the subset of models considered within this notebook) of the SPEI6 change is generally close to zero (Fig. 1). However, the inter-model spread is high, and is particularly pronounced over the Iberian Peninsula, southern France, North Africa, the Balkans region, and the easternmost Mediterranean areas (including Turkey, Lebanon, and Syria). In fact, the spatial pattern of the SPEI6 change is pretty different depending on the considered model (Fig. 6 and 7).\n\n- Seasonal analysis (Figs. 2–5) indicates that, in ERA5, drought intensification is strongest in summer. Spring and autumn also show notable decreases in SPEI6, although localised increases partly offset the spatial mean, especially in spring. In winter, the sign of the SPEI6 change is region-dependent, with large decreases over the eastern Iberian Peninsula, the eastern Mediterranean, and North Africa, and increases over the western Iberian Peninsula, as well as parts of Italy and the Balkans. Considering the CORDEX ensemble median, SPEI6 changes remain close to zero for most seasons and regions, with no clear spatial patterns overall, or at least none consistent with those observed in ERA5.\n\n- The timeseries of spatial means over the Mediterranean confirms an increase in drought conditions during the considered historical period in ERA5, with a trend of –0.24 (Fig. 8). For the ensemble median of CORDEX, the timeseries shows a trend close to zero, which is also not statistically significant. The timeseries plot also shows a high inter-model spread. Timeseries for seasonal aggregations are not shown here, but have been computed offline and exhibit behaviour similar to the annual case.\n\n- The results obtained in this notebook differ from those reported in the quality assessment named [CMIP6 biases in the SPEI6 drought index over the Mediterranean region](https://ecmwf-projects.github.io/c3s2-eqc-quality-assessment/Climate_Projections/CMIP6/climate_projections-cmip6_validation_q13.html), where CMIP6 ensemble median captured the sign of the SPEI6 change for a similar historical period, although with an underestimation compared to ERA5.\n\n- The limited change in SPEI6 simulated by some CORDEX models, compared to ERA5 and the previous CMIP6-based counterpart assessment, may partly reflect methodological and structural factors. In particular, the shorter analysis periods used here (20-year blocks instead of 30) may reduce the robustness of the signal. In addition, many CORDEX RCMs do not include time-evolving aerosols, which can lead to an underestimation of temperature trends in some regions and consequently affect drought-related indices such as SPEI6 [[9]](https://doi.org/10.1038/s43247-024-01332-8).\n\n(climate_projections-cordex_validation_q07:section-3.8)=", "text_with_prefix": "EQC Quality Assessment: \"CORDEX biases in the SPEI6 drought index over the Mediterranean region\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q07 | Category: Climate_Projections\nSection: CORDEX biases in the SPEI6 drought index over the Mediterranean region > Analysis and results > 3. Plot and describe results > 3.7. Results summary\n---\n- ERA5 indicates a general intensification of drought conditions over the Mediterranean region when comparing the baseline (1971–1990) and target (1991–2010) periods.\n\n- The CORDEX ensemble median (obtained from the subset of models considered within this notebook) of the SPEI6 change is generally close to zero (Fig. 1). However, the inter-model spread is high, and is particularly pronounced over the Iberian Peninsula, southern France, North Africa, the Balkans region, and the easternmost Mediterranean areas (including Turkey, Lebanon, and Syria). In fact, the spatial pattern of the SPEI6 change is pretty different depending on the considered model (Fig. 6 and 7).\n\n- Seasonal analysis (Figs. 2–5) indicates that, in ERA5, drought intensification is strongest in summer. Spring and autumn also show notable decreases in SPEI6, although localised increases partly offset the spatial mean, especially in spring. In winter, the sign of the SPEI6 change is region-dependent, with large decreases over the eastern Iberian Peninsula, the eastern Mediterranean, and North Africa, and increases over the western Iberian Peninsula, as well as parts of Italy and the Balkans. Considering the CORDEX ensemble median, SPEI6 changes remain close to zero for most seasons and regions, with no clear spatial patterns overall, or at least none consistent with those observed in ERA5.\n\n- The timeseries of spatial means over the Mediterranean confirms an increase in drought conditions during the considered historical period in ERA5, with a trend of –0.24 (Fig. 8). For the ensemble median of CORDEX, the timeseries shows a trend close to zero, which is also not statistically significant. The timeseries plot also shows a high inter-model spread. Timeseries for seasonal aggregations are not shown here, but have been computed offline and exhibit behaviour similar to the annual case.\n\n- The results obtained in this notebook differ from those reported in the quality assessment named [CMIP6 biases in the SPEI6 drought index over the Mediterranean region](https://ecmwf-projects.github.io/c3s2-eqc-quality-assessment/Climate_Projections/CMIP6/climate_projections-cmip6_validation_q13.html), where CMIP6 ensemble median captured the sign of the SPEI6 change for a similar historical period, although with an underestimation compared to ERA5.\n\n- The limited change in SPEI6 simulated by some CORDEX models, compared to ERA5 and the previous CMIP6-based counterpart assessment, may partly reflect methodological and structural factors. In particular, the shorter analysis periods used here (20-year blocks instead of 30) may reduce the robustness of the signal. In addition, many CORDEX RCMs do not include time-evolving aerosols, which can lead to an underestimation of temperature trends in some regions and consequently affect drought-related indices such as SPEI6 [[9]](https://doi.org/10.1038/s43247-024-01332-8).\n\n(climate_projections-cordex_validation_q07:section-3.8)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q07__ffac4e2f9927", "report_id": "climate_projections-cordex-domains-single-levels_validation_q07", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q07", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CORDEX biases in the SPEI6 drought index over the Mediterranean region > Analysis and results > 3. Plot and describe results > 3.8. Implications for the users", "title": "CORDEX biases in the SPEI6 drought index over the Mediterranean region", "chunk_index": 21, "token_count": 458, "text_raw": "- The pronounced inter-model spread over North Africa, Turkey, and the eastern Mediterranean implies that users in these regions may face higher future uncertainty. This highlights the need to stress-test adaptation strategies against a wide range of plausible futures rather than relying on a single scenario.\n\n- Although the ensemble median does not capture the SPEI6 change signal for the historical period, this does not necessarily imply that the models are not suitable for investigating future changes, as the spatial pattern of change depends on the individual model considered and model biases are not necessarily stationary, but may evolve under changing climate conditions [[10]](https://doi.org/10.5194/hess-25-273-2021).\n\n- The approach used in this notebook illustrates how a subset of GCM–RCM combinations can be selected while maintaining representation of all driving GCMs. This can be useful in situations where computational or practical constraints limit the use of the full ensemble. However, the results also reinforce that broader ensembles provide a more robust basis for assessing uncertainty.\n\n- Finally, users should be aware that results may vary depending on the choice of historical periods and model configurations. In this context, complementary tools such as the [C3S Atlas](https://confluence.ecmwf.int/display/CKB/Gridded+data+underpinning+the+Copernicus+Interactive+Climate+Atlas%3A+Description+of+the+datasets+and+variables#GriddeddataunderpinningtheCopernicusInteractiveClimateAtlas:Descriptionofthedatasetsandvariables), which includes a larger ensemble and alternative period definitions, can provide additional context for interpreting these findings. In addition, this notebook helps users understand the basis of the SPEI index and its methodological assumptions, supporting its application in operational water management and drought preparedness.", "text_with_prefix": "EQC Quality Assessment: \"CORDEX biases in the SPEI6 drought index over the Mediterranean region\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q07 | Category: Climate_Projections\nSection: CORDEX biases in the SPEI6 drought index over the Mediterranean region > Analysis and results > 3. Plot and describe results > 3.8. Implications for the users\n---\n- The pronounced inter-model spread over North Africa, Turkey, and the eastern Mediterranean implies that users in these regions may face higher future uncertainty. This highlights the need to stress-test adaptation strategies against a wide range of plausible futures rather than relying on a single scenario.\n\n- Although the ensemble median does not capture the SPEI6 change signal for the historical period, this does not necessarily imply that the models are not suitable for investigating future changes, as the spatial pattern of change depends on the individual model considered and model biases are not necessarily stationary, but may evolve under changing climate conditions [[10]](https://doi.org/10.5194/hess-25-273-2021).\n\n- The approach used in this notebook illustrates how a subset of GCM–RCM combinations can be selected while maintaining representation of all driving GCMs. This can be useful in situations where computational or practical constraints limit the use of the full ensemble. However, the results also reinforce that broader ensembles provide a more robust basis for assessing uncertainty.\n\n- Finally, users should be aware that results may vary depending on the choice of historical periods and model configurations. In this context, complementary tools such as the [C3S Atlas](https://confluence.ecmwf.int/display/CKB/Gridded+data+underpinning+the+Copernicus+Interactive+Climate+Atlas%3A+Description+of+the+datasets+and+variables#GriddeddataunderpinningtheCopernicusInteractiveClimateAtlas:Descriptionofthedatasetsandvariables), which includes a larger ensemble and alternative period definitions, can provide additional context for interpreting these findings. In addition, this notebook helps users understand the basis of the SPEI index and its methodological assumptions, supporting its application in operational water management and drought preparedness."} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q07__fc07b0fef2dd", "report_id": "climate_projections-cordex-domains-single-levels_validation_q07", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q07", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CORDEX biases in the SPEI6 drought index over the Mediterranean region > ℹ️ If you want to know more > Key resources", "title": "CORDEX biases in the SPEI6 drought index over the Mediterranean region", "chunk_index": 22, "token_count": 296, "text_raw": "Some key resources and further reading were linked throughout this assessment.\n\nThe CDS catalogue entries for the data used were:\n* CORDEX regional climate model data on single levels (daily - Maximum 2m temperature in the last 24 hours, Minimum 2m temperature in the last 24 hours, 2m air temperature and Mean precipitation flux): https://cds.climate.copernicus.eu/datasets/projections-cordex-domains-single-levels?tab=overview\n* ERA5 hourly data on single levels from 1940 to present (2m temperature and total precipitation): https://cds.climate.copernicus.eu/datasets/reanalysis-era5-pressure-levels-monthly-means?tab=overview\n\nCode libraries used:\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)", "text_with_prefix": "EQC Quality Assessment: \"CORDEX biases in the SPEI6 drought index over the Mediterranean region\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q07 | Category: Climate_Projections\nSection: CORDEX biases in the SPEI6 drought index over the Mediterranean region > ℹ️ If you want to know more > Key resources\n---\nSome key resources and further reading were linked throughout this assessment.\n\nThe CDS catalogue entries for the data used were:\n* CORDEX regional climate model data on single levels (daily - Maximum 2m temperature in the last 24 hours, Minimum 2m temperature in the last 24 hours, 2m air temperature and Mean precipitation flux): https://cds.climate.copernicus.eu/datasets/projections-cordex-domains-single-levels?tab=overview\n* ERA5 hourly data on single levels from 1940 to present (2m temperature and total precipitation): https://cds.climate.copernicus.eu/datasets/reanalysis-era5-pressure-levels-monthly-means?tab=overview\n\nCode libraries used:\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)"} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q07__a2b333889aef", "report_id": "climate_projections-cordex-domains-single-levels_validation_q07", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q07", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CORDEX biases in the SPEI6 drought index over the Mediterranean region > ℹ️ If you want to know more > References", "title": "CORDEX biases in the SPEI6 drought index over the Mediterranean region", "chunk_index": 23, "token_count": 1030, "text_raw": "[[1]](https://doi.org/10.1175/2009JCLI2909.1) Vicente-Serrano, S. M., Beguerı́a, S., and López-Moreno, J. I., 2010. A multiscalar drought index sensitive to global warming: The standardized precipitation evapotranspiration index, Journal of Climate, 23, 1696–1718. https://doi.org/10.1175/2009JCLI2909.1\n\n[[2]](https://digitalcommons.unl.edu/droughtfacpub/69/) Wilhite, D. A., 2000. Droughts as a natural hazard: concepts and definitions. In: DROUGHT, A Global Assessment, vol. I and II, Routledge Hazards and Disasters Series, Routledge.\n\n[[3]](https://doi.org/10.1175/2012EI000434.1) Vicente-Serrano, S. M., Beguerı́a, S., Lorenzo-Lacruz, J., Camarero, J. J., López-Moreno, J. I., Azorin-Molina, C., Revuelto, J., Morán-Tejeda, E., and Sanchez-Lorenzo, A., 2012. Performance of drought indices for ecological, agricultural, and hydrological applications, Earth Interactions, 16, https://doi.org/10.1175/2012EI000434.1\n\n[[4]](https://doi.org/10.3390/eesp2025035029) Kalisoras, A., Georgoulias, A. K., Akritidis, D., and Zanis, P., 2025. Future Projections in Agricultural Drought Characteristics for Greece Under Different Climate Change Scenarios. Environmental and Earth Sciences Proceedings, 35(1), 29. https://doi.org/10.3390/eesp2025035029\n\n[[5]](https://doi.org/10.1002/joc.7302) Spinoni, J., Barbosa, P., Bucchignani, E., Cassano, J., Cavazos, T., Cescatti, A., Christensen, J. H., Christensen, O. B., Coppola, E., Evans, J. P., Forzieri, G., Geyer, B., Giorgi, F., Jacob, D., Katzfey, J., Koenigk, T., Laprise, R., Lennard, C. J., Kurnaz, M. L., … Dosio, A., 2021. Global exposure of population and land-use to meteorological droughts under different warming levels and SSPs: A CORDEX-based study. International Journal of Climatology, 41(15), 6825–6853. https://doi.org/10.1002/joc.7302\n\n[[6]](https://doi.org/10.5194/essd-12-2959-2020) Iturbide, M., Gutiérrez, J. M., Alves, L. M., Bedia, J., Cerezo-Mota, R., Cimadevilla, E., Cofiño, A. S., Di Luca, A., Faria, S. H., Gorodetskaya, I. V., Hauser, M., Herrera, S., Hennessy, K., Hewitt, H. T., Jones, R. G., Krakovska, S., Manzanas, R., Martínez-Castro, D., Narisma, G. T., Nurhati, I. S., Pinto, I., Seneviratne, S. I., van den Hurk, B., and Vera, C. S., 2020. An update of IPCC climate reference regions for subcontinental analysis of climate model data: definition and aggregated datasets, Earth Syst. Sci. Data, 12, 2959–2970, https://doi.org/10.5194/essd-12-2959-2020\n\n[[7]](https://doi.org/10.13031/2013.26773) Hargreaves, G. H., and Samani, Z. A., 1985. Reference crop evapotranspiration from temperature, Applied engineering in agriculture, pp. 96–99. https://doi.org/10.13031/2013.26773", "text_with_prefix": "EQC Quality Assessment: \"CORDEX biases in the SPEI6 drought index over the Mediterranean region\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q07 | Category: Climate_Projections\nSection: CORDEX biases in the SPEI6 drought index over the Mediterranean region > ℹ️ If you want to know more > References\n---\n[[1]](https://doi.org/10.1175/2009JCLI2909.1) Vicente-Serrano, S. M., Beguerı́a, S., and López-Moreno, J. I., 2010. A multiscalar drought index sensitive to global warming: The standardized precipitation evapotranspiration index, Journal of Climate, 23, 1696–1718. https://doi.org/10.1175/2009JCLI2909.1\n\n[[2]](https://digitalcommons.unl.edu/droughtfacpub/69/) Wilhite, D. A., 2000. Droughts as a natural hazard: concepts and definitions. In: DROUGHT, A Global Assessment, vol. I and II, Routledge Hazards and Disasters Series, Routledge.\n\n[[3]](https://doi.org/10.1175/2012EI000434.1) Vicente-Serrano, S. M., Beguerı́a, S., Lorenzo-Lacruz, J., Camarero, J. J., López-Moreno, J. I., Azorin-Molina, C., Revuelto, J., Morán-Tejeda, E., and Sanchez-Lorenzo, A., 2012. Performance of drought indices for ecological, agricultural, and hydrological applications, Earth Interactions, 16, https://doi.org/10.1175/2012EI000434.1\n\n[[4]](https://doi.org/10.3390/eesp2025035029) Kalisoras, A., Georgoulias, A. K., Akritidis, D., and Zanis, P., 2025. Future Projections in Agricultural Drought Characteristics for Greece Under Different Climate Change Scenarios. Environmental and Earth Sciences Proceedings, 35(1), 29. https://doi.org/10.3390/eesp2025035029\n\n[[5]](https://doi.org/10.1002/joc.7302) Spinoni, J., Barbosa, P., Bucchignani, E., Cassano, J., Cavazos, T., Cescatti, A., Christensen, J. H., Christensen, O. B., Coppola, E., Evans, J. P., Forzieri, G., Geyer, B., Giorgi, F., Jacob, D., Katzfey, J., Koenigk, T., Laprise, R., Lennard, C. J., Kurnaz, M. L., … Dosio, A., 2021. Global exposure of population and land-use to meteorological droughts under different warming levels and SSPs: A CORDEX-based study. International Journal of Climatology, 41(15), 6825–6853. https://doi.org/10.1002/joc.7302\n\n[[6]](https://doi.org/10.5194/essd-12-2959-2020) Iturbide, M., Gutiérrez, J. M., Alves, L. M., Bedia, J., Cerezo-Mota, R., Cimadevilla, E., Cofiño, A. S., Di Luca, A., Faria, S. H., Gorodetskaya, I. V., Hauser, M., Herrera, S., Hennessy, K., Hewitt, H. T., Jones, R. G., Krakovska, S., Manzanas, R., Martínez-Castro, D., Narisma, G. T., Nurhati, I. S., Pinto, I., Seneviratne, S. I., van den Hurk, B., and Vera, C. S., 2020. An update of IPCC climate reference regions for subcontinental analysis of climate model data: definition and aggregated datasets, Earth Syst. Sci. Data, 12, 2959–2970, https://doi.org/10.5194/essd-12-2959-2020\n\n[[7]](https://doi.org/10.13031/2013.26773) Hargreaves, G. H., and Samani, Z. A., 1985. Reference crop evapotranspiration from temperature, Applied engineering in agriculture, pp. 96–99. https://doi.org/10.13031/2013.26773"} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q07__903d82539c8f", "report_id": "climate_projections-cordex-domains-single-levels_validation_q07", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q07", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "CORDEX biases in the SPEI6 drought index over the Mediterranean region > ℹ️ If you want to know more > References", "title": "CORDEX biases in the SPEI6 drought index over the Mediterranean region", "chunk_index": 24, "token_count": 474, "text_raw": "ani, Z. A., 1985. Reference crop evapotranspiration from temperature, Applied engineering in agriculture, pp. 96–99. https://doi.org/10.13031/2013.26773\n\n[[8]](https://doi.org/10.1002/joc.3887) Beguerı́a, S., Vicente-Serrano, S. M., Reig, F., and Latorre, B., 2014. Standardized precipitation evapotranspiration index (spei) revisited: Parameter fitting, evapotranspiration models, tools, datasets and drought monitoring, International Journal of Climatology, 34, 3001–3023. https://doi.org/10.1002/joc.3887\n\n[[9]](https://doi.org/10.1038/s43247-024-01332-8) Schumacher, D.L., Singh, J., Hauser, M., et al., 2024. Exacerbated summer European warming not captured by climate models neglecting long-term aerosol changes. Commun Earth Environ 5, 182. https://doi.org/10.1038/s43247-024-01332-8\n\n[[10]](https://doi.org/10.5194/hess-25-273-2021) Schmith, T., Thejll, P., Berg, P., Boberg, F., Christensen, O. B., Christiansen, B., Christensen, J. H., Madsen, M. S., and Steger, C., 2021. Identifying robust bias adjustment methods for European extreme precipitation in a multi-model pseudo-reality setting, Hydrol. Earth Syst. Sci., 25, 273–290, https://doi.org/10.5194/hess-25-273-2021", "text_with_prefix": "EQC Quality Assessment: \"CORDEX biases in the SPEI6 drought index over the Mediterranean region\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q07 | Category: Climate_Projections\nSection: CORDEX biases in the SPEI6 drought index over the Mediterranean region > ℹ️ If you want to know more > References\n---\nani, Z. A., 1985. Reference crop evapotranspiration from temperature, Applied engineering in agriculture, pp. 96–99. https://doi.org/10.13031/2013.26773\n\n[[8]](https://doi.org/10.1002/joc.3887) Beguerı́a, S., Vicente-Serrano, S. M., Reig, F., and Latorre, B., 2014. Standardized precipitation evapotranspiration index (spei) revisited: Parameter fitting, evapotranspiration models, tools, datasets and drought monitoring, International Journal of Climatology, 34, 3001–3023. https://doi.org/10.1002/joc.3887\n\n[[9]](https://doi.org/10.1038/s43247-024-01332-8) Schumacher, D.L., Singh, J., Hauser, M., et al., 2024. Exacerbated summer European warming not captured by climate models neglecting long-term aerosol changes. Commun Earth Environ 5, 182. https://doi.org/10.1038/s43247-024-01332-8\n\n[[10]](https://doi.org/10.5194/hess-25-273-2021) Schmith, T., Thejll, P., Berg, P., Boberg, F., Christensen, O. B., Christiansen, B., Christensen, J. H., Madsen, M. S., and Steger, C., 2021. Identifying robust bias adjustment methods for European extreme precipitation in a multi-model pseudo-reality setting, Hydrol. Earth Syst. Sci., 25, 273–290, https://doi.org/10.5194/hess-25-273-2021"} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q08__3db05370bb41", "report_id": "climate_projections-cordex-domains-single-levels_validation_q08", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q08", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region", "title": "Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region", "chunk_index": 0, "token_count": 114, "text_raw": "Production date: 25-03-2026\n\nProduced by: CMCC foundation - Euro-Mediterranean Center on Climate Change. Albert Martinez Boti and Lorenzo Sangelantoni.", "text_with_prefix": "EQC Quality Assessment: \"Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q08 | Category: Climate_Projections\nSection: Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region\n---\nProduction date: 25-03-2026\n\nProduced by: CMCC foundation - Euro-Mediterranean Center on Climate Change. Albert Martinez Boti and Lorenzo Sangelantoni."} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q08__c6c80d6f58ce", "report_id": "climate_projections-cordex-domains-single-levels_validation_q08", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q08", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region > Quality assessment question", "title": "Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region", "chunk_index": 1, "token_count": 575, "text_raw": "* **What are the projected future changes and associated uncertainties in droughts over the Mediterranean?**\n\nThe Standardized Precipitation-Evapotranspiration Index (SPEI) [[1]](https://doi.org/10.1175/2009JCLI2909.1) is a widely used drought indicator that integrates precipitation and potential evapotranspiration, providing a more complete measure of water deficits than precipitation alone. Its multiscalar formulation allows the evaluation of different types of drought [[2]](https://digitalcommons.unl.edu/droughtfacpub/69/), from short-term agricultural or soil moisture deficits to long-term hydrological or ecological events. Being standardised, SPEI is dimensionless, facilitating comparisons across regions and climates. By capturing variations in frequency, duration, and severity, it effectively reflects the complex nature of drought, making it a robust and physically meaningful proxy for drought assessment in the context of climate change [[3]](https://doi.org/10.1175/2012EI000434.1).\n\nUnderstanding SPEI is particularly relevant for the water management sector. By capturing multiple types of drought SPEI serves as a versatile indicator for various users. Climate projections using SPEI enable planners to anticipate future water deficits, guide allocation, and design effective strategies for reservoir management, irrigation, and drought preparedness, helping to reduce socio-economic and ecological impacts [[4]](https://doi.org/10.3390/eesp2025035029)[[5]](https://doi.org/10.1002/joc.7302).\n\nThis assessment uses data from a subset of models from [CORDEX](https://cds.climate.copernicus.eu/datasets/projections-cordex-domains-single-levels?tab=overview) Regional Climate Models (RCMs) to explore the signal and the uncertainty in future projections of the SPEI6 index. The analysis focuses on the Mediterranean region, evaluating changes in SPEI6 for the near-future period (2026–2055; under the RCP8.5 scenario) relative to the 1971–2000 baseline, which serves as the reference period for standardisation. The projected uncertainty is assessed by considering the ensemble inter-model spread of projected changes.\n\nThis notebook complements the evaluation of the biases of the same subset of CORDEX models performed through the assessment named *\"CORDEX biases in the SPEI6 drought index over the Mediterranean region\"*", "text_with_prefix": "EQC Quality Assessment: \"Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q08 | Category: Climate_Projections\nSection: Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region > Quality assessment question\n---\n* **What are the projected future changes and associated uncertainties in droughts over the Mediterranean?**\n\nThe Standardized Precipitation-Evapotranspiration Index (SPEI) [[1]](https://doi.org/10.1175/2009JCLI2909.1) is a widely used drought indicator that integrates precipitation and potential evapotranspiration, providing a more complete measure of water deficits than precipitation alone. Its multiscalar formulation allows the evaluation of different types of drought [[2]](https://digitalcommons.unl.edu/droughtfacpub/69/), from short-term agricultural or soil moisture deficits to long-term hydrological or ecological events. Being standardised, SPEI is dimensionless, facilitating comparisons across regions and climates. By capturing variations in frequency, duration, and severity, it effectively reflects the complex nature of drought, making it a robust and physically meaningful proxy for drought assessment in the context of climate change [[3]](https://doi.org/10.1175/2012EI000434.1).\n\nUnderstanding SPEI is particularly relevant for the water management sector. By capturing multiple types of drought SPEI serves as a versatile indicator for various users. Climate projections using SPEI enable planners to anticipate future water deficits, guide allocation, and design effective strategies for reservoir management, irrigation, and drought preparedness, helping to reduce socio-economic and ecological impacts [[4]](https://doi.org/10.3390/eesp2025035029)[[5]](https://doi.org/10.1002/joc.7302).\n\nThis assessment uses data from a subset of models from [CORDEX](https://cds.climate.copernicus.eu/datasets/projections-cordex-domains-single-levels?tab=overview) Regional Climate Models (RCMs) to explore the signal and the uncertainty in future projections of the SPEI6 index. The analysis focuses on the Mediterranean region, evaluating changes in SPEI6 for the near-future period (2026–2055; under the RCP8.5 scenario) relative to the 1971–2000 baseline, which serves as the reference period for standardisation. The projected uncertainty is assessed by considering the ensemble inter-model spread of projected changes.\n\nThis notebook complements the evaluation of the biases of the same subset of CORDEX models performed through the assessment named *\"CORDEX biases in the SPEI6 drought index over the Mediterranean region\"*"} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q08__1baf0b00ef75", "report_id": "climate_projections-cordex-domains-single-levels_validation_q08", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q08", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region > Quality assessment statement", "title": "Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region", "chunk_index": 2, "token_count": 426, "text_raw": "These are the key outcomes of this assessment\n\n* The CORDEX ensemble median for the subset of models considered in this assessment indicates an overall increase in projected drought conditions for the near-term future (2026–2055) under the RCP8.5 scenario.\n\n* There is substantial inter-model spread, with differences in both the magnitude and, in some cases, the sign of projected SPEI6 changes, particularly at regional and seasonal scales, indicating that projections are highly model-dependent and should be interpreted in an ensemble-based framework.\n\n* Despite this spread, SPEI6 anomalies remain predominantly negative at the Mediterranean scale, indicating a consistent basin-wide tendency towards drier conditions.\n\n* The projected intensification of drought may be conservative, as indicated by a complementary historical assessment showing that the analysed CORDEX models tend to underestimate drought conditions relative to ERA5.\n```\n\nattachment:88b80b6a-fa7c-45f6-bdb0-51e2689e6529.png\n---\nwidth: 1500px\nalt: Visual abstract \n---\nSpatial patterns of SPEI6 change over the Mediterranean region. Panels show (a) CORDEX ensemble median (median of the selected subset of models at each grid cell) and (b) ensemble spread (standard deviation across the selected models). SPEI6 is standardised over the reference period 1971–2000. SPEI6 change is calculated at each grid point as the difference between the monthly climatologies of the target period (2026–2055) and the baseline period (1971–2000), and then averaged to obtain the ANNUAL mean. Future projections used for the target period follow the RCP8.5 scenario.\n```", "text_with_prefix": "EQC Quality Assessment: \"Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q08 | Category: Climate_Projections\nSection: Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region > Quality assessment statement\n---\nThese are the key outcomes of this assessment\n\n* The CORDEX ensemble median for the subset of models considered in this assessment indicates an overall increase in projected drought conditions for the near-term future (2026–2055) under the RCP8.5 scenario.\n\n* There is substantial inter-model spread, with differences in both the magnitude and, in some cases, the sign of projected SPEI6 changes, particularly at regional and seasonal scales, indicating that projections are highly model-dependent and should be interpreted in an ensemble-based framework.\n\n* Despite this spread, SPEI6 anomalies remain predominantly negative at the Mediterranean scale, indicating a consistent basin-wide tendency towards drier conditions.\n\n* The projected intensification of drought may be conservative, as indicated by a complementary historical assessment showing that the analysed CORDEX models tend to underestimate drought conditions relative to ERA5.\n```\n\nattachment:88b80b6a-fa7c-45f6-bdb0-51e2689e6529.png\n---\nwidth: 1500px\nalt: Visual abstract \n---\nSpatial patterns of SPEI6 change over the Mediterranean region. Panels show (a) CORDEX ensemble median (median of the selected subset of models at each grid cell) and (b) ensemble spread (standard deviation across the selected models). SPEI6 is standardised over the reference period 1971–2000. SPEI6 change is calculated at each grid point as the difference between the monthly climatologies of the target period (2026–2055) and the baseline period (1971–2000), and then averaged to obtain the ANNUAL mean. Future projections used for the target period follow the RCP8.5 scenario.\n```"} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q08__ef9c051ab875", "report_id": "climate_projections-cordex-domains-single-levels_validation_q08", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q08", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region > Methodology", "title": "Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region", "chunk_index": 3, "token_count": 1067, "text_raw": "The Standardized Precipitation-Evapotranspiration Index accumulated to six months (SPEI6) has been calculated for this assessment — through the use of [xclim](https://xclim.readthedocs.io/en/stable/) — as a proxy to evaluate projected drought conditions for a subset of [CORDEX](https://cds.climate.copernicus.eu/datasets/projections-cordex-domains-single-levels?tab=overview) models. The study covers the Mediterranean domain, following IPCC-AR6 regional definitions [[6]](https://doi.org/10.5194/essd-12-2959-2020). Results include spatial maps of SPEI6 changes for the near-future period (2026–2055) relative to the 1971–2000 baseline, which serves as the reference period for standardisation. Additionally, time series are computed to show the temporal evolution of future SPEI6 changes. Uncertainty is assessed through the inter-model ensemble spread, defined as the standard deviation across the selected subset of models, of projected SPEI6 changes.\n\nPotential evapotranspiration (PET) is calculated with the Hargreaves method [[7]](https://doi.org/10.13031/2013.26773), which requires only maximum and minimum temperature (and optionally mean temperature), providing a computationally efficient yet robust estimate (e.g., [[8]](https://doi.org/10.1002/joc.3887)). The accumulated series of precipitation minus PET is then standardised over the 1971–2000 reference period.\n\nFive key concepts are defined:\n\n- **Reference period:** The period used to standardise SPEI values, here 1971–2000. \n- **Baseline period:** The period against which changes are calculated. In this notebook the baseline period is selected to be the same as the reference period. \n- **Target period:** The future period for which changes are evaluated, here 2026–2055. The Representative Concentration Pathway is set to the RCP8.5 scenario for this assessment. \n- **SPEI6 change:** Differences in SPEI6 between the target period and the baseline, representing anomalies in drought conditions. Since, in this notebook, the baseline and the reference period are the same, these changes directly represent anomalies relative to the standardisation climatology used to compute SPEI6.\n\nThe analysis and results follow the next outline:\n\n**[](climate_projections-cordex_validation_q08:section-1)**\n * [](climate_projections-cordex_validation_q08:section-1.1)\n * [](climate_projections-cordex_validation_q08:section-1.2)\n * [](climate_projections-cordex_validation_q08:section-1.3)\n * [](climate_projections-cordex_validation_q08:section-1.4)\n * [](climate_projections-cordex_validation_q08:section-1.5)\n * [](climate_projections-cordex_validation_q08:section-1.6)\n\n**[](climate_projections-cordex_validation_q08:section-2)**\n * [](climate_projections-cordex_validation_q08:section-2.1)\n * [](climate_projections-cordex_validation_q08:section-2.2)\n * [](climate_projections-cordex_validation_q08:section-2.3)\n\n**[](climate_projections-cordex_validation_q08:section-3)**\n * [](climate_projections-cordex_validation_q08:section-3.1)\n * [](climate_projections-cordex_validation_q08:section-3.2)\n * [](climate_projections-cordex_validation_q08:section-3.3)\n * [](climate_projections-cordex_validation_q08:section-3.4)\n * [](climate_projections-cordex_validation_q08:section-3.5)\n * [](climate_projections-cordex_validation_q08:section-3.6)\n * [](climate_projections-cordex_validation_q08:section-3.7)\n\n\n
\nNOTE ON REFERENCE PERIOD SELECTION:
\nThis notebook follows the same methodology as the assessment ‘CORDEX biases in the SPEI6 drought index over the Mediterranean region’. However, the reference period (1971–2000) is shorter than the 1971–2010 period used in the historical assessment. This choice avoids including scenario data in the standardisation (as CORDEX historical simulations extend only until 2005), ensures computational efficiency due to the shorter period, and maintains consistency with the 30-year length of the future target period.\n
", "text_with_prefix": "EQC Quality Assessment: \"Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q08 | Category: Climate_Projections\nSection: Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region > Methodology\n---\nThe Standardized Precipitation-Evapotranspiration Index accumulated to six months (SPEI6) has been calculated for this assessment — through the use of [xclim](https://xclim.readthedocs.io/en/stable/) — as a proxy to evaluate projected drought conditions for a subset of [CORDEX](https://cds.climate.copernicus.eu/datasets/projections-cordex-domains-single-levels?tab=overview) models. The study covers the Mediterranean domain, following IPCC-AR6 regional definitions [[6]](https://doi.org/10.5194/essd-12-2959-2020). Results include spatial maps of SPEI6 changes for the near-future period (2026–2055) relative to the 1971–2000 baseline, which serves as the reference period for standardisation. Additionally, time series are computed to show the temporal evolution of future SPEI6 changes. Uncertainty is assessed through the inter-model ensemble spread, defined as the standard deviation across the selected subset of models, of projected SPEI6 changes.\n\nPotential evapotranspiration (PET) is calculated with the Hargreaves method [[7]](https://doi.org/10.13031/2013.26773), which requires only maximum and minimum temperature (and optionally mean temperature), providing a computationally efficient yet robust estimate (e.g., [[8]](https://doi.org/10.1002/joc.3887)). The accumulated series of precipitation minus PET is then standardised over the 1971–2000 reference period.\n\nFive key concepts are defined:\n\n- **Reference period:** The period used to standardise SPEI values, here 1971–2000. \n- **Baseline period:** The period against which changes are calculated. In this notebook the baseline period is selected to be the same as the reference period. \n- **Target period:** The future period for which changes are evaluated, here 2026–2055. The Representative Concentration Pathway is set to the RCP8.5 scenario for this assessment. \n- **SPEI6 change:** Differences in SPEI6 between the target period and the baseline, representing anomalies in drought conditions. Since, in this notebook, the baseline and the reference period are the same, these changes directly represent anomalies relative to the standardisation climatology used to compute SPEI6.\n\nThe analysis and results follow the next outline:\n\n**[](climate_projections-cordex_validation_q08:section-1)**\n * [](climate_projections-cordex_validation_q08:section-1.1)\n * [](climate_projections-cordex_validation_q08:section-1.2)\n * [](climate_projections-cordex_validation_q08:section-1.3)\n * [](climate_projections-cordex_validation_q08:section-1.4)\n * [](climate_projections-cordex_validation_q08:section-1.5)\n * [](climate_projections-cordex_validation_q08:section-1.6)\n\n**[](climate_projections-cordex_validation_q08:section-2)**\n * [](climate_projections-cordex_validation_q08:section-2.1)\n * [](climate_projections-cordex_validation_q08:section-2.2)\n * [](climate_projections-cordex_validation_q08:section-2.3)\n\n**[](climate_projections-cordex_validation_q08:section-3)**\n * [](climate_projections-cordex_validation_q08:section-3.1)\n * [](climate_projections-cordex_validation_q08:section-3.2)\n * [](climate_projections-cordex_validation_q08:section-3.3)\n * [](climate_projections-cordex_validation_q08:section-3.4)\n * [](climate_projections-cordex_validation_q08:section-3.5)\n * [](climate_projections-cordex_validation_q08:section-3.6)\n * [](climate_projections-cordex_validation_q08:section-3.7)\n\n\n
\nNOTE ON REFERENCE PERIOD SELECTION:
\nThis notebook follows the same methodology as the assessment ‘CORDEX biases in the SPEI6 drought index over the Mediterranean region’. However, the reference period (1971–2000) is shorter than the 1971–2010 period used in the historical assessment. This choice avoids including scenario data in the standardisation (as CORDEX historical simulations extend only until 2005), ensures computational efficiency due to the shorter period, and maintains consistency with the 30-year length of the future target period.\n
"} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q08__5500a6fb6df6", "report_id": "climate_projections-cordex-domains-single-levels_validation_q08", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q08", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region > Analysis and results > 1. Parameters, requests and functions definition > 1.2. Define Parameters", "title": "Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region", "chunk_index": 4, "token_count": 612, "text_raw": "In the \"Define Parameters\" section, various customisable options for the notebook are specified:\n\n- The reference and baseline periods are set to be the same in this notebook and can be adjusted by modifying `reference_slice`; the default range is 1971-2000.\n- The target period can be adjusted by modifying `target_slice`; the default range is 2026-2055.\n- A check is performed on the `target_slice` and `reference_slice` to ensure consistency with CDS CORDEX requests and to verify that the reference period falls within the historical experiment, i.e. the last year of `reference_slice` does not exceed 2005.\n- `collection_id` is not customisable for this assessment and is set to 'CORDEX'.\n- The `area` parameter specifies the geographical domain of interest. By default, the Mediterranean domain is used, following IPCC-AR6 regional definitions [[6]](https://doi.org/10.5194/essd-12-2959-2020)\n- The `time_agg` allows selecting the time aggregation for the analysis. Options include: annual, DJF, MAM, JJA, SON, Jan, Feb, Mar, Apr, May, Jun, Jul, Aug, Sep, Oct, Nov, Dec.\n- SPEI accumulation window (`window`): Defines the accumulation period in months for the SPEI calculation. Common options include 3, 6, or 12 months, depending on whether short-term or long-term droughts are of interest. In this assessment, a six-month accumulation is used.\n- `rcm_model_regrid` and `gcm_driven_model_regrid` specify the EURO-CORDEX regional climate model and the driving global climate model, respectively, whose grid will be used as the target for remapping. Note that both models must be included in the matrix of RCM-GCM combinations defined by the `models_cordex` dictionary in Section 1.3.\n- The `chunks` selection allows the user to define if dividing into chunks when downloading the data on their local machine. Although it does not significantly affect the analysis, it is recommended to keep the default value for optimal performance.\n\nPeriods definition\nCollection id\nDefine region for analysis\ntime aggregation\nSPEI accumulation\nChunks for download\n\nRule: start ends in 1 or 6, end ends in 0 or 5\nHistorical constraint for baseline\n\n(climate_projections-cordex_validation_q08:section-1.3)=", "text_with_prefix": "EQC Quality Assessment: \"Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q08 | Category: Climate_Projections\nSection: Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region > Analysis and results > 1. Parameters, requests and functions definition > 1.2. Define Parameters\n---\nIn the \"Define Parameters\" section, various customisable options for the notebook are specified:\n\n- The reference and baseline periods are set to be the same in this notebook and can be adjusted by modifying `reference_slice`; the default range is 1971-2000.\n- The target period can be adjusted by modifying `target_slice`; the default range is 2026-2055.\n- A check is performed on the `target_slice` and `reference_slice` to ensure consistency with CDS CORDEX requests and to verify that the reference period falls within the historical experiment, i.e. the last year of `reference_slice` does not exceed 2005.\n- `collection_id` is not customisable for this assessment and is set to 'CORDEX'.\n- The `area` parameter specifies the geographical domain of interest. By default, the Mediterranean domain is used, following IPCC-AR6 regional definitions [[6]](https://doi.org/10.5194/essd-12-2959-2020)\n- The `time_agg` allows selecting the time aggregation for the analysis. Options include: annual, DJF, MAM, JJA, SON, Jan, Feb, Mar, Apr, May, Jun, Jul, Aug, Sep, Oct, Nov, Dec.\n- SPEI accumulation window (`window`): Defines the accumulation period in months for the SPEI calculation. Common options include 3, 6, or 12 months, depending on whether short-term or long-term droughts are of interest. In this assessment, a six-month accumulation is used.\n- `rcm_model_regrid` and `gcm_driven_model_regrid` specify the EURO-CORDEX regional climate model and the driving global climate model, respectively, whose grid will be used as the target for remapping. Note that both models must be included in the matrix of RCM-GCM combinations defined by the `models_cordex` dictionary in Section 1.3.\n- The `chunks` selection allows the user to define if dividing into chunks when downloading the data on their local machine. Although it does not significantly affect the analysis, it is recommended to keep the default value for optimal performance.\n\nPeriods definition\nCollection id\nDefine region for analysis\ntime aggregation\nSPEI accumulation\nChunks for download\n\nRule: start ends in 1 or 6, end ends in 0 or 5\nHistorical constraint for baseline\n\n(climate_projections-cordex_validation_q08:section-1.3)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q08__4161a6c3283e", "report_id": "climate_projections-cordex-domains-single-levels_validation_q08", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q08", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region > Analysis and results > 1. Parameters, requests and functions definition > 1.3. Define models", "title": "Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region", "chunk_index": 5, "token_count": 258, "text_raw": "The following climate analyses are performed using a subset of CORDEX models that provide both the historical and RCP8.5 experiments, as well as maximum, minimum and mean temperature and precipitation, within the Climate Data Store ([CDS](https://cds.climate.copernicus.eu/datasets/projections-cordex-domains-single-levels?tab=overview)). All Global Climate Models (GCMs) driving the selected CORDEX simulations were included, and the selection ensures a representative coverage of the available Regional Climate Models (RCMs). The selected models are the same as those used for the historical evaluation in the assessment titled *“CORDEX biases in the SPEI6 drought index over the Mediterranean region\"*.\n\n(climate_projections-cordex_validation_q08:section-1.4)=", "text_with_prefix": "EQC Quality Assessment: \"Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q08 | Category: Climate_Projections\nSection: Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region > Analysis and results > 1. Parameters, requests and functions definition > 1.3. Define models\n---\nThe following climate analyses are performed using a subset of CORDEX models that provide both the historical and RCP8.5 experiments, as well as maximum, minimum and mean temperature and precipitation, within the Climate Data Store ([CDS](https://cds.climate.copernicus.eu/datasets/projections-cordex-domains-single-levels?tab=overview)). All Global Climate Models (GCMs) driving the selected CORDEX simulations were included, and the selection ensures a representative coverage of the available Regional Climate Models (RCMs). The selected models are the same as those used for the historical evaluation in the assessment titled *“CORDEX biases in the SPEI6 drought index over the Mediterranean region\"*.\n\n(climate_projections-cordex_validation_q08:section-1.4)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q08__811314f1fdef", "report_id": "climate_projections-cordex-domains-single-levels_validation_q08", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q08", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region > Analysis and results > 1. Parameters, requests and functions definition > 1.4. Define land-sea mask and ERA5 precipitation requests", "title": "Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region", "chunk_index": 6, "token_count": 200, "text_raw": "Within this notebook, ERA5 will be used to download the land-sea mask when plotting. In this section, we set the required parameters for the cds-api data-request of ERA5 land-sea mask. A request for monthly ERA5 precipitation data is also defined to compute a bare-earth (barrean) mask, which needs to be applied to ensure meaningful results in arid regions.\n\n(climate_projections-cordex_validation_q08:section-1.5)=", "text_with_prefix": "EQC Quality Assessment: \"Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q08 | Category: Climate_Projections\nSection: Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region > Analysis and results > 1. Parameters, requests and functions definition > 1.4. Define land-sea mask and ERA5 precipitation requests\n---\nWithin this notebook, ERA5 will be used to download the land-sea mask when plotting. In this section, we set the required parameters for the cds-api data-request of ERA5 land-sea mask. A request for monthly ERA5 precipitation data is also defined to compute a bare-earth (barrean) mask, which needs to be applied to ensure meaningful results in arid regions.\n\n(climate_projections-cordex_validation_q08:section-1.5)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q08__53d69443d506", "report_id": "climate_projections-cordex-domains-single-levels_validation_q08", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q08", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region > Analysis and results > 1. Parameters, requests and functions definition > 1.5. Define model requests", "title": "Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region", "chunk_index": 7, "token_count": 152, "text_raw": "In this section we set the required parameters for the cds-api data-request.\n\nHistorical part (until 2005)\nFuture part (RCP8.5 from 2006 onwards)\n\n(climate_projections-cordex_validation_q08:section-1.6)=", "text_with_prefix": "EQC Quality Assessment: \"Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q08 | Category: Climate_Projections\nSection: Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region > Analysis and results > 1. Parameters, requests and functions definition > 1.5. Define model requests\n---\nIn this section we set the required parameters for the cds-api data-request.\n\nHistorical part (until 2005)\nFuture part (RCP8.5 from 2006 onwards)\n\n(climate_projections-cordex_validation_q08:section-1.6)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q08__5f2b40c867e8", "report_id": "climate_projections-cordex-domains-single-levels_validation_q08", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q08", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region > Analysis and results > 1. Parameters, requests and functions definition > 1.6. Functions to cache", "title": "Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region", "chunk_index": 8, "token_count": 308, "text_raw": "In this section, functions that will be executed in the caching phase are defined. Caching is the process of storing copies of files in a temporary storage location, so that they can be accessed more quickly. This process also checks if the user has already downloaded a file, avoiding redundant downloads.\n\nDescription of the main functions:\n\n- **`compute_spei_cordex_fut`**: Uses the xclim package to calculate the Standardized Precipitation-Evapotranspiration Index (SPEI). The `reference_slice` argument specifies the reference period used for standardization, while the `window` argument defines the number of months over which precipitation and evapotranspiration are accumulated (6 months in this assessment).\n\n- **`get_mask_pr`**: Generates a mask identifying grid points where precipitation is below 0.3 mm/day, allowing filtering of extremely dry regions.\n\nEnsure CF-compliant attributes for model datasets\nOriginal bounds for conservative interpolation\nWrap into dataset\n\n(climate_projections-cordex_validation_q08:section-2)=", "text_with_prefix": "EQC Quality Assessment: \"Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q08 | Category: Climate_Projections\nSection: Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region > Analysis and results > 1. Parameters, requests and functions definition > 1.6. Functions to cache\n---\nIn this section, functions that will be executed in the caching phase are defined. Caching is the process of storing copies of files in a temporary storage location, so that they can be accessed more quickly. This process also checks if the user has already downloaded a file, avoiding redundant downloads.\n\nDescription of the main functions:\n\n- **`compute_spei_cordex_fut`**: Uses the xclim package to calculate the Standardized Precipitation-Evapotranspiration Index (SPEI). The `reference_slice` argument specifies the reference period used for standardization, while the `window` argument defines the number of months over which precipitation and evapotranspiration are accumulated (6 months in this assessment).\n\n- **`get_mask_pr`**: Generates a mask identifying grid points where precipitation is below 0.3 mm/day, allowing filtering of extremely dry regions.\n\nEnsure CF-compliant attributes for model datasets\nOriginal bounds for conservative interpolation\nWrap into dataset\n\n(climate_projections-cordex_validation_q08:section-2)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q08__cf35fa753c1e", "report_id": "climate_projections-cordex-domains-single-levels_validation_q08", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q08", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region > Analysis and results > 2. Downloading and processing > 2.1. Download and transform the regridding model", "title": "Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region", "chunk_index": 9, "token_count": 362, "text_raw": "In this section, the `download.download_and_transform` function from the `c3s_eqc_automatic_quality_control` package is employed to download daily data from the selected EURO-CORDEX regridding model, select the subregion of interest, compute daily potential evapotranspiration, calculate the daily water balance, and derive the SPEI (SPEI6 for this assessment). Results are cached to avoid redundant downloads and repeated processing.\n\nThe regridding model here refers to the model whose grid is used as the target grid for remapping the other models. This ensures all datasets share a common grid, facilitating direct comparison at each grid cell. Within this notebook, the regridding model is set to `\"ichec_ec_earth_dmi_hirham5\"` but it can be changed by modifying the `gcm_driven_model_regrid` and `rcm_model_regrid` parameters in Section 1.2. Note that both models must be included in the matrix of RCM-GCM combinations defined by the `models_cordex` dictionary in Section 1.3. It is important to highlight that the choice of the target grid can impact the analysis depending on the specific application.\n\n(climate_projections-cordex_validation_q08:section-2.2)=", "text_with_prefix": "EQC Quality Assessment: \"Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q08 | Category: Climate_Projections\nSection: Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region > Analysis and results > 2. Downloading and processing > 2.1. Download and transform the regridding model\n---\nIn this section, the `download.download_and_transform` function from the `c3s_eqc_automatic_quality_control` package is employed to download daily data from the selected EURO-CORDEX regridding model, select the subregion of interest, compute daily potential evapotranspiration, calculate the daily water balance, and derive the SPEI (SPEI6 for this assessment). Results are cached to avoid redundant downloads and repeated processing.\n\nThe regridding model here refers to the model whose grid is used as the target grid for remapping the other models. This ensures all datasets share a common grid, facilitating direct comparison at each grid cell. Within this notebook, the regridding model is set to `\"ichec_ec_earth_dmi_hirham5\"` but it can be changed by modifying the `gcm_driven_model_regrid` and `rcm_model_regrid` parameters in Section 1.2. Note that both models must be included in the matrix of RCM-GCM combinations defined by the `models_cordex` dictionary in Section 1.3. It is important to highlight that the choice of the target grid can impact the analysis depending on the specific application.\n\n(climate_projections-cordex_validation_q08:section-2.2)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q08__930d5914d91a", "report_id": "climate_projections-cordex-domains-single-levels_validation_q08", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q08", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region > Analysis and results > 2. Downloading and processing > 2.2. Download and transform models", "title": "Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region", "chunk_index": 10, "token_count": 553, "text_raw": "In this section, the `download.download_and_transform` function from the 'c3s_eqc_automatic_quality_control' package is used to download daily data from CORDEX models, compute daily potential evapotranspiration, calculate the daily water balance, and derive the SPEI (SPEI6 for this assessment). SPEI6 is computed on each model's native grid and also on the `\"ichec_ec_earth_dmi_hirham5\"` grid to enable bias assessment. When regridding is required, it is performed for each essential climate variable before the index calculation to preserve standardisation. Results are cached to avoid redundant downloads and repeated processing.\n\nOriginal model\nInterpolated model\n\n```text\nmodel='cccma_canesm2_clmcom_clm_cclm4_8_17'\nmodel='cccma_canesm2_gerics_remo2015'\nmodel='cnrm_cerfacs_cm5_ipsl_wrf381p'\nmodel='cnrm_cerfacs_cm5_smhi_rca4'\nmodel='ichec_ec_earth_dmi_hirham5'\nmodel='ichec_ec_earth_knmi_racmo22e'\nmodel='ipsl_cm5a_mr_ipsl_wrf381p'\nmodel='ipsl_cm5a_mr_knmi_racmo22e'\nmodel='miroc_miroc5_clmcom_clm_cclm4_8_17'\nmodel='miroc_miroc5_gerics_remo2015'\nmodel='mohc_hadgem2_es_cnrm_aladin63'\nmodel='mohc_hadgem2_es_mohc_hadrem3_ga7_05'\nmodel='mpi_m_mpi_esm_lr_mpi_csc_remo2009'\nmodel='mpi_m_mpi_esm_lr_cnrm_aladin63'\nmodel='ncc_noresm1_m_dmi_hirham5'\nmodel='ncc_noresm1_m_mohc_hadrem3_ga7_05'\nmodel='ncc_noresm1_m_smhi_rca4'\n```\n\n(climate_projections-cordex_validation_q08:section-2.3)=", "text_with_prefix": "EQC Quality Assessment: \"Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q08 | Category: Climate_Projections\nSection: Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region > Analysis and results > 2. Downloading and processing > 2.2. Download and transform models\n---\nIn this section, the `download.download_and_transform` function from the 'c3s_eqc_automatic_quality_control' package is used to download daily data from CORDEX models, compute daily potential evapotranspiration, calculate the daily water balance, and derive the SPEI (SPEI6 for this assessment). SPEI6 is computed on each model's native grid and also on the `\"ichec_ec_earth_dmi_hirham5\"` grid to enable bias assessment. When regridding is required, it is performed for each essential climate variable before the index calculation to preserve standardisation. Results are cached to avoid redundant downloads and repeated processing.\n\nOriginal model\nInterpolated model\n\n```text\nmodel='cccma_canesm2_clmcom_clm_cclm4_8_17'\nmodel='cccma_canesm2_gerics_remo2015'\nmodel='cnrm_cerfacs_cm5_ipsl_wrf381p'\nmodel='cnrm_cerfacs_cm5_smhi_rca4'\nmodel='ichec_ec_earth_dmi_hirham5'\nmodel='ichec_ec_earth_knmi_racmo22e'\nmodel='ipsl_cm5a_mr_ipsl_wrf381p'\nmodel='ipsl_cm5a_mr_knmi_racmo22e'\nmodel='miroc_miroc5_clmcom_clm_cclm4_8_17'\nmodel='miroc_miroc5_gerics_remo2015'\nmodel='mohc_hadgem2_es_cnrm_aladin63'\nmodel='mohc_hadgem2_es_mohc_hadrem3_ga7_05'\nmodel='mpi_m_mpi_esm_lr_mpi_csc_remo2009'\nmodel='mpi_m_mpi_esm_lr_cnrm_aladin63'\nmodel='ncc_noresm1_m_dmi_hirham5'\nmodel='ncc_noresm1_m_mohc_hadrem3_ga7_05'\nmodel='ncc_noresm1_m_smhi_rca4'\n```\n\n(climate_projections-cordex_validation_q08:section-2.3)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q08__bab5c362daf7", "report_id": "climate_projections-cordex-domains-single-levels_validation_q08", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q08", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region > Analysis and results > 2. Downloading and processing > 2.3. Apply land-sea mask and change some attributes", "title": "Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region", "chunk_index": 11, "token_count": 444, "text_raw": "This section changes some attributes and applies a land–sea mask to all models. A bare-earth (barrean) mask is also applied to ensure meaningful results in arid regions, as in the [C3S atlas](https://atlas.climate.copernicus.eu/atlas). In particular, we follow the methodology available in their [Github repository](https://github.com/ecmwf-projects/c3s-atlas/blob/main/auxiliar/barrean_mask.ipynb). While their implementation of the mask also considers factors such as snow depth and vegetation cover, here we only apply the filter based on annual mean precipitation for 1971–2000 being less than 0.3 mm day⁻¹. This simplification is justified because, in the region under consideration, the bare-earth mask obtained with the full set of filters closely matches the one derived from the precipitation threshold alone.\n\n**Note:** `ds_interpolated` contains data from the models regridded to the `\"ichec_ec_earth_dmi_hirham5\"` grid. `model_datasets` contain the same data on the original grid of each model. Regridding is performed for every essential climate variable prior to the index calculation in order to avoid compromising standardisation.\n\nDownload and prepare lsm\nApply land-sea mask to interpolate dataset\nApply land sea mask to each model dataset\nApply PR mask to interpolated dataset\nApply PR mask to each model\n\n```text\n100%|██████████| 1/1 [00:00<00:00, 25.66it/s]\n```\n\n(climate_projections-cordex_validation_q08:section-3)=", "text_with_prefix": "EQC Quality Assessment: \"Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q08 | Category: Climate_Projections\nSection: Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region > Analysis and results > 2. Downloading and processing > 2.3. Apply land-sea mask and change some attributes\n---\nThis section changes some attributes and applies a land–sea mask to all models. A bare-earth (barrean) mask is also applied to ensure meaningful results in arid regions, as in the [C3S atlas](https://atlas.climate.copernicus.eu/atlas). In particular, we follow the methodology available in their [Github repository](https://github.com/ecmwf-projects/c3s-atlas/blob/main/auxiliar/barrean_mask.ipynb). While their implementation of the mask also considers factors such as snow depth and vegetation cover, here we only apply the filter based on annual mean precipitation for 1971–2000 being less than 0.3 mm day⁻¹. This simplification is justified because, in the region under consideration, the bare-earth mask obtained with the full set of filters closely matches the one derived from the precipitation threshold alone.\n\n**Note:** `ds_interpolated` contains data from the models regridded to the `\"ichec_ec_earth_dmi_hirham5\"` grid. `model_datasets` contain the same data on the original grid of each model. Regridding is performed for every essential climate variable prior to the index calculation in order to avoid compromising standardisation.\n\nDownload and prepare lsm\nApply land-sea mask to interpolate dataset\nApply land sea mask to each model dataset\nApply PR mask to interpolated dataset\nApply PR mask to each model\n\n```text\n100%|██████████| 1/1 [00:00<00:00, 25.66it/s]\n```\n\n(climate_projections-cordex_validation_q08:section-3)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q08__317e190aa571", "report_id": "climate_projections-cordex-domains-single-levels_validation_q08", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q08", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region > Analysis and results > 3. Plot and describe results", "title": "Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region", "chunk_index": 12, "token_count": 239, "text_raw": "This section will display the following results:\n\n- SPEI6 change maps. The layout includes the ensemble median (defined as the median of the change values across the selected subset of models at each grid cell) and the ensemble spread (calculated as the standard deviation of the change across the selected subset of models). \n- SPEI6 change maps for each individual model. \n- Time series of SPEI6 change (averaged over the Mediterranean region).\n\n**Note:** The projected SPEI6 change is calculated at each grid point as the arithmetic difference between climatologies in the target period (2026–2055) and the baseline period (1971–2000).\n\n(climate_projections-cordex_validation_q08:section-3.1)=", "text_with_prefix": "EQC Quality Assessment: \"Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q08 | Category: Climate_Projections\nSection: Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region > Analysis and results > 3. Plot and describe results\n---\nThis section will display the following results:\n\n- SPEI6 change maps. The layout includes the ensemble median (defined as the median of the change values across the selected subset of models at each grid cell) and the ensemble spread (calculated as the standard deviation of the change across the selected subset of models). \n- SPEI6 change maps for each individual model. \n- Time series of SPEI6 change (averaged over the Mediterranean region).\n\n**Note:** The projected SPEI6 change is calculated at each grid point as the arithmetic difference between climatologies in the target period (2026–2055) and the baseline period (1971–2000).\n\n(climate_projections-cordex_validation_q08:section-3.1)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q08__0f7ba819edf1", "report_id": "climate_projections-cordex-domains-single-levels_validation_q08", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q08", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region > Analysis and results > 3. Plot and describe results > 3.1. Define plotting functions", "title": "Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region", "chunk_index": 13, "token_count": 475, "text_raw": "The functions presented here are used to calculate SPEI6 changes and generate corresponding layout plots. Three types of layout can be displayed, depending on the plotting function:\n\n1. Reference and ensemble summary: Includes the ensemble median and the ensemble spread. This is generated using `plot_ensemble()`.\n\n2. Individual models: Displays all models individually using `plot_models()`.\n\n**Calculation of SPEI6 change:**\n\n- `compute_spei_change()` calculates SPEI6 change at each grid point as the arithmetic difference between monthly climatologies of the target period (2026–2055) and the baseline period (1971–2000). It then aggregates the change according to the user-specified `time_agg` in Section 1.2.\n\n- `compute_change_4timeseries()` computes SPEI6 anomalies relative to the baseline climatology for each month of the timeseries and aggregates them according to the `time_agg` parameter. This provides monthly, seasonal, or annual-level information depending on the user’s selection.\n\nMapping months to names\nMapping seasons to months\nSelect baseline and target periods\nMonthly climatologies\nAggregate according to user choice\nMapping months to names\nMapping seasons to months\nSelect baseline period\nMonthly climatology for the baseline\nCompute monthly anomalies relative to baseline climatology\nPreserve attributes\nAggregate anomalies\nGroup by year\nSelect months for the season, then group by year\nSelect the month, keep time as is\n--- Determine colour parameters ---\n--- Helper function to split GCM and RCM ---\n--- Plot each model ---\n--- Hide empty axes ---\n--- Shared extent and colourbar ---\nDefault behaviour\nCreate figure and axes\n--- ERA5 plot ---\n--- Ensemble Median ---\n--- Ensemble Bias (Median - ERA5) ---\n--- Ensemble Standard Deviation ---\n\n(climate_projections-cordex_validation_q08:section-3.2)=", "text_with_prefix": "EQC Quality Assessment: \"Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q08 | Category: Climate_Projections\nSection: Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region > Analysis and results > 3. Plot and describe results > 3.1. Define plotting functions\n---\nThe functions presented here are used to calculate SPEI6 changes and generate corresponding layout plots. Three types of layout can be displayed, depending on the plotting function:\n\n1. Reference and ensemble summary: Includes the ensemble median and the ensemble spread. This is generated using `plot_ensemble()`.\n\n2. Individual models: Displays all models individually using `plot_models()`.\n\n**Calculation of SPEI6 change:**\n\n- `compute_spei_change()` calculates SPEI6 change at each grid point as the arithmetic difference between monthly climatologies of the target period (2026–2055) and the baseline period (1971–2000). It then aggregates the change according to the user-specified `time_agg` in Section 1.2.\n\n- `compute_change_4timeseries()` computes SPEI6 anomalies relative to the baseline climatology for each month of the timeseries and aggregates them according to the `time_agg` parameter. This provides monthly, seasonal, or annual-level information depending on the user’s selection.\n\nMapping months to names\nMapping seasons to months\nSelect baseline and target periods\nMonthly climatologies\nAggregate according to user choice\nMapping months to names\nMapping seasons to months\nSelect baseline period\nMonthly climatology for the baseline\nCompute monthly anomalies relative to baseline climatology\nPreserve attributes\nAggregate anomalies\nGroup by year\nSelect months for the season, then group by year\nSelect the month, keep time as is\n--- Determine colour parameters ---\n--- Helper function to split GCM and RCM ---\n--- Plot each model ---\n--- Hide empty axes ---\n--- Shared extent and colourbar ---\nDefault behaviour\nCreate figure and axes\n--- ERA5 plot ---\n--- Ensemble Median ---\n--- Ensemble Bias (Median - ERA5) ---\n--- Ensemble Standard Deviation ---\n\n(climate_projections-cordex_validation_q08:section-3.2)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q08__e8821abee9d4", "report_id": "climate_projections-cordex-domains-single-levels_validation_q08", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q08", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region > Analysis and results > 3. Plot and describe results > 3.2. Plot ensemble maps", "title": "Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region", "chunk_index": 14, "token_count": 370, "text_raw": "In this section, we invoke the `plot_ensemble()` function to visualise the SPEI6 change for (a) the ensemble median (defined as the median of the change values for the selected subset of models at each grid cell) and (b) the ensemble spread (calculated as the standard deviation of the change across the selected subset of models).\n\n**ANNUAL aggregation**\n\nChange default colorbars\nShared colorbar for ERA5 and ensemble median\nGet spei change\n\n
\n
\n

Fig 1. \n Spatial patterns of SPEI6 change over the Mediterranean region. Panels show (a) CORDEX ensemble median (median of the selected subset of models at each grid cell) and (b) ensemble spread (standard deviation across the selected models). SPEI6 is standardised over the reference period 1971–2000. SPEI6 change is calculated at each grid point as the difference between the monthly climatologies of the target period (2026–2055) and the baseline period (1971–2000), and then averaged to obtain the ANNUAL mean. Future projections used for the target period follow the RCP8.5 scenario.\n\n

\n\n(climate_projections-cordex_validation_q08:section-3.3)=", "text_with_prefix": "EQC Quality Assessment: \"Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q08 | Category: Climate_Projections\nSection: Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region > Analysis and results > 3. Plot and describe results > 3.2. Plot ensemble maps\n---\nIn this section, we invoke the `plot_ensemble()` function to visualise the SPEI6 change for (a) the ensemble median (defined as the median of the change values for the selected subset of models at each grid cell) and (b) the ensemble spread (calculated as the standard deviation of the change across the selected subset of models).\n\n**ANNUAL aggregation**\n\nChange default colorbars\nShared colorbar for ERA5 and ensemble median\nGet spei change\n\n
\n
\n

Fig 1. \n Spatial patterns of SPEI6 change over the Mediterranean region. Panels show (a) CORDEX ensemble median (median of the selected subset of models at each grid cell) and (b) ensemble spread (standard deviation across the selected models). SPEI6 is standardised over the reference period 1971–2000. SPEI6 change is calculated at each grid point as the difference between the monthly climatologies of the target period (2026–2055) and the baseline period (1971–2000), and then averaged to obtain the ANNUAL mean. Future projections used for the target period follow the RCP8.5 scenario.\n\n

\n\n(climate_projections-cordex_validation_q08:section-3.3)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q08__3dabc2f2ab36", "report_id": "climate_projections-cordex-domains-single-levels_validation_q08", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q08", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region > Analysis and results > 3. Plot and describe results > 3.3. Plot ensemble maps - Seasonal aggregations", "title": "Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region", "chunk_index": 15, "token_count": 837, "text_raw": "Same as 3.2 section but for: DJF, MAM, JJA and SON\n\n**WINTER (DJF)**\n\nGet spei change\n\n
\n
\n

Fig 2. \n Spatial patterns of SPEI6 change over the Mediterranean region. Panels show (a) CORDEX ensemble median (median of the selected subset of models at each grid cell) and (b) ensemble spread (standard deviation across the selected models). SPEI6 is standardised over the reference period 1971–2000. SPEI6 change is calculated at each grid point as the difference between the monthly climatologies of the target period (2026–2055) and the baseline period (1971–2000), and then averaged across the winter to obtain the DJF mean. Future projections used for the target period follow the RCP8.5 scenario.\n\n

\n\n**SPRING (MAM)**\n\nGet spei change\n\n
\n
\n

Fig 3. \n Spatial patterns of SPEI6 change over the Mediterranean region. Panels show (a) CORDEX ensemble median (median of the selected subset of models at each grid cell) and (b) ensemble spread (standard deviation across the selected models). SPEI6 is standardised over the reference period 1971–2000. SPEI6 change is calculated at each grid point as the difference between the monthly climatologies of the target period (2026–2055) and the baseline period (1971–2000), and then averaged across the spring to obtain the MAM mean. Future projections used for the target period follow the RCP8.5 scenario.\n\n

\n\n**SUMMER (JJA)**\n\nGet spei change\n\n
\n
\n

Fig 4. \n Spatial patterns of SPEI6 change over the Mediterranean region. Panels show (a) CORDEX ensemble median (median of the selected subset of models at each grid cell) and (b) ensemble spread (standard deviation across the selected models). SPEI6 is standardised over the reference period 1971–2000. SPEI6 change is calculated at each grid point as the difference between the monthly climatologies of the target period (2026–2055) and the baseline period (1971–2000), and then averaged across the summer to obtain the JJA mean. Future projections used for the target period follow the RCP8.5 scenario.\n\n

\n\n**AUTUMN (SON)**\n\nGet spei change\n\n
\n
\n

Fig 5. \n Spatial patterns of SPEI6 change over the Mediterranean region. Panels show (a) CORDEX ensemble median (median of the selected subset of models at each grid cell) and (b) ensemble spread (standard deviation across the selected models). SPEI6 is standardised over the reference period 1971–2000. SPEI6 change is calculated at each grid point as the difference between the monthly climatologies of the target period (2026–2055) and the baseline period (1971–2000), and then averaged across the autumn to obtain the SON mean. Future projections used for the target period follow the RCP8.5 scenario.\n\n

\n\n(climate_projections-cordex_validation_q08:section-3.4)=", "text_with_prefix": "EQC Quality Assessment: \"Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q08 | Category: Climate_Projections\nSection: Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region > Analysis and results > 3. Plot and describe results > 3.3. Plot ensemble maps - Seasonal aggregations\n---\nSame as 3.2 section but for: DJF, MAM, JJA and SON\n\n**WINTER (DJF)**\n\nGet spei change\n\n
\n
\n

Fig 2. \n Spatial patterns of SPEI6 change over the Mediterranean region. Panels show (a) CORDEX ensemble median (median of the selected subset of models at each grid cell) and (b) ensemble spread (standard deviation across the selected models). SPEI6 is standardised over the reference period 1971–2000. SPEI6 change is calculated at each grid point as the difference between the monthly climatologies of the target period (2026–2055) and the baseline period (1971–2000), and then averaged across the winter to obtain the DJF mean. Future projections used for the target period follow the RCP8.5 scenario.\n\n

\n\n**SPRING (MAM)**\n\nGet spei change\n\n
\n
\n

Fig 3. \n Spatial patterns of SPEI6 change over the Mediterranean region. Panels show (a) CORDEX ensemble median (median of the selected subset of models at each grid cell) and (b) ensemble spread (standard deviation across the selected models). SPEI6 is standardised over the reference period 1971–2000. SPEI6 change is calculated at each grid point as the difference between the monthly climatologies of the target period (2026–2055) and the baseline period (1971–2000), and then averaged across the spring to obtain the MAM mean. Future projections used for the target period follow the RCP8.5 scenario.\n\n

\n\n**SUMMER (JJA)**\n\nGet spei change\n\n
\n
\n

Fig 4. \n Spatial patterns of SPEI6 change over the Mediterranean region. Panels show (a) CORDEX ensemble median (median of the selected subset of models at each grid cell) and (b) ensemble spread (standard deviation across the selected models). SPEI6 is standardised over the reference period 1971–2000. SPEI6 change is calculated at each grid point as the difference between the monthly climatologies of the target period (2026–2055) and the baseline period (1971–2000), and then averaged across the summer to obtain the JJA mean. Future projections used for the target period follow the RCP8.5 scenario.\n\n

\n\n**AUTUMN (SON)**\n\nGet spei change\n\n
\n
\n

Fig 5. \n Spatial patterns of SPEI6 change over the Mediterranean region. Panels show (a) CORDEX ensemble median (median of the selected subset of models at each grid cell) and (b) ensemble spread (standard deviation across the selected models). SPEI6 is standardised over the reference period 1971–2000. SPEI6 change is calculated at each grid point as the difference between the monthly climatologies of the target period (2026–2055) and the baseline period (1971–2000), and then averaged across the autumn to obtain the SON mean. Future projections used for the target period follow the RCP8.5 scenario.\n\n

\n\n(climate_projections-cordex_validation_q08:section-3.4)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q08__e1cbfc78b840", "report_id": "climate_projections-cordex-domains-single-levels_validation_q08", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q08", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region > Analysis and results > 3. Plot and describe results > 3.4. Plot model maps", "title": "Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region", "chunk_index": 16, "token_count": 297, "text_raw": "In this section, we invoke the `plot_models()` function to visualise the SPEI6 change of every model individually. Note that the model data used in this section maintains its original grid. Only the annual mean is shown here.\n\nReduce horizontal spacing between columns\n\n
\n
\n

Fig 6.\n Spatial patterns of SPEI6 change for each individual CORDEX model over the Mediterranean region. SPEI6 is standardised over the reference period 1971–2000. For each model, SPEI6 change is calculated at each grid point as the difference between the monthly climatologies of the target period (2026–2055) and the baseline period (1971–2000), and then averaged to obtain the annual mean. Future projections used for the target period follow the RCP8.5 scenario. \n

\n\n(climate_projections-cordex_validation_q08:section-3.5)=", "text_with_prefix": "EQC Quality Assessment: \"Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q08 | Category: Climate_Projections\nSection: Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region > Analysis and results > 3. Plot and describe results > 3.4. Plot model maps\n---\nIn this section, we invoke the `plot_models()` function to visualise the SPEI6 change of every model individually. Note that the model data used in this section maintains its original grid. Only the annual mean is shown here.\n\nReduce horizontal spacing between columns\n\n
\n
\n

Fig 6.\n Spatial patterns of SPEI6 change for each individual CORDEX model over the Mediterranean region. SPEI6 is standardised over the reference period 1971–2000. For each model, SPEI6 change is calculated at each grid point as the difference between the monthly climatologies of the target period (2026–2055) and the baseline period (1971–2000), and then averaged to obtain the annual mean. Future projections used for the target period follow the RCP8.5 scenario. \n

\n\n(climate_projections-cordex_validation_q08:section-3.5)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q08__5040904b4fe1", "report_id": "climate_projections-cordex-domains-single-levels_validation_q08", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q08", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region > Analysis and results > 3. Plot and describe results > 3.5. Timeseries", "title": "Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region", "chunk_index": 17, "token_count": 433, "text_raw": "This section shows the timeseries of SPEI6 anomalies relative to the reference period climatology (same period as the baseline for this notebook). We first use `compute_change_4timeseries()`, which calculates SPEI6 anomalies for each month of the target period (relative to the baseline monthly climatologies) and aggregates them according to the `time_agg` parameter. Spatially averaged values are then obtained. The analysis compares the CORDEX ensemble median and individual models and shows the ensemble spread and the slope value of the trend for the CORDEX ensemble median. Only the annual aggregation is shown here.\n\nDefine colors\nCutout region\nCompute change/anomaly for each model and spatially aggregate\nEnsemble statistics\n--- Select ONLY target period ---\nInitialize plot\nPlot individual model time series\nAdd legend entry for individual models\nPlot CORDEX median\nFill ±1 std (spread)\nCompute trend ONLY for CORDEX (already filtered to target period)\nLabels and title\nFormat x-axis as years\nAnnotate trends\n\n
\n
\n

Fig 7.\n Timeseries of SPEI6 anomalies over the Mediterranean region. Monthly anomalies are calculated relative to the monthly baseline climatologies, aggregated according to the specified timescale, and then spatially averaged over the region. The figure compares the CORDEX ensemble median and individual models, showing the trend for the ensemble median, as well as the ensemble spread and the corresponding trend slope. Future projections used for the target period follow the RCP8.5 scenario.\n

\n\n(climate_projections-cordex_validation_q08:section-3.6)=", "text_with_prefix": "EQC Quality Assessment: \"Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q08 | Category: Climate_Projections\nSection: Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region > Analysis and results > 3. Plot and describe results > 3.5. Timeseries\n---\nThis section shows the timeseries of SPEI6 anomalies relative to the reference period climatology (same period as the baseline for this notebook). We first use `compute_change_4timeseries()`, which calculates SPEI6 anomalies for each month of the target period (relative to the baseline monthly climatologies) and aggregates them according to the `time_agg` parameter. Spatially averaged values are then obtained. The analysis compares the CORDEX ensemble median and individual models and shows the ensemble spread and the slope value of the trend for the CORDEX ensemble median. Only the annual aggregation is shown here.\n\nDefine colors\nCutout region\nCompute change/anomaly for each model and spatially aggregate\nEnsemble statistics\n--- Select ONLY target period ---\nInitialize plot\nPlot individual model time series\nAdd legend entry for individual models\nPlot CORDEX median\nFill ±1 std (spread)\nCompute trend ONLY for CORDEX (already filtered to target period)\nLabels and title\nFormat x-axis as years\nAnnotate trends\n\n
\n
\n

Fig 7.\n Timeseries of SPEI6 anomalies over the Mediterranean region. Monthly anomalies are calculated relative to the monthly baseline climatologies, aggregated according to the specified timescale, and then spatially averaged over the region. The figure compares the CORDEX ensemble median and individual models, showing the trend for the ensemble median, as well as the ensemble spread and the corresponding trend slope. Future projections used for the target period follow the RCP8.5 scenario.\n

\n\n(climate_projections-cordex_validation_q08:section-3.6)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q08__4cc5a109a6b2", "report_id": "climate_projections-cordex-domains-single-levels_validation_q08", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q08", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region > Analysis and results > 3. Plot and describe results > 3.6. Results summary", "title": "Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region", "chunk_index": 18, "token_count": 420, "text_raw": "- The projected SPEI6 change for the near-term future (2026–2055) indicates an overall decrease in the index (Fig. 1), suggesting a general intensification of drought conditions. However, a large inter-model spread is evident, with substantial disparities not only in the magnitude of change but also in its sign (see Fig. 6). Some models project wetter conditions than the baseline, particularly over regions such as the Balkans, southern France, the Adriatic coast of Italy, and parts of Turkey.\n\n- The inter-model spread spatial patterns are sensitive to the seasonal aggregation considered (Figs. 2–5), with the largest inter-model discrepancies occurring in spring.\n\n- The timeseries of spatial means over the Mediterranean shows an increase in projected drought conditions for the near-term future period (Fig. 7). For the CORDEX ensemble median, the timeseries shows a slight (but statistically significant) negative trend over this period. Although the trend is weak and the inter-model spread is substantial, most individual models—and the ensemble median—remain consistently below zero. This indicates that drier conditions are projected for the annual aggregation over the region, as anomalies are defined relative to the historical baseline period.\n\n- This assessment is broadly consistent with results from the [interactive C3S Atlas](https://atlas.climate.copernicus.eu/atlas). The Atlas indicates slightly stronger decreases in SPEI6, which may be partly explained by differences in the standardisation periods used and the larger ensemble of models considered.\n\n(climate_projections-cordex_validation_q08:section-3.7)=", "text_with_prefix": "EQC Quality Assessment: \"Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q08 | Category: Climate_Projections\nSection: Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region > Analysis and results > 3. Plot and describe results > 3.6. Results summary\n---\n- The projected SPEI6 change for the near-term future (2026–2055) indicates an overall decrease in the index (Fig. 1), suggesting a general intensification of drought conditions. However, a large inter-model spread is evident, with substantial disparities not only in the magnitude of change but also in its sign (see Fig. 6). Some models project wetter conditions than the baseline, particularly over regions such as the Balkans, southern France, the Adriatic coast of Italy, and parts of Turkey.\n\n- The inter-model spread spatial patterns are sensitive to the seasonal aggregation considered (Figs. 2–5), with the largest inter-model discrepancies occurring in spring.\n\n- The timeseries of spatial means over the Mediterranean shows an increase in projected drought conditions for the near-term future period (Fig. 7). For the CORDEX ensemble median, the timeseries shows a slight (but statistically significant) negative trend over this period. Although the trend is weak and the inter-model spread is substantial, most individual models—and the ensemble median—remain consistently below zero. This indicates that drier conditions are projected for the annual aggregation over the region, as anomalies are defined relative to the historical baseline period.\n\n- This assessment is broadly consistent with results from the [interactive C3S Atlas](https://atlas.climate.copernicus.eu/atlas). The Atlas indicates slightly stronger decreases in SPEI6, which may be partly explained by differences in the standardisation periods used and the larger ensemble of models considered.\n\n(climate_projections-cordex_validation_q08:section-3.7)="} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q08__9385951de052", "report_id": "climate_projections-cordex-domains-single-levels_validation_q08", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q08", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region > Analysis and results > 3. Plot and describe results > 3.7. Implications for the users", "title": "Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region", "chunk_index": 19, "token_count": 306, "text_raw": "- The substantial uncertainty in projected drought conditions, particularly at regional and seasonal scales, supports the use of risk-based approaches in water management and drought preparedness, where strategies should be tested against a range of plausible futures rather than relying on a single model or ensemble summary.\n\n- The historical bias assessment (***CORDEX biases in the SPEI6 drought index over the Mediterranean region***) shows that the analysed CORDEX models tend to underestimate drought conditions relative to ERA5. Based on this, the projected intensification of drought in this notebook may be conservative and should be interpreted with this limitation in mind.\n\n- Complementary tools such as the [interactive C3S Atlas](https://atlas.climate.copernicus.eu/atlas), which include a larger ensemble and additional dimensions such as months, global warming levels, and multiple time horizons, can provide further context for interpreting these results.\n\n- Beyond the direct interpretation of results, this notebook emphasises the basis and methodological construction of the SPEI index, supporting its use in operational contexts.", "text_with_prefix": "EQC Quality Assessment: \"Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q08 | Category: Climate_Projections\nSection: Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region > Analysis and results > 3. Plot and describe results > 3.7. Implications for the users\n---\n- The substantial uncertainty in projected drought conditions, particularly at regional and seasonal scales, supports the use of risk-based approaches in water management and drought preparedness, where strategies should be tested against a range of plausible futures rather than relying on a single model or ensemble summary.\n\n- The historical bias assessment (***CORDEX biases in the SPEI6 drought index over the Mediterranean region***) shows that the analysed CORDEX models tend to underestimate drought conditions relative to ERA5. Based on this, the projected intensification of drought in this notebook may be conservative and should be interpreted with this limitation in mind.\n\n- Complementary tools such as the [interactive C3S Atlas](https://atlas.climate.copernicus.eu/atlas), which include a larger ensemble and additional dimensions such as months, global warming levels, and multiple time horizons, can provide further context for interpreting these results.\n\n- Beyond the direct interpretation of results, this notebook emphasises the basis and methodological construction of the SPEI index, supporting its use in operational contexts."} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q08__03aea013d6a1", "report_id": "climate_projections-cordex-domains-single-levels_validation_q08", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q08", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region > ℹ️ If you want to know more > Key resources", "title": "Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region", "chunk_index": 20, "token_count": 299, "text_raw": "Some key resources and further reading were linked throughout this assessment.\n\nThe CDS catalogue entries for the data used were:\n* CORDEX regional climate model data on single levels (daily - Maximum 2m temperature in the last 24 hours, Minimum 2m temperature in the last 24 hours, 2m air temperature and Mean precipitation flux): https://cds.climate.copernicus.eu/datasets/projections-cordex-domains-single-levels?tab=overview\n* ERA5 monthly averaged data on single levels from 1940 to present (total precipitation): https://cds.climate.copernicus.eu/datasets/reanalysis-era5-single-levels-monthly-means?tab=overview\n\nCode libraries used:\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)", "text_with_prefix": "EQC Quality Assessment: \"Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q08 | Category: Climate_Projections\nSection: Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region > ℹ️ If you want to know more > Key resources\n---\nSome key resources and further reading were linked throughout this assessment.\n\nThe CDS catalogue entries for the data used were:\n* CORDEX regional climate model data on single levels (daily - Maximum 2m temperature in the last 24 hours, Minimum 2m temperature in the last 24 hours, 2m air temperature and Mean precipitation flux): https://cds.climate.copernicus.eu/datasets/projections-cordex-domains-single-levels?tab=overview\n* ERA5 monthly averaged data on single levels from 1940 to present (total precipitation): https://cds.climate.copernicus.eu/datasets/reanalysis-era5-single-levels-monthly-means?tab=overview\n\nCode libraries used:\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)"} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q08__a2b333889aef", "report_id": "climate_projections-cordex-domains-single-levels_validation_q08", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q08", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region > ℹ️ If you want to know more > References", "title": "Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region", "chunk_index": 21, "token_count": 1036, "text_raw": "[[1]](https://doi.org/10.1175/2009JCLI2909.1) Vicente-Serrano, S. M., Beguerı́a, S., and López-Moreno, J. I., 2010. A multiscalar drought index sensitive to global warming: The standardized precipitation evapotranspiration index, Journal of Climate, 23, 1696–1718. https://doi.org/10.1175/2009JCLI2909.1\n\n[[2]](https://digitalcommons.unl.edu/droughtfacpub/69/) Wilhite, D. A., 2000. Droughts as a natural hazard: concepts and definitions. In: DROUGHT, A Global Assessment, vol. I and II, Routledge Hazards and Disasters Series, Routledge.\n\n[[3]](https://doi.org/10.1175/2012EI000434.1) Vicente-Serrano, S. M., Beguerı́a, S., Lorenzo-Lacruz, J., Camarero, J. J., López-Moreno, J. I., Azorin-Molina, C., Revuelto, J., Morán-Tejeda, E., and Sanchez-Lorenzo, A., 2012. Performance of drought indices for ecological, agricultural, and hydrological applications, Earth Interactions, 16, https://doi.org/10.1175/2012EI000434.1\n\n[[4]](https://doi.org/10.3390/eesp2025035029) Kalisoras, A., Georgoulias, A. K., Akritidis, D., and Zanis, P., 2025. Future Projections in Agricultural Drought Characteristics for Greece Under Different Climate Change Scenarios. Environmental and Earth Sciences Proceedings, 35(1), 29. https://doi.org/10.3390/eesp2025035029\n\n[[5]](https://doi.org/10.1002/joc.7302) Spinoni, J., Barbosa, P., Bucchignani, E., Cassano, J., Cavazos, T., Cescatti, A., Christensen, J. H., Christensen, O. B., Coppola, E., Evans, J. P., Forzieri, G., Geyer, B., Giorgi, F., Jacob, D., Katzfey, J., Koenigk, T., Laprise, R., Lennard, C. J., Kurnaz, M. L., … Dosio, A., 2021. Global exposure of population and land-use to meteorological droughts under different warming levels and SSPs: A CORDEX-based study. International Journal of Climatology, 41(15), 6825–6853. https://doi.org/10.1002/joc.7302\n\n[[6]](https://doi.org/10.5194/essd-12-2959-2020) Iturbide, M., Gutiérrez, J. M., Alves, L. M., Bedia, J., Cerezo-Mota, R., Cimadevilla, E., Cofiño, A. S., Di Luca, A., Faria, S. H., Gorodetskaya, I. V., Hauser, M., Herrera, S., Hennessy, K., Hewitt, H. T., Jones, R. G., Krakovska, S., Manzanas, R., Martínez-Castro, D., Narisma, G. T., Nurhati, I. S., Pinto, I., Seneviratne, S. I., van den Hurk, B., and Vera, C. S., 2020. An update of IPCC climate reference regions for subcontinental analysis of climate model data: definition and aggregated datasets, Earth Syst. Sci. Data, 12, 2959–2970, https://doi.org/10.5194/essd-12-2959-2020\n\n[[7]](https://doi.org/10.13031/2013.26773) Hargreaves, G. H., and Samani, Z. A., 1985. Reference crop evapotranspiration from temperature, Applied engineering in agriculture, pp. 96–99. https://doi.org/10.13031/2013.26773", "text_with_prefix": "EQC Quality Assessment: \"Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q08 | Category: Climate_Projections\nSection: Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region > ℹ️ If you want to know more > References\n---\n[[1]](https://doi.org/10.1175/2009JCLI2909.1) Vicente-Serrano, S. M., Beguerı́a, S., and López-Moreno, J. I., 2010. A multiscalar drought index sensitive to global warming: The standardized precipitation evapotranspiration index, Journal of Climate, 23, 1696–1718. https://doi.org/10.1175/2009JCLI2909.1\n\n[[2]](https://digitalcommons.unl.edu/droughtfacpub/69/) Wilhite, D. A., 2000. Droughts as a natural hazard: concepts and definitions. In: DROUGHT, A Global Assessment, vol. I and II, Routledge Hazards and Disasters Series, Routledge.\n\n[[3]](https://doi.org/10.1175/2012EI000434.1) Vicente-Serrano, S. M., Beguerı́a, S., Lorenzo-Lacruz, J., Camarero, J. J., López-Moreno, J. I., Azorin-Molina, C., Revuelto, J., Morán-Tejeda, E., and Sanchez-Lorenzo, A., 2012. Performance of drought indices for ecological, agricultural, and hydrological applications, Earth Interactions, 16, https://doi.org/10.1175/2012EI000434.1\n\n[[4]](https://doi.org/10.3390/eesp2025035029) Kalisoras, A., Georgoulias, A. K., Akritidis, D., and Zanis, P., 2025. Future Projections in Agricultural Drought Characteristics for Greece Under Different Climate Change Scenarios. Environmental and Earth Sciences Proceedings, 35(1), 29. https://doi.org/10.3390/eesp2025035029\n\n[[5]](https://doi.org/10.1002/joc.7302) Spinoni, J., Barbosa, P., Bucchignani, E., Cassano, J., Cavazos, T., Cescatti, A., Christensen, J. H., Christensen, O. B., Coppola, E., Evans, J. P., Forzieri, G., Geyer, B., Giorgi, F., Jacob, D., Katzfey, J., Koenigk, T., Laprise, R., Lennard, C. J., Kurnaz, M. L., … Dosio, A., 2021. Global exposure of population and land-use to meteorological droughts under different warming levels and SSPs: A CORDEX-based study. International Journal of Climatology, 41(15), 6825–6853. https://doi.org/10.1002/joc.7302\n\n[[6]](https://doi.org/10.5194/essd-12-2959-2020) Iturbide, M., Gutiérrez, J. M., Alves, L. M., Bedia, J., Cerezo-Mota, R., Cimadevilla, E., Cofiño, A. S., Di Luca, A., Faria, S. H., Gorodetskaya, I. V., Hauser, M., Herrera, S., Hennessy, K., Hewitt, H. T., Jones, R. G., Krakovska, S., Manzanas, R., Martínez-Castro, D., Narisma, G. T., Nurhati, I. S., Pinto, I., Seneviratne, S. I., van den Hurk, B., and Vera, C. S., 2020. An update of IPCC climate reference regions for subcontinental analysis of climate model data: definition and aggregated datasets, Earth Syst. Sci. Data, 12, 2959–2970, https://doi.org/10.5194/essd-12-2959-2020\n\n[[7]](https://doi.org/10.13031/2013.26773) Hargreaves, G. H., and Samani, Z. A., 1985. Reference crop evapotranspiration from temperature, Applied engineering in agriculture, pp. 96–99. https://doi.org/10.13031/2013.26773"} {"chunk_id": "climate_projections-cordex-domains-single-levels_validation_q08__d0711ad8b064", "report_id": "climate_projections-cordex-domains-single-levels_validation_q08", "dataset_id": "projections-cordex-domains-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q08", "aspect_base": "validation", "category": "Climate_Projections", "match_confidence": "exact", "section": "Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region > ℹ️ If you want to know more > References", "title": "Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region", "chunk_index": 22, "token_count": 248, "text_raw": "ani, Z. A., 1985. Reference crop evapotranspiration from temperature, Applied engineering in agriculture, pp. 96–99. https://doi.org/10.13031/2013.26773\n\n[[8]](https://doi.org/10.1002/joc.3887) Beguerı́a, S., Vicente-Serrano, S. M., Reig, F., and Latorre, B., 2014. Standardized precipitation evapotranspiration index (spei) revisited: Parameter fitting, evapotranspiration models, tools, datasets and drought monitoring, International Journal of Climatology, 34, 3001–3023. https://doi.org/10.1002/joc.3887", "text_with_prefix": "EQC Quality Assessment: \"Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region\"\nDataset: projections-cordex-domains-single-levels [CDS]\nAspect: validation_q08 | Category: Climate_Projections\nSection: Near-future CORDEX projections of the SPEI6 drought index over the Mediterranean region > ℹ️ If you want to know more > References\n---\nani, Z. A., 1985. Reference crop evapotranspiration from temperature, Applied engineering in agriculture, pp. 96–99. https://doi.org/10.13031/2013.26773\n\n[[8]](https://doi.org/10.1002/joc.3887) Beguerı́a, S., Vicente-Serrano, S. M., Reig, F., and Latorre, B., 2014. Standardized precipitation evapotranspiration index (spei) revisited: Parameter fitting, evapotranspiration models, tools, datasets and drought monitoring, International Journal of Climatology, 34, 3001–3023. https://doi.org/10.1002/joc.3887"} {"chunk_id": "derived_multi-origin-c3s-atlas_consistency_q01__9dbf29930b62", "report_id": "derived_multi-origin-c3s-atlas_consistency_q01", "dataset_id": "multi-origin-c3s-atlas", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q01", "aspect_base": "consistency", "category": "Derived_Datasets", "match_confidence": "exact", "section": "Consistency between the C3S Atlas dataset and its origins: Case study", "title": "Consistency between the C3S Atlas dataset and its origins: Case study", "chunk_index": 0, "token_count": 103, "text_raw": "Production date: 2025-09-30.\n\nDataset version: 2.0.\n\nProduced by: Olivier Burggraaff, Nicole Reynolds (National Physical Laboratory).", "text_with_prefix": "EQC Quality Assessment: \"Consistency between the C3S Atlas dataset and its origins: Case study\"\nDataset: multi-origin-c3s-atlas [CDS]\nAspect: consistency_q01 | Category: Derived_Datasets\nSection: Consistency between the C3S Atlas dataset and its origins: Case study\n---\nProduction date: 2025-09-30.\n\nDataset version: 2.0.\n\nProduced by: Olivier Burggraaff, Nicole Reynolds (National Physical Laboratory)."} {"chunk_id": "derived_multi-origin-c3s-atlas_consistency_q01__cde778eaa348", "report_id": "derived_multi-origin-c3s-atlas_consistency_q01", "dataset_id": "multi-origin-c3s-atlas", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q01", "aspect_base": "consistency", "category": "Derived_Datasets", "match_confidence": "exact", "section": "Consistency between the C3S Atlas dataset and its origins: Case study > Quality assessment question", "title": "Consistency between the C3S Atlas dataset and its origins: Case study", "chunk_index": 1, "token_count": 881, "text_raw": "* **Are the climate indicators in the dataset underpinning the Copernicus Interactive Climate Atlas consistent with their origin datasets?**\n* **Can the dataset underpinning the Copernicus Interactive Climate Atlas be reproduced from its origin datasets?**\n\nThe [_Copernicus Interactive Climate Atlas_](https://atlas.climate.copernicus.eu/atlas), or _C3S Atlas_ for short, is a C3S web application providing an easy-to-access tool for exploring climate projections, reanalyses, and observational data [[Gutiérrez+24](https://doi.org/10.21957/ah52ufc369)].\nVersion 2.0 of the application allows the user to interact with 12 datasets:\n\n| Type | Dataset |\n|--------------------|---------------|\n| Climate Projection | CMIP6 |\n| Climate Projection | CMIP5 |\n| Climate Projection | CORDEX-CORE |\n| Climate Projection | CORDEX-EUR-11 |\n| Reanalysis | ERA5 |\n| Reanalysis | ERA5-Land |\n| Reanalysis | ORAS5 |\n| Reanalysis | CERRA |\n| Observations | E-OBS |\n| Observations | BERKEARTH |\n| Observations | CPC |\n| Observations | SST-CCI |\n\nThese datasets are provided through an intermediary dataset, the [_Gridded dataset underpinning the Copernicus Interactive Climate Atlas_](https://doi.org/10.24381/cds.h35hb680) or _C3S Atlas dataset_ for short [[C3S Atlas dataset](https://doi.org/10.24381/cds.h35hb680)].\nCompared to their origins, the versions of the climate datasets within the C3S Atlas dataset have been processed following the workflow in Figure {numref}`{number} `.\n\nattachment:c3s_atlas_dataset_workflow.png\n---\nheight: 360px\nname: multi-origin-c3s-atlas_consistency_q01_workflow-fig\n---\nSchematic representation of the workflow for the production of the C3S Atlas dataset from its origin datasets, from the [User-tools for the C3S Atlas](https://ecmwf-projects.github.io/c3s-atlas/chapter01.html).\n```\n\nBecause a wide range of users interact with climate data through the C3S Atlas application, it is crucial that the underpinning dataset represent its origins correctly.\nIn other words, the C3S Atlas dataset must be consistent with and reproducible from its origins.\nHere, we assess this consistency and reproducibility by comparing climate indicators retrieved from the C3S Atlas dataset with their equivalents calculated from the origin dataset, mirroring the workflow from Figure {numref}`{number} `.\nWhile a full analysis and reproduction of every record within the C3S Atlas dataset is outside the scope of quality assessment\n(and would require high-performance computing infrastructure),\na case study with a narrower scope probes these quality attributes of the dataset\nand can be a jumping-off point for further analysis by the reader.\n\nThis notebook is part of a series:\n| Notebook | Contents |\n|---|---|\n| **Consistency between the C3S Atlas dataset and its origins: Case study** | Comparison between C3S Atlas dataset and one origin dataset (CMIP6) for one indicator (`tx35`), including detailed setup. |\n| [](./derived_multi-origin-c3s-atlas_consistency_q02) | Comparison between C3S Atlas dataset and one origin dataset (CMIP6) for multiple indicators. |\n| [](./derived_multi-origin-c3s-atlas_consistency_q03) | Comparison between C3S Atlas dataset and multiple origin datasets for one indicator. |", "text_with_prefix": "EQC Quality Assessment: \"Consistency between the C3S Atlas dataset and its origins: Case study\"\nDataset: multi-origin-c3s-atlas [CDS]\nAspect: consistency_q01 | Category: Derived_Datasets\nSection: Consistency between the C3S Atlas dataset and its origins: Case study > Quality assessment question\n---\n* **Are the climate indicators in the dataset underpinning the Copernicus Interactive Climate Atlas consistent with their origin datasets?**\n* **Can the dataset underpinning the Copernicus Interactive Climate Atlas be reproduced from its origin datasets?**\n\nThe [_Copernicus Interactive Climate Atlas_](https://atlas.climate.copernicus.eu/atlas), or _C3S Atlas_ for short, is a C3S web application providing an easy-to-access tool for exploring climate projections, reanalyses, and observational data [[Gutiérrez+24](https://doi.org/10.21957/ah52ufc369)].\nVersion 2.0 of the application allows the user to interact with 12 datasets:\n\n| Type | Dataset |\n|--------------------|---------------|\n| Climate Projection | CMIP6 |\n| Climate Projection | CMIP5 |\n| Climate Projection | CORDEX-CORE |\n| Climate Projection | CORDEX-EUR-11 |\n| Reanalysis | ERA5 |\n| Reanalysis | ERA5-Land |\n| Reanalysis | ORAS5 |\n| Reanalysis | CERRA |\n| Observations | E-OBS |\n| Observations | BERKEARTH |\n| Observations | CPC |\n| Observations | SST-CCI |\n\nThese datasets are provided through an intermediary dataset, the [_Gridded dataset underpinning the Copernicus Interactive Climate Atlas_](https://doi.org/10.24381/cds.h35hb680) or _C3S Atlas dataset_ for short [[C3S Atlas dataset](https://doi.org/10.24381/cds.h35hb680)].\nCompared to their origins, the versions of the climate datasets within the C3S Atlas dataset have been processed following the workflow in Figure {numref}`{number} `.\n\nattachment:c3s_atlas_dataset_workflow.png\n---\nheight: 360px\nname: multi-origin-c3s-atlas_consistency_q01_workflow-fig\n---\nSchematic representation of the workflow for the production of the C3S Atlas dataset from its origin datasets, from the [User-tools for the C3S Atlas](https://ecmwf-projects.github.io/c3s-atlas/chapter01.html).\n```\n\nBecause a wide range of users interact with climate data through the C3S Atlas application, it is crucial that the underpinning dataset represent its origins correctly.\nIn other words, the C3S Atlas dataset must be consistent with and reproducible from its origins.\nHere, we assess this consistency and reproducibility by comparing climate indicators retrieved from the C3S Atlas dataset with their equivalents calculated from the origin dataset, mirroring the workflow from Figure {numref}`{number} `.\nWhile a full analysis and reproduction of every record within the C3S Atlas dataset is outside the scope of quality assessment\n(and would require high-performance computing infrastructure),\na case study with a narrower scope probes these quality attributes of the dataset\nand can be a jumping-off point for further analysis by the reader.\n\nThis notebook is part of a series:\n| Notebook | Contents |\n|---|---|\n| **Consistency between the C3S Atlas dataset and its origins: Case study** | Comparison between C3S Atlas dataset and one origin dataset (CMIP6) for one indicator (`tx35`), including detailed setup. |\n| [](./derived_multi-origin-c3s-atlas_consistency_q02) | Comparison between C3S Atlas dataset and one origin dataset (CMIP6) for multiple indicators. |\n| [](./derived_multi-origin-c3s-atlas_consistency_q03) | Comparison between C3S Atlas dataset and multiple origin datasets for one indicator. |"} {"chunk_id": "derived_multi-origin-c3s-atlas_consistency_q01__21923d2a0ce0", "report_id": "derived_multi-origin-c3s-atlas_consistency_q01", "dataset_id": "multi-origin-c3s-atlas", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q01", "aspect_base": "consistency", "category": "Derived_Datasets", "match_confidence": "exact", "section": "Consistency between the C3S Atlas dataset and its origins: Case study > Quality assessment statement", "title": "Consistency between the C3S Atlas dataset and its origins: Case study", "chunk_index": 2, "token_count": 186, "text_raw": "These are the key outcomes of this assessment\n\n* Climate indicators (here `tx35`) provided by the C3S Atlas dataset are highly consistent with values calculated from its origin datasets (here the CMIP6 multi-model ensemble).\n\n* Differences between the C3S Atlas dataset and a manual reproduction are rare (fewer than 0.1% of pixel pairs) and generally negligible (median absolute difference of 0).\n\n* The C3S Atlas is traceable and reproducible, and can be confidently used to view, analyse, and download climate data.\n```", "text_with_prefix": "EQC Quality Assessment: \"Consistency between the C3S Atlas dataset and its origins: Case study\"\nDataset: multi-origin-c3s-atlas [CDS]\nAspect: consistency_q01 | Category: Derived_Datasets\nSection: Consistency between the C3S Atlas dataset and its origins: Case study > Quality assessment statement\n---\nThese are the key outcomes of this assessment\n\n* Climate indicators (here `tx35`) provided by the C3S Atlas dataset are highly consistent with values calculated from its origin datasets (here the CMIP6 multi-model ensemble).\n\n* Differences between the C3S Atlas dataset and a manual reproduction are rare (fewer than 0.1% of pixel pairs) and generally negligible (median absolute difference of 0).\n\n* The C3S Atlas is traceable and reproducible, and can be confidently used to view, analyse, and download climate data.\n```"} {"chunk_id": "derived_multi-origin-c3s-atlas_consistency_q01__c8ec061fba04", "report_id": "derived_multi-origin-c3s-atlas_consistency_q01", "dataset_id": "multi-origin-c3s-atlas", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q01", "aspect_base": "consistency", "category": "Derived_Datasets", "match_confidence": "exact", "section": "Consistency between the C3S Atlas dataset and its origins: Case study > Methodology", "title": "Consistency between the C3S Atlas dataset and its origins: Case study", "chunk_index": 3, "token_count": 442, "text_raw": "This quality assessment tests the consistency between climate indicators retrieved from the [_Gridded dataset underpinning the Copernicus Interactive Climate Atlas_](https://doi.org/10.24381/cds.h35hb680) [[C3S Atlas dataset](https://doi.org/10.24381/cds.h35hb680)] and their equivalents calculated from the origin datasets,\nas well as the reproducibility of said dataset.\n\nThis notebook starts the quality assessment with a simple case study: one climate indicator derived from one origin dataset.\nThis indicator is `tx35`,\nthe _Monthly count of days with maximum near-surface (2-metre) air temperature above 35 °C_,\nderived from the CMIP6 multi-model ensemble [[CMIP6 dataset](https://doi.org/10.24381/cds.c866074c)].\nThis notebook mirrors [one of the reproducibility demonstrations](https://ecmwf-projects.github.io/c3s-atlas/notebooks/tx35.html) written by the data provider.\n\nThe analysis and results are organised in the following steps, which are detailed in the sections below:\n\n**[](section-codesetup)**\n * Install User-tools for the C3S Atlas.\n * Import all required libraries.\n * Define indicator.\n * Define helper functions.\n\n**[](section-origin)**\n * Download data from the origin dataset.\n * Homogenise data.\n * Calculate indicator.\n * Regrid the origin data to the C3S Atlas grid.\n\n**[](section-c3s-atlas)**\n * Download data from the C3S Atlas dataset.\n\n**[](section-results)**\n * Consistency: Compare the C3S Atlas and reproduced datasets on native grids.\n * Reproducibility: Compare the C3S Atlas and reproduced datasets on the C3S Atlas grid.", "text_with_prefix": "EQC Quality Assessment: \"Consistency between the C3S Atlas dataset and its origins: Case study\"\nDataset: multi-origin-c3s-atlas [CDS]\nAspect: consistency_q01 | Category: Derived_Datasets\nSection: Consistency between the C3S Atlas dataset and its origins: Case study > Methodology\n---\nThis quality assessment tests the consistency between climate indicators retrieved from the [_Gridded dataset underpinning the Copernicus Interactive Climate Atlas_](https://doi.org/10.24381/cds.h35hb680) [[C3S Atlas dataset](https://doi.org/10.24381/cds.h35hb680)] and their equivalents calculated from the origin datasets,\nas well as the reproducibility of said dataset.\n\nThis notebook starts the quality assessment with a simple case study: one climate indicator derived from one origin dataset.\nThis indicator is `tx35`,\nthe _Monthly count of days with maximum near-surface (2-metre) air temperature above 35 °C_,\nderived from the CMIP6 multi-model ensemble [[CMIP6 dataset](https://doi.org/10.24381/cds.c866074c)].\nThis notebook mirrors [one of the reproducibility demonstrations](https://ecmwf-projects.github.io/c3s-atlas/notebooks/tx35.html) written by the data provider.\n\nThe analysis and results are organised in the following steps, which are detailed in the sections below:\n\n**[](section-codesetup)**\n * Install User-tools for the C3S Atlas.\n * Import all required libraries.\n * Define indicator.\n * Define helper functions.\n\n**[](section-origin)**\n * Download data from the origin dataset.\n * Homogenise data.\n * Calculate indicator.\n * Regrid the origin data to the C3S Atlas grid.\n\n**[](section-c3s-atlas)**\n * Download data from the C3S Atlas dataset.\n\n**[](section-results)**\n * Consistency: Compare the C3S Atlas and reproduced datasets on native grids.\n * Reproducibility: Compare the C3S Atlas and reproduced datasets on the C3S Atlas grid."} {"chunk_id": "derived_multi-origin-c3s-atlas_consistency_q01__4cbfff22f59a", "report_id": "derived_multi-origin-c3s-atlas_consistency_q01", "dataset_id": "multi-origin-c3s-atlas", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q01", "aspect_base": "consistency", "category": "Derived_Datasets", "match_confidence": "exact", "section": "Consistency between the C3S Atlas dataset and its origins: Case study > Analysis and results > 1. Code setup", "title": "Consistency between the C3S Atlas dataset and its origins: Case study", "chunk_index": 4, "token_count": 183, "text_raw": "This notebook uses [earthkit](https://github.com/ecmwf/earthkit) for\ndownloading ([earthkit-data](https://github.com/ecmwf/earthkit-data))\nand visualising ([earthkit-plots](https://github.com/ecmwf/earthkit-plots)) data.\nBecause earthkit is in active development, some functionality may change after this notebook is published.\nIf any part of the code stops functioning, please raise an issue on our GitHub repository so it can be fixed.\n```", "text_with_prefix": "EQC Quality Assessment: \"Consistency between the C3S Atlas dataset and its origins: Case study\"\nDataset: multi-origin-c3s-atlas [CDS]\nAspect: consistency_q01 | Category: Derived_Datasets\nSection: Consistency between the C3S Atlas dataset and its origins: Case study > Analysis and results > 1. Code setup\n---\nThis notebook uses [earthkit](https://github.com/ecmwf/earthkit) for\ndownloading ([earthkit-data](https://github.com/ecmwf/earthkit-data))\nand visualising ([earthkit-plots](https://github.com/ecmwf/earthkit-plots)) data.\nBecause earthkit is in active development, some functionality may change after this notebook is published.\nIf any part of the code stops functioning, please raise an issue on our GitHub repository so it can be fixed.\n```"} {"chunk_id": "derived_multi-origin-c3s-atlas_consistency_q01__846b80c9a95d", "report_id": "derived_multi-origin-c3s-atlas_consistency_q01", "dataset_id": "multi-origin-c3s-atlas", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q01", "aspect_base": "consistency", "category": "Derived_Datasets", "match_confidence": "exact", "section": "Consistency between the C3S Atlas dataset and its origins: Case study > Analysis and results > 1. Code setup > Install the User-tools for the C3S Atlas", "title": "Consistency between the C3S Atlas dataset and its origins: Case study", "chunk_index": 5, "token_count": 182, "text_raw": "This notebook uses the [User-tools for the C3S Atlas](https://github.com/ecmwf-projects/c3s-atlas), which can be installed from GitHub using `pip`.\nFor convenience, the following cell can do this from within the notebook.\nFurther details and alternative options for installing this library are available in its [documentation](https://github.com/ecmwf-projects/c3s-atlas?tab=readme-ov-file#requirements).", "text_with_prefix": "EQC Quality Assessment: \"Consistency between the C3S Atlas dataset and its origins: Case study\"\nDataset: multi-origin-c3s-atlas [CDS]\nAspect: consistency_q01 | Category: Derived_Datasets\nSection: Consistency between the C3S Atlas dataset and its origins: Case study > Analysis and results > 1. Code setup > Install the User-tools for the C3S Atlas\n---\nThis notebook uses the [User-tools for the C3S Atlas](https://github.com/ecmwf-projects/c3s-atlas), which can be installed from GitHub using `pip`.\nFor convenience, the following cell can do this from within the notebook.\nFurther details and alternative options for installing this library are available in its [documentation](https://github.com/ecmwf-projects/c3s-atlas?tab=readme-ov-file#requirements)."} {"chunk_id": "derived_multi-origin-c3s-atlas_consistency_q01__a88b66f280b2", "report_id": "derived_multi-origin-c3s-atlas_consistency_q01", "dataset_id": "multi-origin-c3s-atlas", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q01", "aspect_base": "consistency", "category": "Derived_Datasets", "match_confidence": "exact", "section": "Consistency between the C3S Atlas dataset and its origins: Case study > Analysis and results > 1. Code setup > Import required libraries", "title": "Consistency between the C3S Atlas dataset and its origins: Case study", "chunk_index": 6, "token_count": 126, "text_raw": "In this section, we import all the relevant packages needed for running the notebook.\n\nInput / Output\nGeneral data handling\nData pre-processing\nClimate indicators\nVisualisation\nVisualisation in Jupyter book -- automatically ignored otherwise", "text_with_prefix": "EQC Quality Assessment: \"Consistency between the C3S Atlas dataset and its origins: Case study\"\nDataset: multi-origin-c3s-atlas [CDS]\nAspect: consistency_q01 | Category: Derived_Datasets\nSection: Consistency between the C3S Atlas dataset and its origins: Case study > Analysis and results > 1. Code setup > Import required libraries\n---\nIn this section, we import all the relevant packages needed for running the notebook.\n\nInput / Output\nGeneral data handling\nData pre-processing\nClimate indicators\nVisualisation\nVisualisation in Jupyter book -- automatically ignored otherwise"} {"chunk_id": "derived_multi-origin-c3s-atlas_consistency_q01__c8979b08301c", "report_id": "derived_multi-origin-c3s-atlas_consistency_q01", "dataset_id": "multi-origin-c3s-atlas", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q01", "aspect_base": "consistency", "category": "Derived_Datasets", "match_confidence": "exact", "section": "Consistency between the C3S Atlas dataset and its origins: Case study > Analysis and results > 1. Code setup > Define indicators", "title": "Consistency between the C3S Atlas dataset and its origins: Case study", "chunk_index": 7, "token_count": 158, "text_raw": "This section defines functions and variables for calculating and using the climate indicators:\n\nFunctions to calculate indicators\n\nThe following cell defines [earthkit-plots styles](https://earthkit-plots.readthedocs.io/en/latest/examples/examples/introduction/05-styles.html) for the indicators.\nThese styles define the colour maps and colour bar ranges for each quantity.\n\nStyles for indicators\nApply general settings", "text_with_prefix": "EQC Quality Assessment: \"Consistency between the C3S Atlas dataset and its origins: Case study\"\nDataset: multi-origin-c3s-atlas [CDS]\nAspect: consistency_q01 | Category: Derived_Datasets\nSection: Consistency between the C3S Atlas dataset and its origins: Case study > Analysis and results > 1. Code setup > Define indicators\n---\nThis section defines functions and variables for calculating and using the climate indicators:\n\nFunctions to calculate indicators\n\nThe following cell defines [earthkit-plots styles](https://earthkit-plots.readthedocs.io/en/latest/examples/examples/introduction/05-styles.html) for the indicators.\nThese styles define the colour maps and colour bar ranges for each quantity.\n\nStyles for indicators\nApply general settings"} {"chunk_id": "derived_multi-origin-c3s-atlas_consistency_q01__563cb91d375a", "report_id": "derived_multi-origin-c3s-atlas_consistency_q01", "dataset_id": "multi-origin-c3s-atlas", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q01", "aspect_base": "consistency", "category": "Derived_Datasets", "match_confidence": "exact", "section": "Consistency between the C3S Atlas dataset and its origins: Case study > Analysis and results > 1. Code setup > Helper functions", "title": "Consistency between the C3S Atlas dataset and its origins: Case study", "chunk_index": 8, "token_count": 847, "text_raw": "This section defines some functions and variables used in the following analysis, allowing code cells in later sections to be shorter and ensuring consistency.\n\n##### Data (pre-)processing\n\nThe following functions handle the homogenisation of origin data to a consistent format using the [User-tools for the C3S Atlas](https://ecmwf-projects.github.io/c3s-atlas/intro.html):\n\nHomogenisation of origin dataset\nHomogenisation of origin dataset, multiple non-consecutive years\n\nThe following functions handle regridding data based on ESMF as implemented in the [User-tools for the C3S Atlas](https://ecmwf-projects.github.io/c3s-atlas/intro.html).\nThis step is explained in more detail in the relevant section below.\n\nRegridding from native grid to C3S Atlas grid\n\nThe following functions aid in sub-selecting data, e.g. selecting one model from the ensemble included in the C3S Atlas dataset or selecting data for a specific time frame:\n\nC3S Atlas dataset: Individual model selection\nEnsure the model ID is provided in the right format\nFind the corresponding model ID in the list of models\nThis cannot use .sel because the coordinate is not indexed\nFind the corresponding data and return those\nSelect (multiple) years in a dataset\nSelect one month in multiple datasets\n\n##### Statistics\nThe following functions calculate the difference (absolute / relative) between datasets, handling metadata etc.:\n\nDifference between datasets\nSelect and calculate\nReplace 0/0 with 0\n\nThe following functions calculate and display metrics for the difference between two datasets, e.g. mean and median deviation:\n\nCalculate differences\nConvert to pandas\nIn this notebook: Group by year first\nCalculate aggregate statistics\nCalculate correlation coefficients\nCombine statistics into one dataframe\n\n##### Visualisation\nThe following cells contain functions for plotting results, starting with some base helper functions (e.g. displaying in Jupyter Notebook or Jupyter Book style, adding textboxes with consistent formatting, etc.):\n\nVisualisation: Helper functions, general\nGet the plt.Axes for each ekp.Subplot\nSet up location\nAdd the text\n\nThe following functions are also base helper functions, but specific to geospatial plots:\n\nVisualisation: Helper functions for geospatial plots\nCreate subplots\nRemove redundant time coordinate\n\nThe following cell contains functions for geospatial comparisons between datasets on their native grids or on a common grid (the latter also showing the per-pixel difference):\n\nVisualisation: Plot indicators geospatially\nPre-process: Select data in one month, group by year\nCreate figure\nPlot indicators\nPlot individual datasets\nDecorate: Text + Colour bar\nTitles on top\nDecorate figure\nVisualisation: Plot indicator + difference geospatially\nPre-process: Select data in one month, group by year\nCreate figure\nPlot indicators\nPlot individual datasets\nPlot difference\nColour bar at the bottom\nTitles on top\nDecorate figure\n\nThe following cell contains functions for histogram comparisons between datasets on their native grids or on a common grid (the latter also showing the per-pixel difference):\n\nVisualisation: Plot data in histograms\nGroup data by year\nCreate figure\nPlot histograms of data\nLoop over rows / indicators\nLoop over columns / data\nFlatten data\nCreate histogram\nIdentify panel\nTitles on top\nVisualisation: Plot data + difference in histograms\nGroup data by year\nCreate figure\nSetup x/y share -- cannot be done in plt.subplots because of difference panel not sharing these\nPlot histograms of data\nLoop over rows / indicators\nCalculate difference\nLoop over columns / data\nFlatten data\nCreate histogram\nPlot difference\nIdentify panel\nTitles on top\nDecorate figure\n\n(section-origin)=", "text_with_prefix": "EQC Quality Assessment: \"Consistency between the C3S Atlas dataset and its origins: Case study\"\nDataset: multi-origin-c3s-atlas [CDS]\nAspect: consistency_q01 | Category: Derived_Datasets\nSection: Consistency between the C3S Atlas dataset and its origins: Case study > Analysis and results > 1. Code setup > Helper functions\n---\nThis section defines some functions and variables used in the following analysis, allowing code cells in later sections to be shorter and ensuring consistency.\n\n##### Data (pre-)processing\n\nThe following functions handle the homogenisation of origin data to a consistent format using the [User-tools for the C3S Atlas](https://ecmwf-projects.github.io/c3s-atlas/intro.html):\n\nHomogenisation of origin dataset\nHomogenisation of origin dataset, multiple non-consecutive years\n\nThe following functions handle regridding data based on ESMF as implemented in the [User-tools for the C3S Atlas](https://ecmwf-projects.github.io/c3s-atlas/intro.html).\nThis step is explained in more detail in the relevant section below.\n\nRegridding from native grid to C3S Atlas grid\n\nThe following functions aid in sub-selecting data, e.g. selecting one model from the ensemble included in the C3S Atlas dataset or selecting data for a specific time frame:\n\nC3S Atlas dataset: Individual model selection\nEnsure the model ID is provided in the right format\nFind the corresponding model ID in the list of models\nThis cannot use .sel because the coordinate is not indexed\nFind the corresponding data and return those\nSelect (multiple) years in a dataset\nSelect one month in multiple datasets\n\n##### Statistics\nThe following functions calculate the difference (absolute / relative) between datasets, handling metadata etc.:\n\nDifference between datasets\nSelect and calculate\nReplace 0/0 with 0\n\nThe following functions calculate and display metrics for the difference between two datasets, e.g. mean and median deviation:\n\nCalculate differences\nConvert to pandas\nIn this notebook: Group by year first\nCalculate aggregate statistics\nCalculate correlation coefficients\nCombine statistics into one dataframe\n\n##### Visualisation\nThe following cells contain functions for plotting results, starting with some base helper functions (e.g. displaying in Jupyter Notebook or Jupyter Book style, adding textboxes with consistent formatting, etc.):\n\nVisualisation: Helper functions, general\nGet the plt.Axes for each ekp.Subplot\nSet up location\nAdd the text\n\nThe following functions are also base helper functions, but specific to geospatial plots:\n\nVisualisation: Helper functions for geospatial plots\nCreate subplots\nRemove redundant time coordinate\n\nThe following cell contains functions for geospatial comparisons between datasets on their native grids or on a common grid (the latter also showing the per-pixel difference):\n\nVisualisation: Plot indicators geospatially\nPre-process: Select data in one month, group by year\nCreate figure\nPlot indicators\nPlot individual datasets\nDecorate: Text + Colour bar\nTitles on top\nDecorate figure\nVisualisation: Plot indicator + difference geospatially\nPre-process: Select data in one month, group by year\nCreate figure\nPlot indicators\nPlot individual datasets\nPlot difference\nColour bar at the bottom\nTitles on top\nDecorate figure\n\nThe following cell contains functions for histogram comparisons between datasets on their native grids or on a common grid (the latter also showing the per-pixel difference):\n\nVisualisation: Plot data in histograms\nGroup data by year\nCreate figure\nPlot histograms of data\nLoop over rows / indicators\nLoop over columns / data\nFlatten data\nCreate histogram\nIdentify panel\nTitles on top\nVisualisation: Plot data + difference in histograms\nGroup data by year\nCreate figure\nSetup x/y share -- cannot be done in plt.subplots because of difference panel not sharing these\nPlot histograms of data\nLoop over rows / indicators\nCalculate difference\nLoop over columns / data\nFlatten data\nCreate histogram\nPlot difference\nIdentify panel\nTitles on top\nDecorate figure\n\n(section-origin)="} {"chunk_id": "derived_multi-origin-c3s-atlas_consistency_q01__b6b1384cfce8", "report_id": "derived_multi-origin-c3s-atlas_consistency_q01", "dataset_id": "multi-origin-c3s-atlas", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q01", "aspect_base": "consistency", "category": "Derived_Datasets", "match_confidence": "exact", "section": "Consistency between the C3S Atlas dataset and its origins: Case study > Analysis and results > 2. Calculate indicator from the origin dataset > Download data", "title": "Consistency between the C3S Atlas dataset and its origins: Case study", "chunk_index": 9, "token_count": 602, "text_raw": "This assessment examines the dataset in two years (2060 and 2080),\nfor one ensemble member\nand\nfor one climate scenario\n– which is not how climate projection data are normally used.\nGood practice in climate science is to look at multi-year statistics and trends,\nacross multiple ensemble members.\nHowever, since the purpose of this assessment is to assess the consistency and reproducibility of the post-processing performed to produce the C3S Atlas dataset,\nit is valid to use a subset of the data here.\nThe specific subset used can be easily tweaked by changing the `EXPERIMENT`, `MODEL`, and `YEARS` variables in this section.\n\nThis notebook uses [earthkit-data](https://github.com/ecmwf/earthkit-data) to download files from the CDS.\nIf you intend to run this notebook multiple times, it is highly recommended that you [enable caching](https://earthkit-data.readthedocs.io/en/latest/concepts/caching.html) to prevent having to download the same files multiple times.\nIf you prefer not to use earthkit, the following requests can also be used with the [cdsapi module](https://cds.climate.copernicus.eu/how-to-api#linux-use-client-step).\nIn either case (earthkit-data or cdsapi), it is required to set up a CDS account and API key as explained [on the CDS website](https://cds.climate.copernicus.eu/how-to-api).\n\nThe first step is to define the parameters that will be shared between the download of the origin dataset (here CMIP6) and the C3S Atlas dataset (in the [next section](section-c3s-atlas)), namely the experiment and model member.\nNext, the request to download the corresponding data from CMIP6 is defined,\nin this notebook choosing only the variable necessary to calculate the `tx35` indicator, namely `daily_maximum_near_surface_air_temperature`.\nBecause the indicator will be calculated for multiple years, the `year` parameter of the request is left out for now.\n\nDefine request\n\nThis case study examines\ntwo years, namely 2060 and 2080, so there are two\ncorresponding requests.\nThis can easily be changed or extended to use different years.\nSeparate requests are defined for each year,\nrather than one combined request,\nto limit the size per request.\n\nDownload data\n\nEarthkit downloads the data as a field list.\nThis is converted into an `xarray` dataset, which can be used in the following analysis.", "text_with_prefix": "EQC Quality Assessment: \"Consistency between the C3S Atlas dataset and its origins: Case study\"\nDataset: multi-origin-c3s-atlas [CDS]\nAspect: consistency_q01 | Category: Derived_Datasets\nSection: Consistency between the C3S Atlas dataset and its origins: Case study > Analysis and results > 2. Calculate indicator from the origin dataset > Download data\n---\nThis assessment examines the dataset in two years (2060 and 2080),\nfor one ensemble member\nand\nfor one climate scenario\n– which is not how climate projection data are normally used.\nGood practice in climate science is to look at multi-year statistics and trends,\nacross multiple ensemble members.\nHowever, since the purpose of this assessment is to assess the consistency and reproducibility of the post-processing performed to produce the C3S Atlas dataset,\nit is valid to use a subset of the data here.\nThe specific subset used can be easily tweaked by changing the `EXPERIMENT`, `MODEL`, and `YEARS` variables in this section.\n\nThis notebook uses [earthkit-data](https://github.com/ecmwf/earthkit-data) to download files from the CDS.\nIf you intend to run this notebook multiple times, it is highly recommended that you [enable caching](https://earthkit-data.readthedocs.io/en/latest/concepts/caching.html) to prevent having to download the same files multiple times.\nIf you prefer not to use earthkit, the following requests can also be used with the [cdsapi module](https://cds.climate.copernicus.eu/how-to-api#linux-use-client-step).\nIn either case (earthkit-data or cdsapi), it is required to set up a CDS account and API key as explained [on the CDS website](https://cds.climate.copernicus.eu/how-to-api).\n\nThe first step is to define the parameters that will be shared between the download of the origin dataset (here CMIP6) and the C3S Atlas dataset (in the [next section](section-c3s-atlas)), namely the experiment and model member.\nNext, the request to download the corresponding data from CMIP6 is defined,\nin this notebook choosing only the variable necessary to calculate the `tx35` indicator, namely `daily_maximum_near_surface_air_temperature`.\nBecause the indicator will be calculated for multiple years, the `year` parameter of the request is left out for now.\n\nDefine request\n\nThis case study examines\ntwo years, namely 2060 and 2080, so there are two\ncorresponding requests.\nThis can easily be changed or extended to use different years.\nSeparate requests are defined for each year,\nrather than one combined request,\nto limit the size per request.\n\nDownload data\n\nEarthkit downloads the data as a field list.\nThis is converted into an `xarray` dataset, which can be used in the following analysis."} {"chunk_id": "derived_multi-origin-c3s-atlas_consistency_q01__2f1ad31bb643", "report_id": "derived_multi-origin-c3s-atlas_consistency_q01", "dataset_id": "multi-origin-c3s-atlas", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q01", "aspect_base": "consistency", "category": "Derived_Datasets", "match_confidence": "exact", "section": "Consistency between the C3S Atlas dataset and its origins: Case study > Analysis and results > 2. Calculate indicator from the origin dataset > Download data", "title": "Consistency between the C3S Atlas dataset and its origins: Case study", "chunk_index": 10, "token_count": 622, "text_raw": "as a field list.\nThis is converted into an `xarray` dataset, which can be used in the following analysis.\n\n```text\n Size: 167MB\nDimensions: (time: 730, bnds: 2, lat: 192, lon: 288)\nCoordinates:\n * time (time) object 6kB 2060-01-01 12:00:00 ... 2080-12-31 12:00:00\n * lat (lat) float64 2kB -90.0 -89.06 -88.12 -87.17 ... 88.12 89.06 90.0\n * lon (lon) float64 2kB 0.0 1.25 2.5 3.75 ... 355.0 356.2 357.5 358.8\n height float64 8B 2.0\nDimensions without coordinates: bnds\nData variables:\n time_bnds (time, bnds) object 12kB dask.array\n lat_bnds (time, lat, bnds) float64 2MB dask.array\n lon_bnds (time, lon, bnds) float64 3MB dask.array\n tasmax (time, lat, lon) float32 161MB dask.array\nAttributes: (12/48)\n Conventions: CF-1.7 CMIP-6.2\n activity_id: ScenarioMIP\n branch_method: standard\n branch_time_in_child: 60225.0\n branch_time_in_parent: 60225.0\n comment: none\n ... ...\n title: CMCC-ESM2 output prepared for CMIP6\n variable_id: tasmax\n variant_label: r1i1p1f1\n license: CMIP6 model data produced by CMCC is licensed und...\n cmor_version: 3.6.0\n tracking_id: hdl:21.14100/a1513d6f-5325-4b99-bbe6-9b096557b100\n```", "text_with_prefix": "EQC Quality Assessment: \"Consistency between the C3S Atlas dataset and its origins: Case study\"\nDataset: multi-origin-c3s-atlas [CDS]\nAspect: consistency_q01 | Category: Derived_Datasets\nSection: Consistency between the C3S Atlas dataset and its origins: Case study > Analysis and results > 2. Calculate indicator from the origin dataset > Download data\n---\nas a field list.\nThis is converted into an `xarray` dataset, which can be used in the following analysis.\n\n```text\n Size: 167MB\nDimensions: (time: 730, bnds: 2, lat: 192, lon: 288)\nCoordinates:\n * time (time) object 6kB 2060-01-01 12:00:00 ... 2080-12-31 12:00:00\n * lat (lat) float64 2kB -90.0 -89.06 -88.12 -87.17 ... 88.12 89.06 90.0\n * lon (lon) float64 2kB 0.0 1.25 2.5 3.75 ... 355.0 356.2 357.5 358.8\n height float64 8B 2.0\nDimensions without coordinates: bnds\nData variables:\n time_bnds (time, bnds) object 12kB dask.array\n lat_bnds (time, lat, bnds) float64 2MB dask.array\n lon_bnds (time, lon, bnds) float64 3MB dask.array\n tasmax (time, lat, lon) float32 161MB dask.array\nAttributes: (12/48)\n Conventions: CF-1.7 CMIP-6.2\n activity_id: ScenarioMIP\n branch_method: standard\n branch_time_in_child: 60225.0\n branch_time_in_parent: 60225.0\n comment: none\n ... ...\n title: CMCC-ESM2 output prepared for CMIP6\n variable_id: tasmax\n variant_label: r1i1p1f1\n license: CMIP6 model data produced by CMCC is licensed und...\n cmor_version: 3.6.0\n tracking_id: hdl:21.14100/a1513d6f-5325-4b99-bbe6-9b096557b100\n```"} {"chunk_id": "derived_multi-origin-c3s-atlas_consistency_q01__84e5e4bed626", "report_id": "derived_multi-origin-c3s-atlas_consistency_q01", "dataset_id": "multi-origin-c3s-atlas", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q01", "aspect_base": "consistency", "category": "Derived_Datasets", "match_confidence": "exact", "section": "Consistency between the C3S Atlas dataset and its origins: Case study > Analysis and results > 2. Calculate indicator from the origin dataset > Homogenise data", "title": "Consistency between the C3S Atlas dataset and its origins: Case study", "chunk_index": 11, "token_count": 300, "text_raw": "One of the steps in the C3S Atlas dataset production chain is homogenisation, i.e. ensuring consistency between data from different origin datasets.\nThis homogenisation is implemented in the [User-tools for the C3S Atlas](https://github.com/ecmwf-projects/c3s-atlas/tree/main/c3s_atlas), specifically the `c3s_atlas.fixers.apply_fixers` function.\nThe following changes are applied:\n\n- The names of the spatial coordinates are standardised to `[lon, lat]`.\n- Longitude is converted from `[0...360]` to `[-180...180]` format.\n- The time coordinate is standardised to the CF standard calendar.\n- Variable units are standardised (e.g. °C for temperature).\n- Variables are resampled / aggregated to the required temporal resolution.\n\nThe homogenisation is applied in the following code cell.\nThe `apply_fixers` function describes the different homogenisation steps as it applies them;\nthis can be read by expanding the following cell outputs.", "text_with_prefix": "EQC Quality Assessment: \"Consistency between the C3S Atlas dataset and its origins: Case study\"\nDataset: multi-origin-c3s-atlas [CDS]\nAspect: consistency_q01 | Category: Derived_Datasets\nSection: Consistency between the C3S Atlas dataset and its origins: Case study > Analysis and results > 2. Calculate indicator from the origin dataset > Homogenise data\n---\nOne of the steps in the C3S Atlas dataset production chain is homogenisation, i.e. ensuring consistency between data from different origin datasets.\nThis homogenisation is implemented in the [User-tools for the C3S Atlas](https://github.com/ecmwf-projects/c3s-atlas/tree/main/c3s_atlas), specifically the `c3s_atlas.fixers.apply_fixers` function.\nThe following changes are applied:\n\n- The names of the spatial coordinates are standardised to `[lon, lat]`.\n- Longitude is converted from `[0...360]` to `[-180...180]` format.\n- The time coordinate is standardised to the CF standard calendar.\n- Variable units are standardised (e.g. °C for temperature).\n- Variables are resampled / aggregated to the required temporal resolution.\n\nThe homogenisation is applied in the following code cell.\nThe `apply_fixers` function describes the different homogenisation steps as it applies them;\nthis can be read by expanding the following cell outputs."} {"chunk_id": "derived_multi-origin-c3s-atlas_consistency_q01__1995a91f4ad4", "report_id": "derived_multi-origin-c3s-atlas_consistency_q01", "dataset_id": "multi-origin-c3s-atlas", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q01", "aspect_base": "consistency", "category": "Derived_Datasets", "match_confidence": "exact", "section": "Consistency between the C3S Atlas dataset and its origins: Case study > Analysis and results > 2. Calculate indicator from the origin dataset > Homogenise data", "title": "Consistency between the C3S Atlas dataset and its origins: Case study", "chunk_index": 12, "token_count": 1097, "text_raw": "can be read by expanding the following cell outputs.\n\n```text\n2025-09-30 20:57:27,598 — Homogenization-fixers — INFO — Dataset has already the correct names for its coordinates\n2025-09-30 20:57:27,609 — Homogenization-fixers — INFO — Fixing calendar for Size: 84MB\nDimensions: (time: 365, bnds: 2, lat: 192, lon: 288)\nCoordinates:\n * time (time) object 3kB 2060-01-01 12:00:00 ... 2060-12-31 12:00:00\n * lat (lat) float64 2kB -90.0 -89.06 -88.12 -87.17 ... 88.12 89.06 90.0\n * lon (lon) float64 2kB 0.0 1.25 2.5 3.75 ... 355.0 356.2 357.5 358.8\n height float64 8B 2.0\nDimensions without coordinates: bnds\nData variables:\n time_bnds (time, bnds) object 6kB dask.array\n lat_bnds (time, lat, bnds) float64 1MB dask.array\n lon_bnds (time, lon, bnds) float64 2MB dask.array\n tasmax (time, lat, lon) float32 81MB dask.array\nAttributes: (12/48)\n Conventions: CF-1.7 CMIP-6.2\n activity_id: ScenarioMIP\n branch_method: standard\n branch_time_in_child: 60225.0\n branch_time_in_parent: 60225.0\n comment: none\n ... ...\n title: CMCC-ESM2 output prepared for CMIP6\n variable_id: tasmax\n variant_label: r1i1p1f1\n license: CMIP6 model data produced by CMCC is licensed und...\n cmor_version: 3.6.0\n tracking_id: hdl:21.14100/a1513d6f-5325-4b99-bbe6-9b096557b100\n2025-09-30 20:57:27,877 — UNITS_TRANSFORM — INFO — The dataset tasmax units are not in the correct magnitude. A conversion from K to Celsius will be performed.\n2025-09-30 20:57:27,962 — Homogenization-fixers — INFO — The dataset is in daily or monthly resolution, we don't need to resample it from hourly frequency\n2025-09-30 20:57:27,999 — Homogenization-fixers — INFO — Dataset has already the correct names for its coordinates\n2025-09-30 20:57:28,005 — Homogenization-fixers — INFO — Fixing calendar for Size: 84MB\nDimensions: (time: 365, bnds: 2, lat: 192, lon: 288)\nCoordinates:\n * time (time) object 3kB 2080-01-01 12:00:00 ... 2080-12-31 12:00:00\n * lat (lat) float64 2kB -90.0 -89.06 -88.12 -87.17 ... 88.12 89.06 90.0\n * lon (lon) float64 2kB 0.0 1.25 2.5 3.75 ... 355.0 356.2 357.5 358.8\n height float64 8B 2.0\nDimensions without coordinates: bnds\nData variables:\n time_bnds (time, bnds) object 6kB dask.array\n lat_bnds (time, lat, bnds) float64 1MB dask.array\n lon_bnds (time, lon, bnds) float64 2MB dask.array Analysis and results > 2. Calculate indicator from the origin dataset > Homogenise data\n---\ncan be read by expanding the following cell outputs.\n\n```text\n2025-09-30 20:57:27,598 — Homogenization-fixers — INFO — Dataset has already the correct names for its coordinates\n2025-09-30 20:57:27,609 — Homogenization-fixers — INFO — Fixing calendar for Size: 84MB\nDimensions: (time: 365, bnds: 2, lat: 192, lon: 288)\nCoordinates:\n * time (time) object 3kB 2060-01-01 12:00:00 ... 2060-12-31 12:00:00\n * lat (lat) float64 2kB -90.0 -89.06 -88.12 -87.17 ... 88.12 89.06 90.0\n * lon (lon) float64 2kB 0.0 1.25 2.5 3.75 ... 355.0 356.2 357.5 358.8\n height float64 8B 2.0\nDimensions without coordinates: bnds\nData variables:\n time_bnds (time, bnds) object 6kB dask.array\n lat_bnds (time, lat, bnds) float64 1MB dask.array\n lon_bnds (time, lon, bnds) float64 2MB dask.array\n tasmax (time, lat, lon) float32 81MB dask.array\nAttributes: (12/48)\n Conventions: CF-1.7 CMIP-6.2\n activity_id: ScenarioMIP\n branch_method: standard\n branch_time_in_child: 60225.0\n branch_time_in_parent: 60225.0\n comment: none\n ... ...\n title: CMCC-ESM2 output prepared for CMIP6\n variable_id: tasmax\n variant_label: r1i1p1f1\n license: CMIP6 model data produced by CMCC is licensed und...\n cmor_version: 3.6.0\n tracking_id: hdl:21.14100/a1513d6f-5325-4b99-bbe6-9b096557b100\n2025-09-30 20:57:27,877 — UNITS_TRANSFORM — INFO — The dataset tasmax units are not in the correct magnitude. A conversion from K to Celsius will be performed.\n2025-09-30 20:57:27,962 — Homogenization-fixers — INFO — The dataset is in daily or monthly resolution, we don't need to resample it from hourly frequency\n2025-09-30 20:57:27,999 — Homogenization-fixers — INFO — Dataset has already the correct names for its coordinates\n2025-09-30 20:57:28,005 — Homogenization-fixers — INFO — Fixing calendar for Size: 84MB\nDimensions: (time: 365, bnds: 2, lat: 192, lon: 288)\nCoordinates:\n * time (time) object 3kB 2080-01-01 12:00:00 ... 2080-12-31 12:00:00\n * lat (lat) float64 2kB -90.0 -89.06 -88.12 -87.17 ... 88.12 89.06 90.0\n * lon (lon) float64 2kB 0.0 1.25 2.5 3.75 ... 355.0 356.2 357.5 358.8\n height float64 8B 2.0\nDimensions without coordinates: bnds\nData variables:\n time_bnds (time, bnds) object 6kB dask.array\n lat_bnds (time, lat, bnds) float64 1MB dask.array\n lon_bnds (time, lon, bnds) float64 2MB dask.array Analysis and results > 2. Calculate indicator from the origin dataset > Homogenise data", "title": "Consistency between the C3S Atlas dataset and its origins: Case study", "chunk_index": 13, "token_count": 454, "text_raw": ") float64 1MB dask.array\n lon_bnds (time, lon, bnds) float64 2MB dask.array\n tasmax (time, lat, lon) float32 81MB dask.array\nAttributes: (12/48)\n Conventions: CF-1.7 CMIP-6.2\n activity_id: ScenarioMIP\n branch_method: standard\n branch_time_in_child: 60225.0\n branch_time_in_parent: 60225.0\n comment: none\n ... ...\n title: CMCC-ESM2 output prepared for CMIP6\n variable_id: tasmax\n variant_label: r1i1p1f1\n license: CMIP6 model data produced by CMCC is licensed und...\n cmor_version: 3.6.0\n tracking_id: hdl:21.14100/a1513d6f-5325-4b99-bbe6-9b096557b100\n2025-09-30 20:57:28,258 — UNITS_TRANSFORM — INFO — The dataset tasmax units are not in the correct magnitude. A conversion from K to Celsius will be performed.\n2025-09-30 20:57:28,384 — Homogenization-fixers — INFO — The dataset is in daily or monthly resolution, we don't need to resample it from hourly frequency\n```", "text_with_prefix": "EQC Quality Assessment: \"Consistency between the C3S Atlas dataset and its origins: Case study\"\nDataset: multi-origin-c3s-atlas [CDS]\nAspect: consistency_q01 | Category: Derived_Datasets\nSection: Consistency between the C3S Atlas dataset and its origins: Case study > Analysis and results > 2. Calculate indicator from the origin dataset > Homogenise data\n---\n) float64 1MB dask.array\n lon_bnds (time, lon, bnds) float64 2MB dask.array\n tasmax (time, lat, lon) float32 81MB dask.array\nAttributes: (12/48)\n Conventions: CF-1.7 CMIP-6.2\n activity_id: ScenarioMIP\n branch_method: standard\n branch_time_in_child: 60225.0\n branch_time_in_parent: 60225.0\n comment: none\n ... ...\n title: CMCC-ESM2 output prepared for CMIP6\n variable_id: tasmax\n variant_label: r1i1p1f1\n license: CMIP6 model data produced by CMCC is licensed und...\n cmor_version: 3.6.0\n tracking_id: hdl:21.14100/a1513d6f-5325-4b99-bbe6-9b096557b100\n2025-09-30 20:57:28,258 — UNITS_TRANSFORM — INFO — The dataset tasmax units are not in the correct magnitude. A conversion from K to Celsius will be performed.\n2025-09-30 20:57:28,384 — Homogenization-fixers — INFO — The dataset is in daily or monthly resolution, we don't need to resample it from hourly frequency\n```"} {"chunk_id": "derived_multi-origin-c3s-atlas_consistency_q01__7454b4c5ce28", "report_id": "derived_multi-origin-c3s-atlas_consistency_q01", "dataset_id": "multi-origin-c3s-atlas", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q01", "aspect_base": "consistency", "category": "Derived_Datasets", "match_confidence": "exact", "section": "Consistency between the C3S Atlas dataset and its origins: Case study > Analysis and results > 2. Calculate indicator from the origin dataset > Homogenise data", "title": "Consistency between the C3S Atlas dataset and its origins: Case study", "chunk_index": 14, "token_count": 492, "text_raw": "resolution, we don't need to resample it from hourly frequency\n```\n\n```text\n Size: 162MB\nDimensions: (time: 732, lat: 192, lon: 288)\nCoordinates:\n * lat (lat) float64 2kB -90.0 -89.06 -88.12 -87.17 ... 88.12 89.06 90.0\n * lon (lon) float64 2kB -178.8 -177.5 -176.2 -175.0 ... 177.5 178.8 180.0\n * time (time) datetime64[ns] 6kB 2060-01-01 2060-01-02 ... 2080-12-31\n height float64 8B 2.0\nData variables:\n tasmax (time, lat, lon) float32 162MB dask.array\nAttributes: (12/48)\n Conventions: CF-1.7 CMIP-6.2\n activity_id: ScenarioMIP\n branch_method: standard\n branch_time_in_child: 60225.0\n branch_time_in_parent: 60225.0\n comment: none\n ... ...\n title: CMCC-ESM2 output prepared for CMIP6\n variable_id: tasmax\n variant_label: r1i1p1f1\n license: CMIP6 model data produced by CMCC is licensed und...\n cmor_version: 3.6.0\n tracking_id: hdl:21.14100/a1513d6f-5325-4b99-bbe6-9b096557b100\n```", "text_with_prefix": "EQC Quality Assessment: \"Consistency between the C3S Atlas dataset and its origins: Case study\"\nDataset: multi-origin-c3s-atlas [CDS]\nAspect: consistency_q01 | Category: Derived_Datasets\nSection: Consistency between the C3S Atlas dataset and its origins: Case study > Analysis and results > 2. Calculate indicator from the origin dataset > Homogenise data\n---\nresolution, we don't need to resample it from hourly frequency\n```\n\n```text\n Size: 162MB\nDimensions: (time: 732, lat: 192, lon: 288)\nCoordinates:\n * lat (lat) float64 2kB -90.0 -89.06 -88.12 -87.17 ... 88.12 89.06 90.0\n * lon (lon) float64 2kB -178.8 -177.5 -176.2 -175.0 ... 177.5 178.8 180.0\n * time (time) datetime64[ns] 6kB 2060-01-01 2060-01-02 ... 2080-12-31\n height float64 8B 2.0\nData variables:\n tasmax (time, lat, lon) float32 162MB dask.array\nAttributes: (12/48)\n Conventions: CF-1.7 CMIP-6.2\n activity_id: ScenarioMIP\n branch_method: standard\n branch_time_in_child: 60225.0\n branch_time_in_parent: 60225.0\n comment: none\n ... ...\n title: CMCC-ESM2 output prepared for CMIP6\n variable_id: tasmax\n variant_label: r1i1p1f1\n license: CMIP6 model data produced by CMCC is licensed und...\n cmor_version: 3.6.0\n tracking_id: hdl:21.14100/a1513d6f-5325-4b99-bbe6-9b096557b100\n```"} {"chunk_id": "derived_multi-origin-c3s-atlas_consistency_q01__cf2c8c2f46d3", "report_id": "derived_multi-origin-c3s-atlas_consistency_q01", "dataset_id": "multi-origin-c3s-atlas", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q01", "aspect_base": "consistency", "category": "Derived_Datasets", "match_confidence": "exact", "section": "Consistency between the C3S Atlas dataset and its origins: Case study > Analysis and results > 2. Calculate indicator from the origin dataset > Calculate indicator", "title": "Consistency between the C3S Atlas dataset and its origins: Case study", "chunk_index": 15, "token_count": 366, "text_raw": "The climate indicator,\nin this notebook only `tx35`,\nis calculated using [xclim](https://xclim.readthedocs.io/en/stable/).\nThe function `cal_tx35`, defined [above](section-codesetup), performs the calculation.\nThis function needs to be applied year-by-year, which is achieved using `groupby`.\n\n```text\n Size: 11MB\nDimensions: (time: 24, lat: 192, lon: 288)\nCoordinates:\n * lat (lat) float64 2kB -90.0 -89.06 -88.12 -87.17 ... 88.12 89.06 90.0\n * lon (lon) float64 2kB -178.8 -177.5 -176.2 -175.0 ... 177.5 178.8 180.0\n height float64 8B 2.0\n * time (time) datetime64[ns] 192B 2060-01-01 2060-02-01 ... 2080-12-01\nData variables:\n tx35 (time, lat, lon) int64 11MB dask.array\n```", "text_with_prefix": "EQC Quality Assessment: \"Consistency between the C3S Atlas dataset and its origins: Case study\"\nDataset: multi-origin-c3s-atlas [CDS]\nAspect: consistency_q01 | Category: Derived_Datasets\nSection: Consistency between the C3S Atlas dataset and its origins: Case study > Analysis and results > 2. Calculate indicator from the origin dataset > Calculate indicator\n---\nThe climate indicator,\nin this notebook only `tx35`,\nis calculated using [xclim](https://xclim.readthedocs.io/en/stable/).\nThe function `cal_tx35`, defined [above](section-codesetup), performs the calculation.\nThis function needs to be applied year-by-year, which is achieved using `groupby`.\n\n```text\n Size: 11MB\nDimensions: (time: 24, lat: 192, lon: 288)\nCoordinates:\n * lat (lat) float64 2kB -90.0 -89.06 -88.12 -87.17 ... 88.12 89.06 90.0\n * lon (lon) float64 2kB -178.8 -177.5 -176.2 -175.0 ... 177.5 178.8 180.0\n height float64 8B 2.0\n * time (time) datetime64[ns] 192B 2060-01-01 2060-02-01 ... 2080-12-01\nData variables:\n tx35 (time, lat, lon) int64 11MB dask.array\n```"} {"chunk_id": "derived_multi-origin-c3s-atlas_consistency_q01__923c6a9af281", "report_id": "derived_multi-origin-c3s-atlas_consistency_q01", "dataset_id": "multi-origin-c3s-atlas", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q01", "aspect_base": "consistency", "category": "Derived_Datasets", "match_confidence": "exact", "section": "Consistency between the C3S Atlas dataset and its origins: Case study > Analysis and results > 2. Calculate indicator from the origin dataset > Regrid to C3S Atlas grid", "title": "Consistency between the C3S Atlas dataset and its origins: Case study", "chunk_index": 16, "token_count": 822, "text_raw": "This notebook uses [xESMF](https://github.com/pangeo-data/xESMF) for regridding data.\nxESMF is most easily installed using mamba/conda as explained in its documentation.\nUsers who cannot or do not wish to use mamba/conda can manually compile and install [ESMF](https://earthsystemmodeling.org/docs/release/latest/ESMF_usrdoc/node10.html) on their machines.\nIn future, this notebook will use [earthkit-regrid](https://github.com/ecmwf/earthkit-regrid) instead, once it reaches suitable maturity.\n```\n\nThe final step in the processing is regridding to the standardised grid used in the C3S Atlas dataset (Figure {numref}`{number} `).\nThis is performed through a custom function in the [User-tools for the C3S Atlas](https://github.com/ecmwf-projects/c3s-atlas/tree/main/c3s_atlas),\nspecifically `c3s_atlas.interpolation`.\nThis function is based on ESMF, as noted above.\n\nNote that the C3S Atlas workflow calculates indicators first, then regrids.\nFor operations that involve averaging, like smoothing and regridding, the order of operations can affect the result, especially in areas with steep gradients [[Avila+15](https://doi.org/10.1016/j.wace.2015.06.003)].\nExamples of such areas for a temperature index are coastlines and mountain ranges.\nIn the case of C3S Atlas, this order of operations was a conscious choice to preserve the \"raw\" signals,\ne.g. preventing extreme temperatures from being smoothed out.\nHowever, it can affect the indicator values and therefore must be considered when using the C3S Atlas application or dataset,\nparticularly when intercomparing it with another dataset.\n\nApply regridding function\n\n```text\n Size: 12MB\nDimensions: (lon: 360, lat: 180, time: 24, bnds: 2)\nCoordinates:\n * lon (lon) float64 3kB -179.5 -178.5 -177.5 ... 177.5 178.5 179.5\n * lat (lat) float64 1kB -89.5 -88.5 -87.5 -86.5 ... 86.5 87.5 88.5 89.5\n * time (time) datetime64[ns] 192B 2060-01-01 2060-02-01 ... 2080-12-01\nDimensions without coordinates: bnds\nData variables:\n tx35 (time, lat, lon) int64 12MB dask.array\n lon_bnds (lon, bnds) float64 6kB -180.0 -179.0 -179.0 ... 179.0 179.0 180.0\n lat_bnds (lat, bnds) float64 3kB -90.0 -89.0 -89.0 -88.0 ... 89.0 89.0 90.0\n crs int64 8B 0\n height float64 8B 2.0\n```\n\n(section-c3s-atlas)=", "text_with_prefix": "EQC Quality Assessment: \"Consistency between the C3S Atlas dataset and its origins: Case study\"\nDataset: multi-origin-c3s-atlas [CDS]\nAspect: consistency_q01 | Category: Derived_Datasets\nSection: Consistency between the C3S Atlas dataset and its origins: Case study > Analysis and results > 2. Calculate indicator from the origin dataset > Regrid to C3S Atlas grid\n---\nThis notebook uses [xESMF](https://github.com/pangeo-data/xESMF) for regridding data.\nxESMF is most easily installed using mamba/conda as explained in its documentation.\nUsers who cannot or do not wish to use mamba/conda can manually compile and install [ESMF](https://earthsystemmodeling.org/docs/release/latest/ESMF_usrdoc/node10.html) on their machines.\nIn future, this notebook will use [earthkit-regrid](https://github.com/ecmwf/earthkit-regrid) instead, once it reaches suitable maturity.\n```\n\nThe final step in the processing is regridding to the standardised grid used in the C3S Atlas dataset (Figure {numref}`{number} `).\nThis is performed through a custom function in the [User-tools for the C3S Atlas](https://github.com/ecmwf-projects/c3s-atlas/tree/main/c3s_atlas),\nspecifically `c3s_atlas.interpolation`.\nThis function is based on ESMF, as noted above.\n\nNote that the C3S Atlas workflow calculates indicators first, then regrids.\nFor operations that involve averaging, like smoothing and regridding, the order of operations can affect the result, especially in areas with steep gradients [[Avila+15](https://doi.org/10.1016/j.wace.2015.06.003)].\nExamples of such areas for a temperature index are coastlines and mountain ranges.\nIn the case of C3S Atlas, this order of operations was a conscious choice to preserve the \"raw\" signals,\ne.g. preventing extreme temperatures from being smoothed out.\nHowever, it can affect the indicator values and therefore must be considered when using the C3S Atlas application or dataset,\nparticularly when intercomparing it with another dataset.\n\nApply regridding function\n\n```text\n Size: 12MB\nDimensions: (lon: 360, lat: 180, time: 24, bnds: 2)\nCoordinates:\n * lon (lon) float64 3kB -179.5 -178.5 -177.5 ... 177.5 178.5 179.5\n * lat (lat) float64 1kB -89.5 -88.5 -87.5 -86.5 ... 86.5 87.5 88.5 89.5\n * time (time) datetime64[ns] 192B 2060-01-01 2060-02-01 ... 2080-12-01\nDimensions without coordinates: bnds\nData variables:\n tx35 (time, lat, lon) int64 12MB dask.array\n lon_bnds (lon, bnds) float64 6kB -180.0 -179.0 -179.0 ... 179.0 179.0 180.0\n lat_bnds (lat, bnds) float64 3kB -90.0 -89.0 -89.0 -88.0 ... 89.0 89.0 90.0\n crs int64 8B 0\n height float64 8B 2.0\n```\n\n(section-c3s-atlas)="} {"chunk_id": "derived_multi-origin-c3s-atlas_consistency_q01__c62673abc47e", "report_id": "derived_multi-origin-c3s-atlas_consistency_q01", "dataset_id": "multi-origin-c3s-atlas", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q01", "aspect_base": "consistency", "category": "Derived_Datasets", "match_confidence": "exact", "section": "Consistency between the C3S Atlas dataset and its origins: Case study > Analysis and results > 3. Retrieve indicator from the C3S Atlas dataset", "title": "Consistency between the C3S Atlas dataset and its origins: Case study", "chunk_index": 17, "token_count": 804, "text_raw": "Here, we download the same indicator as above directly from the [Gridded dataset underpinning the Copernicus Interactive Climate Atlas](https://doi.org/10.24381/cds.h35hb680) so the values can be compared.\n\nDefine request\nDownload data\n\nThe C3S Atlas dataset provides all years and members at the same time, so for convenience we pull out only the relevant entries:\n\n```text\n Size: 6MB\nDimensions: (lat: 180, bnds: 2, lon: 360, time: 24)\nCoordinates:\n * lat (lat) float64 1kB -89.5 -88.5 -87.5 ... 87.5 88.5 89.5\n * lon (lon) float64 3kB -179.5 -178.5 -177.5 ... 178.5 179.5\n * time (time) datetime64[ns] 192B 2060-01-01 ... 2080-12-01\n member_id \n gcm_institution \n gcm_model \n gcm_variant \n threshold35c float64 8B ...\n height2m float64 8B ...\nDimensions without coordinates: bnds\nData variables:\n lat_bnds (lat, bnds) float64 3kB dask.array\n lon_bnds (lon, bnds) float64 6kB dask.array\n time_bnds (time, bnds) datetime64[ns] 384B dask.array\n tx35 (time, lat, lon) float32 6MB dask.array\n crs int32 4B ...\nAttributes: (12/26)\n Conventions: CF-1.9 ACDD-1.3\n title: Copernicus Interactive Climate Atlas: gridded...\n summary: Monthly/annual gridded data from observations...\n institution: Copernicus Climate Change Service (C3S)\n producers: Institute of Physics of Cantabria (IFCA, CSIC...\n license: CC-BY 4.0, https://creativecommons.org/licens...\n ... ...\n geospatial_lon_min: -180.0\n geospatial_lon_max: 180.0\n geospatial_lon_resolution: 1.0\n geospatial_lon_units: degrees_east\n date_created: 2024-12-05 16:37:49.749769+01:00\n tracking_id: 1a7a60e7-7787-48b5-b18f-a7bf7b4de4be\n```\n\n(section-results)=", "text_with_prefix": "EQC Quality Assessment: \"Consistency between the C3S Atlas dataset and its origins: Case study\"\nDataset: multi-origin-c3s-atlas [CDS]\nAspect: consistency_q01 | Category: Derived_Datasets\nSection: Consistency between the C3S Atlas dataset and its origins: Case study > Analysis and results > 3. Retrieve indicator from the C3S Atlas dataset\n---\nHere, we download the same indicator as above directly from the [Gridded dataset underpinning the Copernicus Interactive Climate Atlas](https://doi.org/10.24381/cds.h35hb680) so the values can be compared.\n\nDefine request\nDownload data\n\nThe C3S Atlas dataset provides all years and members at the same time, so for convenience we pull out only the relevant entries:\n\n```text\n Size: 6MB\nDimensions: (lat: 180, bnds: 2, lon: 360, time: 24)\nCoordinates:\n * lat (lat) float64 1kB -89.5 -88.5 -87.5 ... 87.5 88.5 89.5\n * lon (lon) float64 3kB -179.5 -178.5 -177.5 ... 178.5 179.5\n * time (time) datetime64[ns] 192B 2060-01-01 ... 2080-12-01\n member_id \n gcm_institution \n gcm_model \n gcm_variant \n threshold35c float64 8B ...\n height2m float64 8B ...\nDimensions without coordinates: bnds\nData variables:\n lat_bnds (lat, bnds) float64 3kB dask.array\n lon_bnds (lon, bnds) float64 6kB dask.array\n time_bnds (time, bnds) datetime64[ns] 384B dask.array\n tx35 (time, lat, lon) float32 6MB dask.array\n crs int32 4B ...\nAttributes: (12/26)\n Conventions: CF-1.9 ACDD-1.3\n title: Copernicus Interactive Climate Atlas: gridded...\n summary: Monthly/annual gridded data from observations...\n institution: Copernicus Climate Change Service (C3S)\n producers: Institute of Physics of Cantabria (IFCA, CSIC...\n license: CC-BY 4.0, https://creativecommons.org/licens...\n ... ...\n geospatial_lon_min: -180.0\n geospatial_lon_max: 180.0\n geospatial_lon_resolution: 1.0\n geospatial_lon_units: degrees_east\n date_created: 2024-12-05 16:37:49.749769+01:00\n tracking_id: 1a7a60e7-7787-48b5-b18f-a7bf7b4de4be\n```\n\n(section-results)="} {"chunk_id": "derived_multi-origin-c3s-atlas_consistency_q01__eb73512fcbc6", "report_id": "derived_multi-origin-c3s-atlas_consistency_q01", "dataset_id": "multi-origin-c3s-atlas", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q01", "aspect_base": "consistency", "category": "Derived_Datasets", "match_confidence": "exact", "section": "Consistency between the C3S Atlas dataset and its origins: Case study > Analysis and results > 4. Results", "title": "Consistency between the C3S Atlas dataset and its origins: Case study", "chunk_index": 18, "token_count": 451, "text_raw": "This section contains the comparison between the indicator values retrieved from the C3S Atlas dataset vs those reproduced from the origin dataset.\n\nThe datasets are first compared on their native grids.\nThis means a point-by-point comparison is not possible\n(because the points are not equivalent),\nbut the distributions can be compared geospatially and overall.\nThis qualitative comparison probes the consistency quality attribute:\nAre the climate indicators in the dataset underpinning the Copernicus Interactive Climate Atlas consistent with their origin datasets?\n\nSecond, the C3S Atlas dataset is compared to the indicators derived from the origin dataset and regridded to the C3S Atlas grid.\nThis makes a quantitative point-by-point comparison possible.\nThis second comparison probes how well the dataset underpinning the Copernicus Interactive Climate Atlas can be reproduced from its origin datasets,\nbased on the workflow (Figure {numref}`{number} `).\n\nFor the geospatial comparison,\nwe display the values of the indicator for one month,\nacross one region and globally.\nFor the example of `tx35`,\nJune\nshould provide many warm days in the Northern hemisphere with significant spatial variation.\n\nAs an example, we display the results across\nEurope.\nThis region can easily be modified in the following code cell using the [domains provided by earthkit-plots](https://earthkit-plots.readthedocs.io/en/latest/examples/examples/introduction/07-domains.html).\nSome examples are provided in the cell (commented out using `#`).\n\nSetup: Choose a month to display\nSetup: Pick domain using earthkit-plots\nOther examples -- uncomment where desired\ndomain = ekp.geo.domains.union([\"Portugal\", \"Spain\"], name=\"Iberia\")\ndomain = \"Italy\"\ndomain = \"South America\"", "text_with_prefix": "EQC Quality Assessment: \"Consistency between the C3S Atlas dataset and its origins: Case study\"\nDataset: multi-origin-c3s-atlas [CDS]\nAspect: consistency_q01 | Category: Derived_Datasets\nSection: Consistency between the C3S Atlas dataset and its origins: Case study > Analysis and results > 4. Results\n---\nThis section contains the comparison between the indicator values retrieved from the C3S Atlas dataset vs those reproduced from the origin dataset.\n\nThe datasets are first compared on their native grids.\nThis means a point-by-point comparison is not possible\n(because the points are not equivalent),\nbut the distributions can be compared geospatially and overall.\nThis qualitative comparison probes the consistency quality attribute:\nAre the climate indicators in the dataset underpinning the Copernicus Interactive Climate Atlas consistent with their origin datasets?\n\nSecond, the C3S Atlas dataset is compared to the indicators derived from the origin dataset and regridded to the C3S Atlas grid.\nThis makes a quantitative point-by-point comparison possible.\nThis second comparison probes how well the dataset underpinning the Copernicus Interactive Climate Atlas can be reproduced from its origin datasets,\nbased on the workflow (Figure {numref}`{number} `).\n\nFor the geospatial comparison,\nwe display the values of the indicator for one month,\nacross one region and globally.\nFor the example of `tx35`,\nJune\nshould provide many warm days in the Northern hemisphere with significant spatial variation.\n\nAs an example, we display the results across\nEurope.\nThis region can easily be modified in the following code cell using the [domains provided by earthkit-plots](https://earthkit-plots.readthedocs.io/en/latest/examples/examples/introduction/07-domains.html).\nSome examples are provided in the cell (commented out using `#`).\n\nSetup: Choose a month to display\nSetup: Pick domain using earthkit-plots\nOther examples -- uncomment where desired\ndomain = ekp.geo.domains.union([\"Portugal\", \"Spain\"], name=\"Iberia\")\ndomain = \"Italy\"\ndomain = \"South America\""} {"chunk_id": "derived_multi-origin-c3s-atlas_consistency_q01__3e9cc88c041b", "report_id": "derived_multi-origin-c3s-atlas_consistency_q01", "dataset_id": "multi-origin-c3s-atlas", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q01", "aspect_base": "consistency", "category": "Derived_Datasets", "match_confidence": "exact", "section": "Consistency between the C3S Atlas dataset and its origins: Case study > Analysis and results > 4. Results > Consistency: Comparison on native grids", "title": "Consistency between the C3S Atlas dataset and its origins: Case study", "chunk_index": 19, "token_count": 1079, "text_raw": "It is clear from\nFigures {numref}`{number} ` and {numref}`{number} `\nthat the C3S Atlas dataset and its reproduction from the origin display very similar patterns,\nas expected.\nThe general distribution of indicator values\n(in both years)\nis the same,\nin this example showing more hot days in northern Africa, the Middle East, and eastern Europe than in western and northern Europe.\nSmall differences appear in various places,\nlikely the result of the regridding step in the C3S Atlas workflow as noted before.\nLarger differences can be seen in areas with steep temperature gradients,\nsuch as Italy with its mountains and coastlines.\nFigure {numref}`{number} `\nconfirms that the overall distributions in indicator values\n(across all time and space dimensions)\nare very similar,\nbut small differences exist.\n\nFor reference,\nFigure {numref}`{number} `\ndisplays the equivalent view\nin the C3S Atlas application.\nThe application export is visually similar to\nFigure {numref}`{number} `,\nshowing the consistency between application and dataset,\nand hence showing that the conclusions from this quality assessment transfer to the C3S Atlas application.\nDifferences are caused by the application displaying projection data as a multi-year average\n(here 2081–2100)\nas well as the difference in visualisation style.\n\nOverall, we can conclude that the C3S Atlas dataset and its origins are highly consistent,\nbut small differences exist due to the difference in grid.\nUsers of the C3S Atlas dataset\n– and thus users of the C3S Atlas application –\nshould be aware that the indicator values retrieved for a specific location may differ slightly from a manual analysis of the origin dataset.\nThis effect may be larger or smaller for values aggregated over larger areas,\nsuch as the country and IPCC-AR6 region averages available in the C3S Atlas application,\ndepending on the exact region and values used.\n\nLastly,\nit should be noted that the order of operations in the C3S Atlas workflow\n(regridding _after_ the indicator calculation;\nFigure {numref}`{number} `)\nresults in values that may differ from the opposite order of operations\n(regridding _before_ indicator calculation)\n[[Avila+15](https://doi.org/10.1016/j.wace.2015.06.003)].\nAssessing these differences is outside the scope of this assessment.\n\nCreate plot with function defined at the start\n\nderived_multi-origin-c3s-atlas_consistency_q01_fig-geo-regional\n:figwidth: 700px\n:name: \"derived_multi-origin-c3s-atlas_consistency_q01_fig-geo-regional\"\n\nComparison between C3S Atlas dataset and reproduction for one indicator (`tx35`) in one month,\nacross Europe,\non the native grid of each dataset.\n```\n\nderived_multi-origin-c3s-atlas_consistency_q01_fig-geo-global\n:figwidth: 700px\n:name: \"derived_multi-origin-c3s-atlas_consistency_q01_fig-geo-global\"\n\nComparison between C3S Atlas dataset and reproduction for one indicator (`tx35`) in one month,\nacross the globe,\non the native grid of each dataset.\n```\n\nattachment:c3s_atlas_cmip6_tx35_export.png\n---\nheight: 600px\nname: \"derived_multi-origin-c3s-atlas_consistency_q01_fig-geo-export\"\n---\nC3S Atlas application view of `tx35` indicator in June, averaged 2081–2100 in the SSP5-8.5 scenario,\nfor comparison with\nFigure {numref}`{number} `\n([permalink: YNZ3InLU](https://atlas.climate.copernicus.eu/atlas/YNZ3InLU)).\n```\n", "text_with_prefix": "EQC Quality Assessment: \"Consistency between the C3S Atlas dataset and its origins: Case study\"\nDataset: multi-origin-c3s-atlas [CDS]\nAspect: consistency_q01 | Category: Derived_Datasets\nSection: Consistency between the C3S Atlas dataset and its origins: Case study > Analysis and results > 4. Results > Consistency: Comparison on native grids\n---\nIt is clear from\nFigures {numref}`{number} ` and {numref}`{number} `\nthat the C3S Atlas dataset and its reproduction from the origin display very similar patterns,\nas expected.\nThe general distribution of indicator values\n(in both years)\nis the same,\nin this example showing more hot days in northern Africa, the Middle East, and eastern Europe than in western and northern Europe.\nSmall differences appear in various places,\nlikely the result of the regridding step in the C3S Atlas workflow as noted before.\nLarger differences can be seen in areas with steep temperature gradients,\nsuch as Italy with its mountains and coastlines.\nFigure {numref}`{number} `\nconfirms that the overall distributions in indicator values\n(across all time and space dimensions)\nare very similar,\nbut small differences exist.\n\nFor reference,\nFigure {numref}`{number} `\ndisplays the equivalent view\nin the C3S Atlas application.\nThe application export is visually similar to\nFigure {numref}`{number} `,\nshowing the consistency between application and dataset,\nand hence showing that the conclusions from this quality assessment transfer to the C3S Atlas application.\nDifferences are caused by the application displaying projection data as a multi-year average\n(here 2081–2100)\nas well as the difference in visualisation style.\n\nOverall, we can conclude that the C3S Atlas dataset and its origins are highly consistent,\nbut small differences exist due to the difference in grid.\nUsers of the C3S Atlas dataset\n– and thus users of the C3S Atlas application –\nshould be aware that the indicator values retrieved for a specific location may differ slightly from a manual analysis of the origin dataset.\nThis effect may be larger or smaller for values aggregated over larger areas,\nsuch as the country and IPCC-AR6 region averages available in the C3S Atlas application,\ndepending on the exact region and values used.\n\nLastly,\nit should be noted that the order of operations in the C3S Atlas workflow\n(regridding _after_ the indicator calculation;\nFigure {numref}`{number} `)\nresults in values that may differ from the opposite order of operations\n(regridding _before_ indicator calculation)\n[[Avila+15](https://doi.org/10.1016/j.wace.2015.06.003)].\nAssessing these differences is outside the scope of this assessment.\n\nCreate plot with function defined at the start\n\nderived_multi-origin-c3s-atlas_consistency_q01_fig-geo-regional\n:figwidth: 700px\n:name: \"derived_multi-origin-c3s-atlas_consistency_q01_fig-geo-regional\"\n\nComparison between C3S Atlas dataset and reproduction for one indicator (`tx35`) in one month,\nacross Europe,\non the native grid of each dataset.\n```\n\nderived_multi-origin-c3s-atlas_consistency_q01_fig-geo-global\n:figwidth: 700px\n:name: \"derived_multi-origin-c3s-atlas_consistency_q01_fig-geo-global\"\n\nComparison between C3S Atlas dataset and reproduction for one indicator (`tx35`) in one month,\nacross the globe,\non the native grid of each dataset.\n```\n\nattachment:c3s_atlas_cmip6_tx35_export.png\n---\nheight: 600px\nname: \"derived_multi-origin-c3s-atlas_consistency_q01_fig-geo-export\"\n---\nC3S Atlas application view of `tx35` indicator in June, averaged 2081–2100 in the SSP5-8.5 scenario,\nfor comparison with\nFigure {numref}`{number} `\n([permalink: YNZ3InLU](https://atlas.climate.copernicus.eu/atlas/YNZ3InLU)).\n```\n"} {"chunk_id": "derived_multi-origin-c3s-atlas_consistency_q01__8f02c02bedd8", "report_id": "derived_multi-origin-c3s-atlas_consistency_q01", "dataset_id": "multi-origin-c3s-atlas", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q01", "aspect_base": "consistency", "category": "Derived_Datasets", "match_confidence": "exact", "section": "Consistency between the C3S Atlas dataset and its origins: Case study > Analysis and results > 4. Results > Consistency: Comparison on native grids", "title": "Consistency between the C3S Atlas dataset and its origins: Case study", "chunk_index": 20, "token_count": 218, "text_raw": "term\n* Scenario: SSP5-8.5\n* Quantity: Climatology\n* Season: June\n* Drag colourbar max to 35.71\n* Projection: WGS 84\n* Download map as PNG\n-->\n\nderived_multi-origin-c3s-atlas_consistency_q01_fig-hist-native\n:figwidth: 700px\n:name: \"derived_multi-origin-c3s-atlas_consistency_q01_fig-hist-native\"\n\nComparison between overall distributions of indicator (`tx35`) values in the C3S Atlas dataset and its reproduction,\nacross all spatial and temporal dimensions,\non the native grid of each dataset.\n```", "text_with_prefix": "EQC Quality Assessment: \"Consistency between the C3S Atlas dataset and its origins: Case study\"\nDataset: multi-origin-c3s-atlas [CDS]\nAspect: consistency_q01 | Category: Derived_Datasets\nSection: Consistency between the C3S Atlas dataset and its origins: Case study > Analysis and results > 4. Results > Consistency: Comparison on native grids\n---\nterm\n* Scenario: SSP5-8.5\n* Quantity: Climatology\n* Season: June\n* Drag colourbar max to 35.71\n* Projection: WGS 84\n* Download map as PNG\n-->\n\nderived_multi-origin-c3s-atlas_consistency_q01_fig-hist-native\n:figwidth: 700px\n:name: \"derived_multi-origin-c3s-atlas_consistency_q01_fig-hist-native\"\n\nComparison between overall distributions of indicator (`tx35`) values in the C3S Atlas dataset and its reproduction,\nacross all spatial and temporal dimensions,\non the native grid of each dataset.\n```"} {"chunk_id": "derived_multi-origin-c3s-atlas_consistency_q01__a11550745f47", "report_id": "derived_multi-origin-c3s-atlas_consistency_q01", "dataset_id": "multi-origin-c3s-atlas", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q01", "aspect_base": "consistency", "category": "Derived_Datasets", "match_confidence": "exact", "section": "Consistency between the C3S Atlas dataset and its origins: Case study > Analysis and results > 4. Results > Reproducibility: Comparison on C3S Atlas grid", "title": "Consistency between the C3S Atlas dataset and its origins: Case study", "chunk_index": 21, "token_count": 727, "text_raw": "After regridding to the C3S Atlas grid,\nthe indicator values reproduced from the origin dataset\ncan be compared point-by-point to the values retrieved from the C3S Atlas dataset.\nWe first examine some metrics that describe the difference Δ between corresponding pixels:\n\nIt is clear that the two datasets are extremely close:\nthe median difference and median absolute difference are both 0\nand\nfewer than 0.1% of pixel pairs show a non-zero difference\n(defined here as |Δ| ≥ ε with ε = 10{sup}`–5` to avoid floating-point errors).\nWe can immediately conclude that the C3S Atlas dataset is highly reproducible\nfrom its origins\nusing the provided workflow.\n\nThis conclusion is strengthened by examining the distributions of indicator values and their point-by-point differences\n(Figure {numref}`{number} `).\nThe vast majority of pixel pairs have a difference of 0,\nwith very few ±1 and even fewer ±2\n(note the logarithmic y-axis).\n\nHowever, this raises an obvious question:\nwhy are there any differences at all?\nThe geospatial distribution of differences\n(Figure {numref}`{number} `)\ndoes not appear to correlate with any patterns in the indicator values themselves nor in geography.\nSince the differences are very small and seemingly random,\na potential explanation is that they simply result from small changes in\nthe origin dataset\nand/or\nthe C3S Atlas workflow software\nbetween the production of the C3S Atlas dataset and the present day.\nTesting this hypothesis would require an extremely detailed analysis of all steps in the workflow,\nwhich is outside the scope of this assessment.\n\nSetting this question affecting <0.1% of pixels aside,\nwe can confidently conclude\nfrom the metrics and figures\nthat the C3S Atlas dataset is highly reproducible from its origins using the provided workflow,\nat least for the `tx35` indicator and CMIP6 dataset.\n\nderived_multi-origin-c3s-atlas_consistency_q01_fig-hist-c3s-atlas\n:figwidth: 700px\n:name: \"derived_multi-origin-c3s-atlas_consistency_q01_fig-hist-c3s-atlas\"\n\nComparison between overall distributions of indicator (`tx35`) values in the C3S Atlas dataset and its reproduction\non the C3S Atlas grid,\nacross all spatial and temporal dimensions,\nincluding the per-pixel difference.\n```\n\nCreate plot with function defined at the start\n\nderived_multi-origin-c3s-atlas_consistency_q01_fig-geocomp-regional\n:figwidth: 700px\n:name: \"derived_multi-origin-c3s-atlas_consistency_q01_fig-geocomp-regional\"\n\nComparison between C3S Atlas dataset and reproduction for one indicator (`tx35`) in one month,\nacross Europe,\non the C3S Atlas dataset grid,\nincluding the per-pixel difference.\n```", "text_with_prefix": "EQC Quality Assessment: \"Consistency between the C3S Atlas dataset and its origins: Case study\"\nDataset: multi-origin-c3s-atlas [CDS]\nAspect: consistency_q01 | Category: Derived_Datasets\nSection: Consistency between the C3S Atlas dataset and its origins: Case study > Analysis and results > 4. Results > Reproducibility: Comparison on C3S Atlas grid\n---\nAfter regridding to the C3S Atlas grid,\nthe indicator values reproduced from the origin dataset\ncan be compared point-by-point to the values retrieved from the C3S Atlas dataset.\nWe first examine some metrics that describe the difference Δ between corresponding pixels:\n\nIt is clear that the two datasets are extremely close:\nthe median difference and median absolute difference are both 0\nand\nfewer than 0.1% of pixel pairs show a non-zero difference\n(defined here as |Δ| ≥ ε with ε = 10{sup}`–5` to avoid floating-point errors).\nWe can immediately conclude that the C3S Atlas dataset is highly reproducible\nfrom its origins\nusing the provided workflow.\n\nThis conclusion is strengthened by examining the distributions of indicator values and their point-by-point differences\n(Figure {numref}`{number} `).\nThe vast majority of pixel pairs have a difference of 0,\nwith very few ±1 and even fewer ±2\n(note the logarithmic y-axis).\n\nHowever, this raises an obvious question:\nwhy are there any differences at all?\nThe geospatial distribution of differences\n(Figure {numref}`{number} `)\ndoes not appear to correlate with any patterns in the indicator values themselves nor in geography.\nSince the differences are very small and seemingly random,\na potential explanation is that they simply result from small changes in\nthe origin dataset\nand/or\nthe C3S Atlas workflow software\nbetween the production of the C3S Atlas dataset and the present day.\nTesting this hypothesis would require an extremely detailed analysis of all steps in the workflow,\nwhich is outside the scope of this assessment.\n\nSetting this question affecting <0.1% of pixels aside,\nwe can confidently conclude\nfrom the metrics and figures\nthat the C3S Atlas dataset is highly reproducible from its origins using the provided workflow,\nat least for the `tx35` indicator and CMIP6 dataset.\n\nderived_multi-origin-c3s-atlas_consistency_q01_fig-hist-c3s-atlas\n:figwidth: 700px\n:name: \"derived_multi-origin-c3s-atlas_consistency_q01_fig-hist-c3s-atlas\"\n\nComparison between overall distributions of indicator (`tx35`) values in the C3S Atlas dataset and its reproduction\non the C3S Atlas grid,\nacross all spatial and temporal dimensions,\nincluding the per-pixel difference.\n```\n\nCreate plot with function defined at the start\n\nderived_multi-origin-c3s-atlas_consistency_q01_fig-geocomp-regional\n:figwidth: 700px\n:name: \"derived_multi-origin-c3s-atlas_consistency_q01_fig-geocomp-regional\"\n\nComparison between C3S Atlas dataset and reproduction for one indicator (`tx35`) in one month,\nacross Europe,\non the C3S Atlas dataset grid,\nincluding the per-pixel difference.\n```"} {"chunk_id": "derived_multi-origin-c3s-atlas_consistency_q01__92d17688398c", "report_id": "derived_multi-origin-c3s-atlas_consistency_q01", "dataset_id": "multi-origin-c3s-atlas", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q01", "aspect_base": "consistency", "category": "Derived_Datasets", "match_confidence": "exact", "section": "Consistency between the C3S Atlas dataset and its origins: Case study > ℹ️ If you want to know more > Key resources", "title": "Consistency between the C3S Atlas dataset and its origins: Case study", "chunk_index": 22, "token_count": 618, "text_raw": "The CDS catalogue entries for the data used were:\n* Gridded dataset underpinning the Copernicus Interactive Climate Atlas: [multi-origin-c3s-atlas](https://doi.org/10.24381/cds.h35hb680)\n * **Consistency between the C3S Atlas dataset and its origins: Case study**\n * [](./derived_multi-origin-c3s-atlas_consistency_q02)\n * [](./derived_multi-origin-c3s-atlas_consistency_q03)\n* CMIP6 climate projections: [projections-cmip6](https://doi.org/10.24381/cds.c866074c)\n * [Quality assessments for CMIP6](../Climate_Projections/CMIP6/CMIP6.md)\n\nCode libraries used:\n* [earthkit](https://github.com/ecmwf/earthkit)\n * [earthkit-data](https://github.com/ecmwf/earthkit-data)\n * [earthkit-plots](https://github.com/ecmwf/earthkit-plots)\n* [User-tools for the C3S Atlas](https://github.com/ecmwf-projects/c3s-atlas)\n* [xclim](https://xclim.readthedocs.io/en/stable/) climate indicator tools\n\nMore about the Copernicus Interactive Climate Atlas and its IPCC predecessor:\n* [Copernicus Interactive Climate Atlas application](https://atlas.climate.copernicus.eu/)\n* [Gridded data underpinning the Copernicus Interactive Climate Atlas: Description of the datasets and variables](https://confluence.ecmwf.int/display/CKB/Gridded+data+underpinning+the+Copernicus+Interactive+Climate+Atlas%3A+Description+of+the+datasets+and+variables)\n* [The Copernicus Interactive Climate Atlas: a tool to explore regional climate change](https://doi.org/10.21957/ah52ufc369)\n* [Copernicus Interactive Climate Atlas: a new tool to visualise climate variability and change](https://www.ecmwf.int/en/newsletter/179/news/copernicus-interactive-climate-atlas-new-tool-visualise-climate-variability)\n* [Implementation of FAIR principles in the IPCC: the WGI AR6 Atlas repository](https://doi.org/10.1038/s41597-022-01739-y)\n* [Climate Change 2021 – The Physical Science Basis: Atlas](https://doi.org/10.1017/9781009157896.021)", "text_with_prefix": "EQC Quality Assessment: \"Consistency between the C3S Atlas dataset and its origins: Case study\"\nDataset: multi-origin-c3s-atlas [CDS]\nAspect: consistency_q01 | Category: Derived_Datasets\nSection: Consistency between the C3S Atlas dataset and its origins: Case study > ℹ️ If you want to know more > Key resources\n---\nThe CDS catalogue entries for the data used were:\n* Gridded dataset underpinning the Copernicus Interactive Climate Atlas: [multi-origin-c3s-atlas](https://doi.org/10.24381/cds.h35hb680)\n * **Consistency between the C3S Atlas dataset and its origins: Case study**\n * [](./derived_multi-origin-c3s-atlas_consistency_q02)\n * [](./derived_multi-origin-c3s-atlas_consistency_q03)\n* CMIP6 climate projections: [projections-cmip6](https://doi.org/10.24381/cds.c866074c)\n * [Quality assessments for CMIP6](../Climate_Projections/CMIP6/CMIP6.md)\n\nCode libraries used:\n* [earthkit](https://github.com/ecmwf/earthkit)\n * [earthkit-data](https://github.com/ecmwf/earthkit-data)\n * [earthkit-plots](https://github.com/ecmwf/earthkit-plots)\n* [User-tools for the C3S Atlas](https://github.com/ecmwf-projects/c3s-atlas)\n* [xclim](https://xclim.readthedocs.io/en/stable/) climate indicator tools\n\nMore about the Copernicus Interactive Climate Atlas and its IPCC predecessor:\n* [Copernicus Interactive Climate Atlas application](https://atlas.climate.copernicus.eu/)\n* [Gridded data underpinning the Copernicus Interactive Climate Atlas: Description of the datasets and variables](https://confluence.ecmwf.int/display/CKB/Gridded+data+underpinning+the+Copernicus+Interactive+Climate+Atlas%3A+Description+of+the+datasets+and+variables)\n* [The Copernicus Interactive Climate Atlas: a tool to explore regional climate change](https://doi.org/10.21957/ah52ufc369)\n* [Copernicus Interactive Climate Atlas: a new tool to visualise climate variability and change](https://www.ecmwf.int/en/newsletter/179/news/copernicus-interactive-climate-atlas-new-tool-visualise-climate-variability)\n* [Implementation of FAIR principles in the IPCC: the WGI AR6 Atlas repository](https://doi.org/10.1038/s41597-022-01739-y)\n* [Climate Change 2021 – The Physical Science Basis: Atlas](https://doi.org/10.1017/9781009157896.021)"} {"chunk_id": "derived_multi-origin-c3s-atlas_consistency_q01__cc0c0f0a560b", "report_id": "derived_multi-origin-c3s-atlas_consistency_q01", "dataset_id": "multi-origin-c3s-atlas", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q01", "aspect_base": "consistency", "category": "Derived_Datasets", "match_confidence": "exact", "section": "Consistency between the C3S Atlas dataset and its origins: Case study > ℹ️ If you want to know more > References", "title": "Consistency between the C3S Atlas dataset and its origins: Case study", "chunk_index": 23, "token_count": 439, "text_raw": "[[Avila+15](https://doi.org/10.1016/j.wace.2015.06.003)] F. B. Avila et al., ‘Systematic investigation of gridding-related scaling effects on annual statistics of daily temperature and precipitation maxima: A case study for south-east Australia’, Weather and Climate Extremes, vol. 9, pp. 6–16, Aug. 2015, doi: 10.1016/j.wace.2015.06.003.\n\n[[C3S Atlas dataset](https://doi.org/10.24381/cds.h35hb680)] Copernicus Climate Change Service, ‘Gridded dataset underpinning the Copernicus Interactive Climate Atlas’. Copernicus Climate Change Service (C3S) Climate Data Store (CDS), Jun. 17, 2024. doi: 10.24381/cds.h35hb680.\n\n[[CMIP6 dataset](https://doi.org/10.24381/cds.c866074c)] Copernicus Climate Change Service, ‘CMIP6 climate projections’. Copernicus Climate Change Service (C3S) Climate Data Store (CDS), Mar. 23, 2021. doi: 10.24381/cds.c866074c.\n\n[[Gutiérrez+24](https://doi.org/10.21957/ah52ufc369)] J. M. Gutiérrez et al., ‘The Copernicus Interactive Climate Atlas: a tool to explore regional climate change’, ECMWF Newsletter, vol. 181, pp. 38–45, Oct. 2024, doi: 10.21957/ah52ufc369.", "text_with_prefix": "EQC Quality Assessment: \"Consistency between the C3S Atlas dataset and its origins: Case study\"\nDataset: multi-origin-c3s-atlas [CDS]\nAspect: consistency_q01 | Category: Derived_Datasets\nSection: Consistency between the C3S Atlas dataset and its origins: Case study > ℹ️ If you want to know more > References\n---\n[[Avila+15](https://doi.org/10.1016/j.wace.2015.06.003)] F. B. Avila et al., ‘Systematic investigation of gridding-related scaling effects on annual statistics of daily temperature and precipitation maxima: A case study for south-east Australia’, Weather and Climate Extremes, vol. 9, pp. 6–16, Aug. 2015, doi: 10.1016/j.wace.2015.06.003.\n\n[[C3S Atlas dataset](https://doi.org/10.24381/cds.h35hb680)] Copernicus Climate Change Service, ‘Gridded dataset underpinning the Copernicus Interactive Climate Atlas’. Copernicus Climate Change Service (C3S) Climate Data Store (CDS), Jun. 17, 2024. doi: 10.24381/cds.h35hb680.\n\n[[CMIP6 dataset](https://doi.org/10.24381/cds.c866074c)] Copernicus Climate Change Service, ‘CMIP6 climate projections’. Copernicus Climate Change Service (C3S) Climate Data Store (CDS), Mar. 23, 2021. doi: 10.24381/cds.c866074c.\n\n[[Gutiérrez+24](https://doi.org/10.21957/ah52ufc369)] J. M. Gutiérrez et al., ‘The Copernicus Interactive Climate Atlas: a tool to explore regional climate change’, ECMWF Newsletter, vol. 181, pp. 38–45, Oct. 2024, doi: 10.21957/ah52ufc369."} {"chunk_id": "derived_multi-origin-c3s-atlas_consistency_q02__c97c355add95", "report_id": "derived_multi-origin-c3s-atlas_consistency_q02", "dataset_id": "multi-origin-c3s-atlas", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q02", "aspect_base": "consistency", "category": "Derived_Datasets", "match_confidence": "exact", "section": "Consistency between the C3S Atlas dataset and its origins: Multiple indicators", "title": "Consistency between the C3S Atlas dataset and its origins: Multiple indicators", "chunk_index": 0, "token_count": 103, "text_raw": "Production date: 2025-10-01.\n\nDataset version: 2.0.\n\nProduced by: Olivier Burggraaff, Nicole Reynolds (National Physical Laboratory).", "text_with_prefix": "EQC Quality Assessment: \"Consistency between the C3S Atlas dataset and its origins: Multiple indicators\"\nDataset: multi-origin-c3s-atlas [CDS]\nAspect: consistency_q02 | Category: Derived_Datasets\nSection: Consistency between the C3S Atlas dataset and its origins: Multiple indicators\n---\nProduction date: 2025-10-01.\n\nDataset version: 2.0.\n\nProduced by: Olivier Burggraaff, Nicole Reynolds (National Physical Laboratory)."} {"chunk_id": "derived_multi-origin-c3s-atlas_consistency_q02__9ed08f15eb5c", "report_id": "derived_multi-origin-c3s-atlas_consistency_q02", "dataset_id": "multi-origin-c3s-atlas", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q02", "aspect_base": "consistency", "category": "Derived_Datasets", "match_confidence": "exact", "section": "Consistency between the C3S Atlas dataset and its origins: Multiple indicators > Quality assessment question", "title": "Consistency between the C3S Atlas dataset and its origins: Multiple indicators", "chunk_index": 1, "token_count": 881, "text_raw": "* **Are the climate indicators in the dataset underpinning the Copernicus Interactive Climate Atlas consistent with their origin datasets?**\n* **Can the dataset underpinning the Copernicus Interactive Climate Atlas be reproduced from its origin datasets?**\n\nThe [_Copernicus Interactive Climate Atlas_](https://atlas.climate.copernicus.eu/atlas), or _C3S Atlas_ for short, is a C3S web application providing an easy-to-access tool for exploring climate projections, reanalyses, and observational data [[Gutiérrez+24](https://doi.org/10.21957/ah52ufc369)].\nVersion 2.0 of the application allows the user to interact with 12 datasets:\n\n| Type | Dataset |\n|--------------------|---------------|\n| Climate Projection | CMIP6 |\n| Climate Projection | CMIP5 |\n| Climate Projection | CORDEX-CORE |\n| Climate Projection | CORDEX-EUR-11 |\n| Reanalysis | ERA5 |\n| Reanalysis | ERA5-Land |\n| Reanalysis | ORAS5 |\n| Reanalysis | CERRA |\n| Observations | E-OBS |\n| Observations | BERKEARTH |\n| Observations | CPC |\n| Observations | SST-CCI |\n\nThese datasets are provided through an intermediary dataset, the [_Gridded dataset underpinning the Copernicus Interactive Climate Atlas_](https://doi.org/10.24381/cds.h35hb680) or _C3S Atlas dataset_ for short [[C3S Atlas dataset](https://doi.org/10.24381/cds.h35hb680)].\nCompared to their origins, the versions of the climate datasets within the C3S Atlas dataset have been processed following the workflow in Figure {numref}`{number} `.\n\nattachment:c3s_atlas_dataset_workflow.png\n---\nheight: 360px\nname: multi-origin-c3s-atlas_consistency_q02_workflow-fig\n---\nSchematic representation of the workflow for the production of the C3S Atlas dataset from its origin datasets, from the [User-tools for the C3S Atlas](https://ecmwf-projects.github.io/c3s-atlas/chapter01.html).\n```\n\nBecause a wide range of users interact with climate data through the C3S Atlas application, it is crucial that the underpinning dataset represent its origins correctly.\nIn other words, the C3S Atlas dataset must be consistent with and reproducible from its origins.\nHere, we assess this consistency and reproducibility by comparing climate indicators retrieved from the C3S Atlas dataset with their equivalents calculated from the origin dataset, mirroring the workflow from Figure {numref}`{number} `.\nWhile a full analysis and reproduction of every record within the C3S Atlas dataset is outside the scope of quality assessment\n(and would require high-performance computing infrastructure),\na case study with a narrower scope probes these quality attributes of the dataset\nand can be a jumping-off point for further analysis by the reader.\n\nThis notebook is part of a series:\n| Notebook | Contents |\n|---|---|\n| [](./derived_multi-origin-c3s-atlas_consistency_q01) | Comparison between C3S Atlas dataset and one origin dataset (CMIP6) for one indicator (`tx35`), including detailed setup. |\n| **Consistency between the C3S Atlas dataset and its origins: Multiple indicators** | Comparison between C3S Atlas dataset and one origin dataset (CMIP6) for multiple indicators. |\n| [](./derived_multi-origin-c3s-atlas_consistency_q03) | Comparison between C3S Atlas dataset and multiple origin datasets for one indicator. |", "text_with_prefix": "EQC Quality Assessment: \"Consistency between the C3S Atlas dataset and its origins: Multiple indicators\"\nDataset: multi-origin-c3s-atlas [CDS]\nAspect: consistency_q02 | Category: Derived_Datasets\nSection: Consistency between the C3S Atlas dataset and its origins: Multiple indicators > Quality assessment question\n---\n* **Are the climate indicators in the dataset underpinning the Copernicus Interactive Climate Atlas consistent with their origin datasets?**\n* **Can the dataset underpinning the Copernicus Interactive Climate Atlas be reproduced from its origin datasets?**\n\nThe [_Copernicus Interactive Climate Atlas_](https://atlas.climate.copernicus.eu/atlas), or _C3S Atlas_ for short, is a C3S web application providing an easy-to-access tool for exploring climate projections, reanalyses, and observational data [[Gutiérrez+24](https://doi.org/10.21957/ah52ufc369)].\nVersion 2.0 of the application allows the user to interact with 12 datasets:\n\n| Type | Dataset |\n|--------------------|---------------|\n| Climate Projection | CMIP6 |\n| Climate Projection | CMIP5 |\n| Climate Projection | CORDEX-CORE |\n| Climate Projection | CORDEX-EUR-11 |\n| Reanalysis | ERA5 |\n| Reanalysis | ERA5-Land |\n| Reanalysis | ORAS5 |\n| Reanalysis | CERRA |\n| Observations | E-OBS |\n| Observations | BERKEARTH |\n| Observations | CPC |\n| Observations | SST-CCI |\n\nThese datasets are provided through an intermediary dataset, the [_Gridded dataset underpinning the Copernicus Interactive Climate Atlas_](https://doi.org/10.24381/cds.h35hb680) or _C3S Atlas dataset_ for short [[C3S Atlas dataset](https://doi.org/10.24381/cds.h35hb680)].\nCompared to their origins, the versions of the climate datasets within the C3S Atlas dataset have been processed following the workflow in Figure {numref}`{number} `.\n\nattachment:c3s_atlas_dataset_workflow.png\n---\nheight: 360px\nname: multi-origin-c3s-atlas_consistency_q02_workflow-fig\n---\nSchematic representation of the workflow for the production of the C3S Atlas dataset from its origin datasets, from the [User-tools for the C3S Atlas](https://ecmwf-projects.github.io/c3s-atlas/chapter01.html).\n```\n\nBecause a wide range of users interact with climate data through the C3S Atlas application, it is crucial that the underpinning dataset represent its origins correctly.\nIn other words, the C3S Atlas dataset must be consistent with and reproducible from its origins.\nHere, we assess this consistency and reproducibility by comparing climate indicators retrieved from the C3S Atlas dataset with their equivalents calculated from the origin dataset, mirroring the workflow from Figure {numref}`{number} `.\nWhile a full analysis and reproduction of every record within the C3S Atlas dataset is outside the scope of quality assessment\n(and would require high-performance computing infrastructure),\na case study with a narrower scope probes these quality attributes of the dataset\nand can be a jumping-off point for further analysis by the reader.\n\nThis notebook is part of a series:\n| Notebook | Contents |\n|---|---|\n| [](./derived_multi-origin-c3s-atlas_consistency_q01) | Comparison between C3S Atlas dataset and one origin dataset (CMIP6) for one indicator (`tx35`), including detailed setup. |\n| **Consistency between the C3S Atlas dataset and its origins: Multiple indicators** | Comparison between C3S Atlas dataset and one origin dataset (CMIP6) for multiple indicators. |\n| [](./derived_multi-origin-c3s-atlas_consistency_q03) | Comparison between C3S Atlas dataset and multiple origin datasets for one indicator. |"} {"chunk_id": "derived_multi-origin-c3s-atlas_consistency_q02__ba02302be65a", "report_id": "derived_multi-origin-c3s-atlas_consistency_q02", "dataset_id": "multi-origin-c3s-atlas", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q02", "aspect_base": "consistency", "category": "Derived_Datasets", "match_confidence": "exact", "section": "Consistency between the C3S Atlas dataset and its origins: Multiple indicators > Quality assessment statement", "title": "Consistency between the C3S Atlas dataset and its origins: Multiple indicators", "chunk_index": 2, "token_count": 310, "text_raw": "These are the key outcomes of this assessment\n\n* Climate indicators (here 25 monthly indicators) provided by the C3S Atlas dataset are highly consistent with values calculated from its origin datasets (here the CMIP6 multi-model ensemble).\n\n* There are some differences in coverage between the C3S Atlas dataset and its origins due to deliberate masking, e.g. agricultural indicators in the Arctic circle, but these do not affect the vast majority of use cases for the C3S Atlas.\n\n* Differences between the C3S Atlas dataset and a manual reproduction are rare and generally negligible (median absolute difference ≤0.02% for all but 2 indicators). Where differences occur, they can be explained by details of the workflow implementation or deliberate choices.\n\n* The C3S Atlas is traceable and reproducible, and can be confidently used to view, analyse, and download climate data.\n\n* For specific use cases, like scientific papers, it is recommended to manually process the origin dataset instead of using the C3S Atlas. This is not necessary for other use cases, such as climate risk assessments and climate reports, in which case the C3S Atlas can be used as is.\n```", "text_with_prefix": "EQC Quality Assessment: \"Consistency between the C3S Atlas dataset and its origins: Multiple indicators\"\nDataset: multi-origin-c3s-atlas [CDS]\nAspect: consistency_q02 | Category: Derived_Datasets\nSection: Consistency between the C3S Atlas dataset and its origins: Multiple indicators > Quality assessment statement\n---\nThese are the key outcomes of this assessment\n\n* Climate indicators (here 25 monthly indicators) provided by the C3S Atlas dataset are highly consistent with values calculated from its origin datasets (here the CMIP6 multi-model ensemble).\n\n* There are some differences in coverage between the C3S Atlas dataset and its origins due to deliberate masking, e.g. agricultural indicators in the Arctic circle, but these do not affect the vast majority of use cases for the C3S Atlas.\n\n* Differences between the C3S Atlas dataset and a manual reproduction are rare and generally negligible (median absolute difference ≤0.02% for all but 2 indicators). Where differences occur, they can be explained by details of the workflow implementation or deliberate choices.\n\n* The C3S Atlas is traceable and reproducible, and can be confidently used to view, analyse, and download climate data.\n\n* For specific use cases, like scientific papers, it is recommended to manually process the origin dataset instead of using the C3S Atlas. This is not necessary for other use cases, such as climate risk assessments and climate reports, in which case the C3S Atlas can be used as is.\n```"} {"chunk_id": "derived_multi-origin-c3s-atlas_consistency_q02__6897cfea64ba", "report_id": "derived_multi-origin-c3s-atlas_consistency_q02", "dataset_id": "multi-origin-c3s-atlas", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q02", "aspect_base": "consistency", "category": "Derived_Datasets", "match_confidence": "exact", "section": "Consistency between the C3S Atlas dataset and its origins: Multiple indicators > Methodology", "title": "Consistency between the C3S Atlas dataset and its origins: Multiple indicators", "chunk_index": 3, "token_count": 259, "text_raw": "This quality assessment tests the consistency between climate indicators retrieved from the [_Gridded dataset underpinning the Copernicus Interactive Climate Atlas_](https://doi.org/10.24381/cds.h35hb680) [[C3S Atlas dataset](https://doi.org/10.24381/cds.h35hb680)] and their equivalents calculated from the origin datasets,\nas well as the reproducibility of said dataset.\n\nThis notebook expands the analysis set out in [the case study](./derived_multi-origin-c3s-atlas_consistency_q01)\nto investigate multiple indicators\n(listed below, descriptions from the [User-tools for the C3S Atlas](https://ecmwf-projects.github.io/c3s-atlas/chapter01.html))\nderived from the CMIP6 multi-model ensemble [[CMIP6 dataset](https://doi.org/10.24381/cds.c866074c)]:", "text_with_prefix": "EQC Quality Assessment: \"Consistency between the C3S Atlas dataset and its origins: Multiple indicators\"\nDataset: multi-origin-c3s-atlas [CDS]\nAspect: consistency_q02 | Category: Derived_Datasets\nSection: Consistency between the C3S Atlas dataset and its origins: Multiple indicators > Methodology\n---\nThis quality assessment tests the consistency between climate indicators retrieved from the [_Gridded dataset underpinning the Copernicus Interactive Climate Atlas_](https://doi.org/10.24381/cds.h35hb680) [[C3S Atlas dataset](https://doi.org/10.24381/cds.h35hb680)] and their equivalents calculated from the origin datasets,\nas well as the reproducibility of said dataset.\n\nThis notebook expands the analysis set out in [the case study](./derived_multi-origin-c3s-atlas_consistency_q01)\nto investigate multiple indicators\n(listed below, descriptions from the [User-tools for the C3S Atlas](https://ecmwf-projects.github.io/c3s-atlas/chapter01.html))\nderived from the CMIP6 multi-model ensemble [[CMIP6 dataset](https://doi.org/10.24381/cds.c866074c)]:"} {"chunk_id": "derived_multi-origin-c3s-atlas_consistency_q02__31f1bf47c070", "report_id": "derived_multi-origin-c3s-atlas_consistency_q02", "dataset_id": "multi-origin-c3s-atlas", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q02", "aspect_base": "consistency", "category": "Derived_Datasets", "match_confidence": "exact", "section": "Consistency between the C3S Atlas dataset and its origins: Multiple indicators > Methodology", "title": "Consistency between the C3S Atlas dataset and its origins: Multiple indicators", "chunk_index": 4, "token_count": 1080, "text_raw": "81/cds.c866074c)]:\n\n| Indicator | Unit | Full description |\n|-----------|------|------------------|\n| `t` | °C | Monthly mean of daily mean near-surface (2-metre) air temperature |\n| `tn` | °C | Monthly mean of daily minimum near-surface (2-metre) air temperature |\n| `tx` | °C | Monthly mean of daily maximum near-surface (2-metre) air temperature |\n| `dtr` | °C | Monthly mean of near-surface (2-metre) air temperature difference between the maximum and minimum daily temperature |\n| `tnn` | °C | Monthly minimum of daily minimum near-surface (2-metre) air temperature |\n| `txx` | °C | Monthly maximum of daily maximum near-surface (2-metre) air temperature |\n| `tx35` | d | Monthly count of days with maximum near-surface (2-metre) temperature above 35 °C |\n| `tx40` | d | Monthly count of days with maximum near-surface (2-metre) temperature above 40 °C |\n| `tr` | d | Monthly count of tropical nights (days with minimum temperature above 20 °C) |\n| `fd` | d | Monthly count of days with minimum near-surface (2-metre) temperature below 0 °C |\n| `r` | mm | Monthly mean of daily accumulated precipitation of liquid water equivalent from all phases |\n| `sdii` | mm | Monthly average of daily precipitation amount of liquid water equivalent from all phases on days with precipitation amount above or equal to 1 mm |\n| `prsn` | mm | Monthly mean of daily accumulated liquid water equivalent thickness snowfall |\n| `rx1day` | mm | Monthly maximum of 1-day accumulated precipitation of liquid water equivalent from all phases |\n| `r01` | d | Monthly count of days with daily accumulated precipitation of liquid water equivalent from all phases above 1 mm |\n| `r10` | d | Monthly count of days with daily accumulated precipitation of liquid water equivalent from all phases above 10 mm |\n| `r20` | d | Monthly count of days with daily accumulated precipitation of liquid water equivalent from all phases above 20 mm |\n| `pet` | mm | Monthly mean daily accumulated potential evapotranspiration (Hargreaves method, 1985), which is the rate at which evapotranspiration would occur under ambient conditions from a uniformly vegetated area when the water supply is not limiting |\n| `evspsbl` | mm | Monthly mean of daily amount of water in the atmosphere due to conversion of both liquid and solid phases to vapor (from underlying surface and vegetation) |\n| `huss` | g/kg | Monthly amount of moisture in the air near the surface divided by amount of air plus moisture at that location |\n| `psl` | hPa | Monthly average air pressure at mean sea level |\n| `sfcwind` | m/s | Monthly mean of daily mean near-surface (10-metre) wind speed |\n| `clt` | % | Monthly mean cloud cover area percentage |\n| `rsds` | W/m² | Monthly mean incident solar (shortwave) radiation that reaches a horizontal plane at the surface |\n| `rlds` | W/m² | Monthly mean incident thermal (longwave) radiation at the surface (during cloudless and overcast conditions) |\n\n\n\n\n", "text_with_prefix": "EQC Quality Assessment: \"Consistency between the C3S Atlas dataset and its origins: Multiple indicators\"\nDataset: multi-origin-c3s-atlas [CDS]\nAspect: consistency_q02 | Category: Derived_Datasets\nSection: Consistency between the C3S Atlas dataset and its origins: Multiple indicators > Methodology\n---\n| — | Monthly index that compares accumulated precipitation for 6 months with the long-term precipitation distribution for the same location and accumulation period, as the number of standard deviations from the mean. The reference period corresponds to 1971-2010\n\n. |\n| `spei6` | — | Monthly index that compares accumulated precipitation minus potential evapotranspiration (Hargreaves method, 1985) for 6 months with the long-term distribution for the same location and accumulation period, as the number of standard deviations from the mean. The reference period corresponds to 1971-2010.\n-->"} {"chunk_id": "derived_multi-origin-c3s-atlas_consistency_q02__5b4b14a3a95e", "report_id": "derived_multi-origin-c3s-atlas_consistency_q02", "dataset_id": "multi-origin-c3s-atlas", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q02", "aspect_base": "consistency", "category": "Derived_Datasets", "match_confidence": "exact", "section": "Consistency between the C3S Atlas dataset and its origins: Multiple indicators > Methodology", "title": "Consistency between the C3S Atlas dataset and its origins: Multiple indicators", "chunk_index": 6, "token_count": 235, "text_raw": "0.\n-->\n\nThe analysis and results are organised in the following steps, which are detailed in the sections below:\n\n**[](section-codesetup)**\n * Install User-tools for the C3S Atlas.\n * Import all required libraries.\n * Define indicators.\n * Define helper functions.\n\n**[](section-origin)**\n * Download data from the origin dataset(s).\n * Homogenise data.\n * Calculate indicator(s).\n * Regrid the origin data to the C3S Atlas grid.\n\n**[](section-c3s-atlas)**\n * Download data from the C3S Atlas dataset.\n\n**[](section-results)**\n * Consistency: Compare the C3S Atlas and reproduced datasets on native grids.\n * Reproducibility: Compare the C3S Atlas and reproduced datasets on the C3S Atlas grid.", "text_with_prefix": "EQC Quality Assessment: \"Consistency between the C3S Atlas dataset and its origins: Multiple indicators\"\nDataset: multi-origin-c3s-atlas [CDS]\nAspect: consistency_q02 | Category: Derived_Datasets\nSection: Consistency between the C3S Atlas dataset and its origins: Multiple indicators > Methodology\n---\n0.\n-->\n\nThe analysis and results are organised in the following steps, which are detailed in the sections below:\n\n**[](section-codesetup)**\n * Install User-tools for the C3S Atlas.\n * Import all required libraries.\n * Define indicators.\n * Define helper functions.\n\n**[](section-origin)**\n * Download data from the origin dataset(s).\n * Homogenise data.\n * Calculate indicator(s).\n * Regrid the origin data to the C3S Atlas grid.\n\n**[](section-c3s-atlas)**\n * Download data from the C3S Atlas dataset.\n\n**[](section-results)**\n * Consistency: Compare the C3S Atlas and reproduced datasets on native grids.\n * Reproducibility: Compare the C3S Atlas and reproduced datasets on the C3S Atlas grid."} {"chunk_id": "derived_multi-origin-c3s-atlas_consistency_q02__02a55325eb09", "report_id": "derived_multi-origin-c3s-atlas_consistency_q02", "dataset_id": "multi-origin-c3s-atlas", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q02", "aspect_base": "consistency", "category": "Derived_Datasets", "match_confidence": "exact", "section": "Consistency between the C3S Atlas dataset and its origins: Multiple indicators > Analysis and results > 1. Code setup", "title": "Consistency between the C3S Atlas dataset and its origins: Multiple indicators", "chunk_index": 7, "token_count": 184, "text_raw": "This notebook uses [earthkit](https://github.com/ecmwf/earthkit) for \ndownloading ([earthkit-data](https://github.com/ecmwf/earthkit-data)) \nand visualising ([earthkit-plots](https://github.com/ecmwf/earthkit-plots)) data.\nBecause earthkit is in active development, some functionality may change after this notebook is published.\nIf any part of the code stops functioning, please raise an issue on our GitHub repository so it can be fixed.\n```", "text_with_prefix": "EQC Quality Assessment: \"Consistency between the C3S Atlas dataset and its origins: Multiple indicators\"\nDataset: multi-origin-c3s-atlas [CDS]\nAspect: consistency_q02 | Category: Derived_Datasets\nSection: Consistency between the C3S Atlas dataset and its origins: Multiple indicators > Analysis and results > 1. Code setup\n---\nThis notebook uses [earthkit](https://github.com/ecmwf/earthkit) for \ndownloading ([earthkit-data](https://github.com/ecmwf/earthkit-data)) \nand visualising ([earthkit-plots](https://github.com/ecmwf/earthkit-plots)) data.\nBecause earthkit is in active development, some functionality may change after this notebook is published.\nIf any part of the code stops functioning, please raise an issue on our GitHub repository so it can be fixed.\n```"} {"chunk_id": "derived_multi-origin-c3s-atlas_consistency_q02__846b80c9a95d", "report_id": "derived_multi-origin-c3s-atlas_consistency_q02", "dataset_id": "multi-origin-c3s-atlas", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q02", "aspect_base": "consistency", "category": "Derived_Datasets", "match_confidence": "exact", "section": "Consistency between the C3S Atlas dataset and its origins: Multiple indicators > Analysis and results > 1. Code setup > Install the User-tools for the C3S Atlas", "title": "Consistency between the C3S Atlas dataset and its origins: Multiple indicators", "chunk_index": 8, "token_count": 182, "text_raw": "This notebook uses the [User-tools for the C3S Atlas](https://github.com/ecmwf-projects/c3s-atlas), which can be installed from GitHub using `pip`.\nFor convenience, the following cell can do this from within the notebook.\nFurther details and alternative options for installing this library are available in its [documentation](https://github.com/ecmwf-projects/c3s-atlas?tab=readme-ov-file#requirements).", "text_with_prefix": "EQC Quality Assessment: \"Consistency between the C3S Atlas dataset and its origins: Multiple indicators\"\nDataset: multi-origin-c3s-atlas [CDS]\nAspect: consistency_q02 | Category: Derived_Datasets\nSection: Consistency between the C3S Atlas dataset and its origins: Multiple indicators > Analysis and results > 1. Code setup > Install the User-tools for the C3S Atlas\n---\nThis notebook uses the [User-tools for the C3S Atlas](https://github.com/ecmwf-projects/c3s-atlas), which can be installed from GitHub using `pip`.\nFor convenience, the following cell can do this from within the notebook.\nFurther details and alternative options for installing this library are available in its [documentation](https://github.com/ecmwf-projects/c3s-atlas?tab=readme-ov-file#requirements)."} {"chunk_id": "derived_multi-origin-c3s-atlas_consistency_q02__62f78432e454", "report_id": "derived_multi-origin-c3s-atlas_consistency_q02", "dataset_id": "multi-origin-c3s-atlas", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q02", "aspect_base": "consistency", "category": "Derived_Datasets", "match_confidence": "exact", "section": "Consistency between the C3S Atlas dataset and its origins: Multiple indicators > Analysis and results > 1. Code setup > Import required libraries", "title": "Consistency between the C3S Atlas dataset and its origins: Multiple indicators", "chunk_index": 9, "token_count": 135, "text_raw": "In this section, we import all the relevant packages needed for running the notebook.\n\nInput / Output\nGeneral data handling\nData pre-processing\nClimate indicators\nFix for bug affecting sfcWind units\nVisualisation\nVisualisation in Jupyter book -- automatically ignored otherwise", "text_with_prefix": "EQC Quality Assessment: \"Consistency between the C3S Atlas dataset and its origins: Multiple indicators\"\nDataset: multi-origin-c3s-atlas [CDS]\nAspect: consistency_q02 | Category: Derived_Datasets\nSection: Consistency between the C3S Atlas dataset and its origins: Multiple indicators > Analysis and results > 1. Code setup > Import required libraries\n---\nIn this section, we import all the relevant packages needed for running the notebook.\n\nInput / Output\nGeneral data handling\nData pre-processing\nClimate indicators\nFix for bug affecting sfcWind units\nVisualisation\nVisualisation in Jupyter book -- automatically ignored otherwise"} {"chunk_id": "derived_multi-origin-c3s-atlas_consistency_q02__f8d481098a7c", "report_id": "derived_multi-origin-c3s-atlas_consistency_q02", "dataset_id": "multi-origin-c3s-atlas", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q02", "aspect_base": "consistency", "category": "Derived_Datasets", "match_confidence": "exact", "section": "Consistency between the C3S Atlas dataset and its origins: Multiple indicators > Analysis and results > 1. Code setup > Define indicators", "title": "Consistency between the C3S Atlas dataset and its origins: Multiple indicators", "chunk_index": 10, "token_count": 912, "text_raw": "This section defines functions and variables for calculating and using the climate indicators.\nThese are split into three cells, corresponding to their input and output data types.\n\nDaily data → Monthly indicators:\n\nDaily data -> Monthly indicators\nInput: tasmin\nInput: tasmax\nInput: tasmin + tasmax\nConvert to mm (assuming kg water /m² = mm)\nConvert to g/kg\nDefunct -- problems with units\ndef cal_rx5day(ds: xr.Dataset) -> xr.Dataset:\n\"\"\" Monthly maximum of 5-day accumulated precipitation of liquid water equivalent from all phases \"\"\"\npr_flux = ds['pr'].copy().assign_attrs(units = 'mm/day')\nds_rx5day = xclim.indicators.atmos.max_n_day_precipitation_amount(pr_flux, window=5, freq='MS').to_dataset(name='rx5day')\nreturn ds_rx5day\ndef cal_spi6(ds: xr.Dataset) -> xr.Dataset:\n\"\"\" Monthly index that compares accumulated precipitation for 6 months with the long-term precipitation distribution for the same location and accumulation period, as the number of standard deviations from the mean. The reference period corresponds to 1971-2010 \"\"\"\npr_flux = ds['pr'].copy().assign_attrs(units = 'mm/day')\nds_spi6 = xclim.indices.standardized_precipitation_index(pr_flux, freq='MS', window=6, dist='gamma', method='ML').to_dataset(name='spi6')\nreturn ds_spi6\ndef cal_spei6(ds: xr.Dataset) -> xr.Dataset:\n\"\"\" Monthly index that compares accumulated precipitation minus potential evapotranspiration (Hargreaves method, 1985) for 6 months with the long-term distribution for the same location and accumulation period, as the number of standard deviations from the mean. The reference period corresponds to 1971-2010 \"\"\"\npr_flux = ds['pr'].copy().assign_attrs(units = 'mm/day')\nds_spei6 = xclim.indices.standardized_precipitation_index(pr_flux, freq='MS', window=6, dist='gamma', method='ML').to_dataset(name='spei6')\nreturn ds_spei6\n\nMonthly data → Monthly indicators:\n\nMonthly data -> Monthly indicators\nInput: evpsbl\nInput: mrsos\nInput: sfcWind\nConvert to hPa\n\nMonthly data → Yearly indicators.\nThese yearly indicators are not used in the rest of the analysis, but are included as a jumping-off point for the reader's own analysis.\n\nDaily data -> Yearly indicators\nInput: tas + tasmin + tasmax\n\nThe following cell defines some convenience functions for calculating multiple indicators in a loop, grouped according to their input and output data types:\n\nCombine functions into dictionaries for easy access by name\nNote: Skipping rx5day due to window length error\n\"rx5day\": cal_rx5day,\nNote: Skipping spi6, spei6 due to unit issues\n\"spi6\": cal_spi6,\n\"spei6\": cal_spei6,\nNote: Skipping mrsos and mrro due to CMIP6 grid issues\n\"mrsos\": cal_mrsos,\n\"mrro\": cal_mrro,\nEasier access in loops\nHard-code monthly indicator names to match table in intro\n\"mrsos\", \"mrro\",\nIndicators for which a relative difference can be calculated (e.g. not temperature in °C)\nindicators_monthly_names = list(indicators_monthly.keys())\nCalculation of multiple indicators\n\nThe following cell defines [earthkit-plots styles](https://earthkit-plots.readthedocs.io/en/latest/examples/examples/introduction/05-styles.html) for the indicators.\nThese styles define the colour maps and colour bar ranges for each quantity.\n\nStyles for indicators\nTemperature day indices\nPrecipitation\nEvapotranspiration\nSoil moisture\nGeneral meteorology\nIndividual styles\nSet up like this so they can still be edited individually\n\"rx5day\": Style(),\n\"spi6\": Style(),\n\"spei6\": Style(),\nApply general settings", "text_with_prefix": "EQC Quality Assessment: \"Consistency between the C3S Atlas dataset and its origins: Multiple indicators\"\nDataset: multi-origin-c3s-atlas [CDS]\nAspect: consistency_q02 | Category: Derived_Datasets\nSection: Consistency between the C3S Atlas dataset and its origins: Multiple indicators > Analysis and results > 1. Code setup > Define indicators\n---\nThis section defines functions and variables for calculating and using the climate indicators.\nThese are split into three cells, corresponding to their input and output data types.\n\nDaily data → Monthly indicators:\n\nDaily data -> Monthly indicators\nInput: tasmin\nInput: tasmax\nInput: tasmin + tasmax\nConvert to mm (assuming kg water /m² = mm)\nConvert to g/kg\nDefunct -- problems with units\ndef cal_rx5day(ds: xr.Dataset) -> xr.Dataset:\n\"\"\" Monthly maximum of 5-day accumulated precipitation of liquid water equivalent from all phases \"\"\"\npr_flux = ds['pr'].copy().assign_attrs(units = 'mm/day')\nds_rx5day = xclim.indicators.atmos.max_n_day_precipitation_amount(pr_flux, window=5, freq='MS').to_dataset(name='rx5day')\nreturn ds_rx5day\ndef cal_spi6(ds: xr.Dataset) -> xr.Dataset:\n\"\"\" Monthly index that compares accumulated precipitation for 6 months with the long-term precipitation distribution for the same location and accumulation period, as the number of standard deviations from the mean. The reference period corresponds to 1971-2010 \"\"\"\npr_flux = ds['pr'].copy().assign_attrs(units = 'mm/day')\nds_spi6 = xclim.indices.standardized_precipitation_index(pr_flux, freq='MS', window=6, dist='gamma', method='ML').to_dataset(name='spi6')\nreturn ds_spi6\ndef cal_spei6(ds: xr.Dataset) -> xr.Dataset:\n\"\"\" Monthly index that compares accumulated precipitation minus potential evapotranspiration (Hargreaves method, 1985) for 6 months with the long-term distribution for the same location and accumulation period, as the number of standard deviations from the mean. The reference period corresponds to 1971-2010 \"\"\"\npr_flux = ds['pr'].copy().assign_attrs(units = 'mm/day')\nds_spei6 = xclim.indices.standardized_precipitation_index(pr_flux, freq='MS', window=6, dist='gamma', method='ML').to_dataset(name='spei6')\nreturn ds_spei6\n\nMonthly data → Monthly indicators:\n\nMonthly data -> Monthly indicators\nInput: evpsbl\nInput: mrsos\nInput: sfcWind\nConvert to hPa\n\nMonthly data → Yearly indicators.\nThese yearly indicators are not used in the rest of the analysis, but are included as a jumping-off point for the reader's own analysis.\n\nDaily data -> Yearly indicators\nInput: tas + tasmin + tasmax\n\nThe following cell defines some convenience functions for calculating multiple indicators in a loop, grouped according to their input and output data types:\n\nCombine functions into dictionaries for easy access by name\nNote: Skipping rx5day due to window length error\n\"rx5day\": cal_rx5day,\nNote: Skipping spi6, spei6 due to unit issues\n\"spi6\": cal_spi6,\n\"spei6\": cal_spei6,\nNote: Skipping mrsos and mrro due to CMIP6 grid issues\n\"mrsos\": cal_mrsos,\n\"mrro\": cal_mrro,\nEasier access in loops\nHard-code monthly indicator names to match table in intro\n\"mrsos\", \"mrro\",\nIndicators for which a relative difference can be calculated (e.g. not temperature in °C)\nindicators_monthly_names = list(indicators_monthly.keys())\nCalculation of multiple indicators\n\nThe following cell defines [earthkit-plots styles](https://earthkit-plots.readthedocs.io/en/latest/examples/examples/introduction/05-styles.html) for the indicators.\nThese styles define the colour maps and colour bar ranges for each quantity.\n\nStyles for indicators\nTemperature day indices\nPrecipitation\nEvapotranspiration\nSoil moisture\nGeneral meteorology\nIndividual styles\nSet up like this so they can still be edited individually\n\"rx5day\": Style(),\n\"spi6\": Style(),\n\"spei6\": Style(),\nApply general settings"} {"chunk_id": "derived_multi-origin-c3s-atlas_consistency_q02__ae0219a4160c", "report_id": "derived_multi-origin-c3s-atlas_consistency_q02", "dataset_id": "multi-origin-c3s-atlas", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q02", "aspect_base": "consistency", "category": "Derived_Datasets", "match_confidence": "exact", "section": "Consistency between the C3S Atlas dataset and its origins: Multiple indicators > Analysis and results > 1. Code setup > Helper functions", "title": "Consistency between the C3S Atlas dataset and its origins: Multiple indicators", "chunk_index": 11, "token_count": 1013, "text_raw": "This section defines some functions and variables used in the following analysis, allowing code cells in later sections to be shorter and ensuring consistency.\n\n##### Data downloading & (pre-)processing\n\nThe following functions help with downloading data from datasets with limits, by generating multiple CDS requests with similar parameters:\n\nCreate requests for multiple daily / monthly variables\nDrop \"bnds\" variables which can mess with merge later on\n\nThe following functions aid in sub-selecting data, e.g. selecting one model from the ensemble included in the C3S Atlas dataset or selecting data for a specific time frame:\n\nC3S Atlas dataset: Individual model selection\nEnsure the model ID is provided in the right format\nFind the corresponding model ID in the list of models\nThis cannot use .sel because the coordinate is not indexed\nFind the corresponding data and return those\nSelect (multiple) years in a dataset\nSelect one month in multiple datasets\n\nThe following functions help with downloading C3S Atlas data for one specific ensemble member and a subset of the temporal range (e.g. one year):\n\nC3S Atlas dataset: Download all data, pulling out years and model members of interest\nDownload data\nDrop \"bnds\" variables which can mess with merge later on\nPick out member\nPick out years\n\nThe following functions handle the homogenisation of origin data to a consistent format using the [User-tools for the C3S Atlas](https://ecmwf-projects.github.io/c3s-atlas/intro.html):\n\nHomogenisation of origin dataset\n\nThe following functions handle regridding data based on ESMF as implemented in the [User-tools for the C3S Atlas](https://ecmwf-projects.github.io/c3s-atlas/intro.html).\nThis step is explained in more detail in the relevant section below.\n\nRegridding from native grid to C3S Atlas grid\nNote: this only works for monthly, not annual indicators\nIgnore repeated warnings on [-90, 90] bounds\n\n##### Statistics\nThe following functions calculate the difference (absolute / relative) between datasets, handling metadata etc.:\n\nDifference between datasets\nSelect and calculate\nReplace 0/0 with 0\n\nThe following functions calculate and display metrics for the difference between two datasets, e.g. mean and median deviation:\n\nIf indicators_relative is not specified explicitly, assume it is the same as indicators\nCalculate differences\nConvert to pandas\nCalculate aggregate statistics\nCalculate correlation coefficients\nCombine statistics into one dataframe\n\n##### Visualisation\nThe following cells contain functions for plotting results, starting with some base helper functions (e.g. displaying in Jupyter Notebook or Jupyter Book style, adding textboxes with consistent formatting, etc.):\n\nVisualisation: Helper functions, general\nGet the plt.Axes for each ekp.Subplot\nSet up location\nAdd the text\n\nThe following functions are also base helper functions, but specific to geospatial plots:\n\nVisualisation: Helper functions for geospatial plots\nCreate subplots\n\nThe following cell contains functions for geospatial comparisons between datasets on their native grids or on a common grid (the latter also showing the per-pixel difference):\n\nVisualisation: Plot indicators geospatially\nPre-process: Select data in one month\nSetup indicators\nCreate figure\nPlot indicators\nPlot individual datasets\nDecorate: Text + Colour bar\n_add_textbox_to_subplots(year, *subplots_data)\nTitles on top\nDecorate figure\nVisualisation: Plot indicators geospatially\nPre-process: Select data in one month, calculate difference\nSetup indicators\nCreate figure\nPlot indicators\nPlot individual datasets\nPlot difference\nDecorate: Text + Colour bar\nTitles on top\nDecorate figure\n\nThe following cell contains functions for histogram comparisons between datasets on their native grids or on a common grid (the latter also showing the per-pixel difference):\n\nVisualisation: Plot data in histograms\nSetup indicators\nCreate figure\nPlot histograms of data\nLoop over rows / indicators\nLoop over columns / data\nFlatten data\nCreate histogram\nIdentify panel\nTitles on top\nVisualisation: Plot data + difference in histograms\nIf indicators_relative is not specified explicitly, assume it is the same as indicators\nCalculate difference\nSetup indicators\nCreate figure\nSetup x/y share -- cannot be done in plt.subplots because of difference panel not sharing these\nPlot histograms of data\nLoop over rows / indicators\nLoop over columns / data\nFlatten data\nCreate histogram\nPlot difference\nPlot relative difference\nIdentify panel\nAdjust ylim if too close\nTitles on top\n\n(section-origin)=", "text_with_prefix": "EQC Quality Assessment: \"Consistency between the C3S Atlas dataset and its origins: Multiple indicators\"\nDataset: multi-origin-c3s-atlas [CDS]\nAspect: consistency_q02 | Category: Derived_Datasets\nSection: Consistency between the C3S Atlas dataset and its origins: Multiple indicators > Analysis and results > 1. Code setup > Helper functions\n---\nThis section defines some functions and variables used in the following analysis, allowing code cells in later sections to be shorter and ensuring consistency.\n\n##### Data downloading & (pre-)processing\n\nThe following functions help with downloading data from datasets with limits, by generating multiple CDS requests with similar parameters:\n\nCreate requests for multiple daily / monthly variables\nDrop \"bnds\" variables which can mess with merge later on\n\nThe following functions aid in sub-selecting data, e.g. selecting one model from the ensemble included in the C3S Atlas dataset or selecting data for a specific time frame:\n\nC3S Atlas dataset: Individual model selection\nEnsure the model ID is provided in the right format\nFind the corresponding model ID in the list of models\nThis cannot use .sel because the coordinate is not indexed\nFind the corresponding data and return those\nSelect (multiple) years in a dataset\nSelect one month in multiple datasets\n\nThe following functions help with downloading C3S Atlas data for one specific ensemble member and a subset of the temporal range (e.g. one year):\n\nC3S Atlas dataset: Download all data, pulling out years and model members of interest\nDownload data\nDrop \"bnds\" variables which can mess with merge later on\nPick out member\nPick out years\n\nThe following functions handle the homogenisation of origin data to a consistent format using the [User-tools for the C3S Atlas](https://ecmwf-projects.github.io/c3s-atlas/intro.html):\n\nHomogenisation of origin dataset\n\nThe following functions handle regridding data based on ESMF as implemented in the [User-tools for the C3S Atlas](https://ecmwf-projects.github.io/c3s-atlas/intro.html).\nThis step is explained in more detail in the relevant section below.\n\nRegridding from native grid to C3S Atlas grid\nNote: this only works for monthly, not annual indicators\nIgnore repeated warnings on [-90, 90] bounds\n\n##### Statistics\nThe following functions calculate the difference (absolute / relative) between datasets, handling metadata etc.:\n\nDifference between datasets\nSelect and calculate\nReplace 0/0 with 0\n\nThe following functions calculate and display metrics for the difference between two datasets, e.g. mean and median deviation:\n\nIf indicators_relative is not specified explicitly, assume it is the same as indicators\nCalculate differences\nConvert to pandas\nCalculate aggregate statistics\nCalculate correlation coefficients\nCombine statistics into one dataframe\n\n##### Visualisation\nThe following cells contain functions for plotting results, starting with some base helper functions (e.g. displaying in Jupyter Notebook or Jupyter Book style, adding textboxes with consistent formatting, etc.):\n\nVisualisation: Helper functions, general\nGet the plt.Axes for each ekp.Subplot\nSet up location\nAdd the text\n\nThe following functions are also base helper functions, but specific to geospatial plots:\n\nVisualisation: Helper functions for geospatial plots\nCreate subplots\n\nThe following cell contains functions for geospatial comparisons between datasets on their native grids or on a common grid (the latter also showing the per-pixel difference):\n\nVisualisation: Plot indicators geospatially\nPre-process: Select data in one month\nSetup indicators\nCreate figure\nPlot indicators\nPlot individual datasets\nDecorate: Text + Colour bar\n_add_textbox_to_subplots(year, *subplots_data)\nTitles on top\nDecorate figure\nVisualisation: Plot indicators geospatially\nPre-process: Select data in one month, calculate difference\nSetup indicators\nCreate figure\nPlot indicators\nPlot individual datasets\nPlot difference\nDecorate: Text + Colour bar\nTitles on top\nDecorate figure\n\nThe following cell contains functions for histogram comparisons between datasets on their native grids or on a common grid (the latter also showing the per-pixel difference):\n\nVisualisation: Plot data in histograms\nSetup indicators\nCreate figure\nPlot histograms of data\nLoop over rows / indicators\nLoop over columns / data\nFlatten data\nCreate histogram\nIdentify panel\nTitles on top\nVisualisation: Plot data + difference in histograms\nIf indicators_relative is not specified explicitly, assume it is the same as indicators\nCalculate difference\nSetup indicators\nCreate figure\nSetup x/y share -- cannot be done in plt.subplots because of difference panel not sharing these\nPlot histograms of data\nLoop over rows / indicators\nLoop over columns / data\nFlatten data\nCreate histogram\nPlot difference\nPlot relative difference\nIdentify panel\nAdjust ylim if too close\nTitles on top\n\n(section-origin)="} {"chunk_id": "derived_multi-origin-c3s-atlas_consistency_q02__6a5068b77961", "report_id": "derived_multi-origin-c3s-atlas_consistency_q02", "dataset_id": "multi-origin-c3s-atlas", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q02", "aspect_base": "consistency", "category": "Derived_Datasets", "match_confidence": "exact", "section": "Consistency between the C3S Atlas dataset and its origins: Multiple indicators > Analysis and results > 2. Calculate indicators from the origin dataset > Download data", "title": "Consistency between the C3S Atlas dataset and its origins: Multiple indicators", "chunk_index": 12, "token_count": 1083, "text_raw": "This assessment examines the dataset in one year (2080),\nfor one ensemble member\nand\nfor one climate scenario\n– which is not how climate projection data are normally used.\nGood practice in climate science is to look at multi-year statistics and trends,\nacross multiple ensemble members.\nHowever, since the purpose of this assessment is to assess the consistency and reproducibility of the post-processing performed to produce the C3S Atlas dataset,\nit is valid to use a subset of the data here.\nThe specific subset used can be easily tweaked by changing the `EXPERIMENT`, `MODEL`, and `YEARS` variables in this section.\n\nThis notebook uses [earthkit-data](https://github.com/ecmwf/earthkit-data) to download files from the CDS.\nIf you intend to run this notebook multiple times, it is highly recommended that you [enable caching](https://earthkit-data.readthedocs.io/en/latest/concepts/caching.html) to prevent having to download the same files multiple times.\nIf you prefer not to use earthkit, the following requests can also be used with the [cdsapi module](https://cds.climate.copernicus.eu/how-to-api#linux-use-client-step).\nIn either case (earthkit-data or cdsapi), it is required to set up a CDS account and API key as explained [on the CDS website](https://cds.climate.copernicus.eu/how-to-api).\n\nThe first step is to define the parameters that will be shared between the download of the origin dataset (here CMIP6) and the C3S Atlas dataset (in the [next section](section-c3s-atlas)), namely the experiment and model member.\nNext, the request to download the corresponding data from CMIP6 is defined,\nin this notebook choosing several daily and monthly variables\ndefined in the following cell.\nTo simplify the data processing,\ndaily and monthly variables are downloaded separately.\n\nVARIABLES_MONTHLY = [\"evspsbl\", \"mrsos\", \"mrro\", \"prsn\", \"sfcWind\", \"clt\", \"rsds\", \"rlds\", \"psl\"]\n\nSetup: General request\n\nDownload CMIP6 daily data\n\n```text\n Size: 404MB\nDimensions: (time: 365, lat: 192, lon: 288)\nCoordinates:\n * time (time) object 3kB 2080-01-01 12:00:00 ... 2080-12-31 12:00:00\n * lat (lat) float64 2kB -90.0 -89.06 -88.12 -87.17 ... 88.12 89.06 90.0\n * lon (lon) float64 2kB 0.0 1.25 2.5 3.75 5.0 ... 355.0 356.2 357.5 358.8\n height float64 8B 2.0\nData variables:\n tas (time, lat, lon) float32 81MB dask.array\n tasmin (time, lat, lon) float32 81MB dask.array\n tasmax (time, lat, lon) float32 81MB dask.array\n pr (time, lat, lon) float32 81MB dask.array\n huss (time, lat, lon) float32 81MB dask.array\nAttributes: (12/48)\n Conventions: CF-1.7 CMIP-6.2\n activity_id: ScenarioMIP\n branch_method: standard\n branch_time_in_child: 60225.0\n branch_time_in_parent: 60225.0\n comment: none\n ... ...\n title: CMCC-ESM2 output prepared for CMIP6\n variable_id: tas\n variant_label: r1i1p1f1\n license: CMIP6 model data produced by CMCC is licensed und...\n cmor_version: 3.6.0\n tracking_id: hdl:21.14100/25c823a7-e043-463c-a6b7-4a6f067f78a4\n```\n\nDownload CMIP6 monthly data", "text_with_prefix": "EQC Quality Assessment: \"Consistency between the C3S Atlas dataset and its origins: Multiple indicators\"\nDataset: multi-origin-c3s-atlas [CDS]\nAspect: consistency_q02 | Category: Derived_Datasets\nSection: Consistency between the C3S Atlas dataset and its origins: Multiple indicators > Analysis and results > 2. Calculate indicators from the origin dataset > Download data\n---\nThis assessment examines the dataset in one year (2080),\nfor one ensemble member\nand\nfor one climate scenario\n– which is not how climate projection data are normally used.\nGood practice in climate science is to look at multi-year statistics and trends,\nacross multiple ensemble members.\nHowever, since the purpose of this assessment is to assess the consistency and reproducibility of the post-processing performed to produce the C3S Atlas dataset,\nit is valid to use a subset of the data here.\nThe specific subset used can be easily tweaked by changing the `EXPERIMENT`, `MODEL`, and `YEARS` variables in this section.\n\nThis notebook uses [earthkit-data](https://github.com/ecmwf/earthkit-data) to download files from the CDS.\nIf you intend to run this notebook multiple times, it is highly recommended that you [enable caching](https://earthkit-data.readthedocs.io/en/latest/concepts/caching.html) to prevent having to download the same files multiple times.\nIf you prefer not to use earthkit, the following requests can also be used with the [cdsapi module](https://cds.climate.copernicus.eu/how-to-api#linux-use-client-step).\nIn either case (earthkit-data or cdsapi), it is required to set up a CDS account and API key as explained [on the CDS website](https://cds.climate.copernicus.eu/how-to-api).\n\nThe first step is to define the parameters that will be shared between the download of the origin dataset (here CMIP6) and the C3S Atlas dataset (in the [next section](section-c3s-atlas)), namely the experiment and model member.\nNext, the request to download the corresponding data from CMIP6 is defined,\nin this notebook choosing several daily and monthly variables\ndefined in the following cell.\nTo simplify the data processing,\ndaily and monthly variables are downloaded separately.\n\nVARIABLES_MONTHLY = [\"evspsbl\", \"mrsos\", \"mrro\", \"prsn\", \"sfcWind\", \"clt\", \"rsds\", \"rlds\", \"psl\"]\n\nSetup: General request\n\nDownload CMIP6 daily data\n\n```text\n Size: 404MB\nDimensions: (time: 365, lat: 192, lon: 288)\nCoordinates:\n * time (time) object 3kB 2080-01-01 12:00:00 ... 2080-12-31 12:00:00\n * lat (lat) float64 2kB -90.0 -89.06 -88.12 -87.17 ... 88.12 89.06 90.0\n * lon (lon) float64 2kB 0.0 1.25 2.5 3.75 5.0 ... 355.0 356.2 357.5 358.8\n height float64 8B 2.0\nData variables:\n tas (time, lat, lon) float32 81MB dask.array\n tasmin (time, lat, lon) float32 81MB dask.array\n tasmax (time, lat, lon) float32 81MB dask.array\n pr (time, lat, lon) float32 81MB dask.array\n huss (time, lat, lon) float32 81MB dask.array\nAttributes: (12/48)\n Conventions: CF-1.7 CMIP-6.2\n activity_id: ScenarioMIP\n branch_method: standard\n branch_time_in_child: 60225.0\n branch_time_in_parent: 60225.0\n comment: none\n ... ...\n title: CMCC-ESM2 output prepared for CMIP6\n variable_id: tas\n variant_label: r1i1p1f1\n license: CMIP6 model data produced by CMCC is licensed und...\n cmor_version: 3.6.0\n tracking_id: hdl:21.14100/25c823a7-e043-463c-a6b7-4a6f067f78a4\n```\n\nDownload CMIP6 monthly data"} {"chunk_id": "derived_multi-origin-c3s-atlas_consistency_q02__5ca05e97f18a", "report_id": "derived_multi-origin-c3s-atlas_consistency_q02", "dataset_id": "multi-origin-c3s-atlas", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q02", "aspect_base": "consistency", "category": "Derived_Datasets", "match_confidence": "exact", "section": "Consistency between the C3S Atlas dataset and its origins: Multiple indicators > Analysis and results > 2. Calculate indicators from the origin dataset > Download data", "title": "Consistency between the C3S Atlas dataset and its origins: Multiple indicators", "chunk_index": 13, "token_count": 744, "text_raw": ".0\n tracking_id: hdl:21.14100/25c823a7-e043-463c-a6b7-4a6f067f78a4\n```\n\nDownload CMIP6 monthly data\n\n```text\n Size: 19MB\nDimensions: (time: 12, lat: 192, lon: 288)\nCoordinates:\n * time (time) object 96B 2080-01-16 12:00:00 ... 2080-12-16 12:00:00\n * lat (lat) float64 2kB -90.0 -89.06 -88.12 -87.17 ... 88.12 89.06 90.0\n * lon (lon) float64 2kB 0.0 1.25 2.5 3.75 5.0 ... 355.0 356.2 357.5 358.8\n height float64 8B ...\nData variables:\n evspsbl (time, lat, lon) float32 3MB dask.array\n prsn (time, lat, lon) float32 3MB dask.array\n sfcWind (time, lat, lon) float32 3MB dask.array\n clt (time, lat, lon) float32 3MB dask.array\n rsds (time, lat, lon) float32 3MB dask.array\n rlds (time, lat, lon) float32 3MB dask.array\n psl (time, lat, lon) float32 3MB dask.array\nAttributes: (12/48)\n Conventions: CF-1.7 CMIP-6.2\n activity_id: ScenarioMIP\n branch_method: standard\n branch_time_in_child: 60225.0\n branch_time_in_parent: 60225.0\n comment: none\n ... ...\n title: CMCC-ESM2 output prepared for CMIP6\n variable_id: evspsbl\n variant_label: r1i1p1f1\n license: CMIP6 model data produced by CMCC is licensed und...\n cmor_version: 3.6.0\n tracking_id: hdl:21.14100/11eb4243-d050-40aa-bc3a-34b8c1195cb6\n```", "text_with_prefix": "EQC Quality Assessment: \"Consistency between the C3S Atlas dataset and its origins: Multiple indicators\"\nDataset: multi-origin-c3s-atlas [CDS]\nAspect: consistency_q02 | Category: Derived_Datasets\nSection: Consistency between the C3S Atlas dataset and its origins: Multiple indicators > Analysis and results > 2. Calculate indicators from the origin dataset > Download data\n---\n.0\n tracking_id: hdl:21.14100/25c823a7-e043-463c-a6b7-4a6f067f78a4\n```\n\nDownload CMIP6 monthly data\n\n```text\n Size: 19MB\nDimensions: (time: 12, lat: 192, lon: 288)\nCoordinates:\n * time (time) object 96B 2080-01-16 12:00:00 ... 2080-12-16 12:00:00\n * lat (lat) float64 2kB -90.0 -89.06 -88.12 -87.17 ... 88.12 89.06 90.0\n * lon (lon) float64 2kB 0.0 1.25 2.5 3.75 5.0 ... 355.0 356.2 357.5 358.8\n height float64 8B ...\nData variables:\n evspsbl (time, lat, lon) float32 3MB dask.array\n prsn (time, lat, lon) float32 3MB dask.array\n sfcWind (time, lat, lon) float32 3MB dask.array\n clt (time, lat, lon) float32 3MB dask.array\n rsds (time, lat, lon) float32 3MB dask.array\n rlds (time, lat, lon) float32 3MB dask.array\n psl (time, lat, lon) float32 3MB dask.array\nAttributes: (12/48)\n Conventions: CF-1.7 CMIP-6.2\n activity_id: ScenarioMIP\n branch_method: standard\n branch_time_in_child: 60225.0\n branch_time_in_parent: 60225.0\n comment: none\n ... ...\n title: CMCC-ESM2 output prepared for CMIP6\n variable_id: evspsbl\n variant_label: r1i1p1f1\n license: CMIP6 model data produced by CMCC is licensed und...\n cmor_version: 3.6.0\n tracking_id: hdl:21.14100/11eb4243-d050-40aa-bc3a-34b8c1195cb6\n```"} {"chunk_id": "derived_multi-origin-c3s-atlas_consistency_q02__9544489c79a4", "report_id": "derived_multi-origin-c3s-atlas_consistency_q02", "dataset_id": "multi-origin-c3s-atlas", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q02", "aspect_base": "consistency", "category": "Derived_Datasets", "match_confidence": "exact", "section": "Consistency between the C3S Atlas dataset and its origins: Multiple indicators > Analysis and results > 2. Calculate indicators from the origin dataset > Homogenise data", "title": "Consistency between the C3S Atlas dataset and its origins: Multiple indicators", "chunk_index": 14, "token_count": 307, "text_raw": "One of the steps in the C3S Atlas dataset production chain is homogenisation, i.e. ensuring consistency between data from different origin datasets.\nThis homogenisation is implemented in the [User-tools for the C3S Atlas](https://github.com/ecmwf-projects/c3s-atlas/tree/main/c3s_atlas), specifically the `c3s_atlas.fixers.apply_fixers` function.\nThe following changes are applied:\n\n- The names of the spatial coordinates are standardised to `[lon, lat]`.\n- Longitude is converted from `[0...360]` to `[-180...180]` format.\n- The time coordinate is standardised to the CF standard calendar.\n- Variable units are standardised (e.g. °C for temperature).\n- Variables are resampled / aggregated to the required temporal resolution.\n\nThe homogenisation is applied in the following code cells, separately for the daily and monthly data.\nMore details can be found in the [case study for `tx35`](./derived_multi-origin-c3s-atlas_consistency_q01).", "text_with_prefix": "EQC Quality Assessment: \"Consistency between the C3S Atlas dataset and its origins: Multiple indicators\"\nDataset: multi-origin-c3s-atlas [CDS]\nAspect: consistency_q02 | Category: Derived_Datasets\nSection: Consistency between the C3S Atlas dataset and its origins: Multiple indicators > Analysis and results > 2. Calculate indicators from the origin dataset > Homogenise data\n---\nOne of the steps in the C3S Atlas dataset production chain is homogenisation, i.e. ensuring consistency between data from different origin datasets.\nThis homogenisation is implemented in the [User-tools for the C3S Atlas](https://github.com/ecmwf-projects/c3s-atlas/tree/main/c3s_atlas), specifically the `c3s_atlas.fixers.apply_fixers` function.\nThe following changes are applied:\n\n- The names of the spatial coordinates are standardised to `[lon, lat]`.\n- Longitude is converted from `[0...360]` to `[-180...180]` format.\n- The time coordinate is standardised to the CF standard calendar.\n- Variable units are standardised (e.g. °C for temperature).\n- Variables are resampled / aggregated to the required temporal resolution.\n\nThe homogenisation is applied in the following code cells, separately for the daily and monthly data.\nMore details can be found in the [case study for `tx35`](./derived_multi-origin-c3s-atlas_consistency_q01)."} {"chunk_id": "derived_multi-origin-c3s-atlas_consistency_q02__691743e20aa5", "report_id": "derived_multi-origin-c3s-atlas_consistency_q02", "dataset_id": "multi-origin-c3s-atlas", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q02", "aspect_base": "consistency", "category": "Derived_Datasets", "match_confidence": "exact", "section": "Consistency between the C3S Atlas dataset and its origins: Multiple indicators > Analysis and results > 2. Calculate indicators from the origin dataset > Calculate indicators", "title": "Consistency between the C3S Atlas dataset and its origins: Multiple indicators", "chunk_index": 15, "token_count": 770, "text_raw": "The climate indicators\nare calculated using [xclim](https://xclim.readthedocs.io/en/stable/).\nThe functions defined [above](section-codesetup) perform the calculations.\n\nAnnual indicators temporarily disabled due to issues with CF metadata\nindicators_cmip6_annual = calculate_indicators_annual(data_cmip6_daily_homogenised)\nindicators_cmip6_annual\n\n```text\n Size: 90MB\nDimensions: (lat: 192, lon: 288, time: 12)\nCoordinates:\n * lat (lat) float64 2kB -90.0 -89.06 -88.12 -87.17 ... 88.12 89.06 90.0\n * lon (lon) float64 2kB -178.8 -177.5 -176.2 -175.0 ... 177.5 178.8 180.0\n height float64 8B 2.0\n * time (time) datetime64[ns] 96B 2080-01-01 2080-02-01 ... 2080-12-01\nData variables: (12/25)\n t (time, lat, lon) float32 3MB dask.array\n tn (time, lat, lon) float32 3MB dask.array\n tx (time, lat, lon) float32 3MB dask.array\n dtr (time, lat, lon) float32 3MB dask.array\n tnn (time, lat, lon) float32 3MB dask.array\n txx (time, lat, lon) float32 3MB dask.array\n ... ...\n evspsbl (time, lat, lon) float32 3MB dask.array\n psl (time, lat, lon) float32 3MB dask.array\n sfcwind (time, lat, lon) float32 3MB dask.array\n clt (time, lat, lon) float32 3MB dask.array\n rsds (time, lat, lon) float32 3MB dask.array\n rlds (time, lat, lon) float32 3MB dask.array\n```", "text_with_prefix": "EQC Quality Assessment: \"Consistency between the C3S Atlas dataset and its origins: Multiple indicators\"\nDataset: multi-origin-c3s-atlas [CDS]\nAspect: consistency_q02 | Category: Derived_Datasets\nSection: Consistency between the C3S Atlas dataset and its origins: Multiple indicators > Analysis and results > 2. Calculate indicators from the origin dataset > Calculate indicators\n---\nThe climate indicators\nare calculated using [xclim](https://xclim.readthedocs.io/en/stable/).\nThe functions defined [above](section-codesetup) perform the calculations.\n\nAnnual indicators temporarily disabled due to issues with CF metadata\nindicators_cmip6_annual = calculate_indicators_annual(data_cmip6_daily_homogenised)\nindicators_cmip6_annual\n\n```text\n Size: 90MB\nDimensions: (lat: 192, lon: 288, time: 12)\nCoordinates:\n * lat (lat) float64 2kB -90.0 -89.06 -88.12 -87.17 ... 88.12 89.06 90.0\n * lon (lon) float64 2kB -178.8 -177.5 -176.2 -175.0 ... 177.5 178.8 180.0\n height float64 8B 2.0\n * time (time) datetime64[ns] 96B 2080-01-01 2080-02-01 ... 2080-12-01\nData variables: (12/25)\n t (time, lat, lon) float32 3MB dask.array\n tn (time, lat, lon) float32 3MB dask.array\n tx (time, lat, lon) float32 3MB dask.array\n dtr (time, lat, lon) float32 3MB dask.array\n tnn (time, lat, lon) float32 3MB dask.array\n txx (time, lat, lon) float32 3MB dask.array\n ... ...\n evspsbl (time, lat, lon) float32 3MB dask.array\n psl (time, lat, lon) float32 3MB dask.array\n sfcwind (time, lat, lon) float32 3MB dask.array\n clt (time, lat, lon) float32 3MB dask.array\n rsds (time, lat, lon) float32 3MB dask.array\n rlds (time, lat, lon) float32 3MB dask.array\n```"} {"chunk_id": "derived_multi-origin-c3s-atlas_consistency_q02__a034f088e06a", "report_id": "derived_multi-origin-c3s-atlas_consistency_q02", "dataset_id": "multi-origin-c3s-atlas", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q02", "aspect_base": "consistency", "category": "Derived_Datasets", "match_confidence": "exact", "section": "Consistency between the C3S Atlas dataset and its origins: Multiple indicators > Analysis and results > 2. Calculate indicators from the origin dataset > Regrid to C3S Atlas grid", "title": "Consistency between the C3S Atlas dataset and its origins: Multiple indicators", "chunk_index": 16, "token_count": 1080, "text_raw": "This notebook uses [xESMF](https://github.com/pangeo-data/xESMF) for regridding data.\nxESMF is most easily installed using mamba/conda as explained in its documentation.\nUsers who cannot or do not wish to use mamba/conda can manually compile and install [ESMF](https://earthsystemmodeling.org/docs/release/latest/ESMF_usrdoc/node10.html) on their machines.\nIn future, this notebook will use [earthkit-regrid](https://github.com/ecmwf/earthkit-regrid) instead, once it reaches suitable maturity.\n```\n\nThe final step in the processing is regridding to the standardised grid used in the C3S Atlas dataset (Figure {numref}`{number} `).\nThis is performed through a custom function in the [User-tools for the C3S Atlas](https://github.com/ecmwf-projects/c3s-atlas/tree/main/c3s_atlas),\nspecifically `c3s_atlas.interpolation`.\nThis function is based on ESMF, as noted above.\n\nNote that the C3S Atlas workflow calculates indicators first, then regrids.\nFor operations that involve averaging, like smoothing and regridding, the order of operations can affect the result, especially in areas with steep gradients [[Avila+15](https://doi.org/10.1016/j.wace.2015.06.003)].\nExamples of such areas for a temperature index are coastlines and mountain ranges.\nIn the case of C3S Atlas, this order of operations was a conscious choice to preserve the \"raw\" signals,\ne.g. preventing extreme temperatures from being smoothed out.\nHowever, it can affect the indicator values and therefore must be considered when using the C3S Atlas application or dataset,\nparticularly when intercomparing it with another dataset.\n\nDue to the number of variables involved, this section can take several minutes to run.\n\n```text\n Size: 106MB\nDimensions: (lon: 360, lat: 180, time: 12, bnds: 2)\nCoordinates:\n * lon (lon) float64 3kB -179.5 -178.5 -177.5 ... 177.5 178.5 179.5\n * lat (lat) float64 1kB -89.5 -88.5 -87.5 -86.5 ... 86.5 87.5 88.5 89.5\n * time (time) datetime64[ns] 96B 2080-01-01 2080-02-01 ... 2080-12-01\nDimensions without coordinates: bnds\nData variables: (12/29)\n t (time, lat, lon) float32 3MB dask.array\n lon_bnds (lon, bnds) float64 6kB -180.0 -179.0 -179.0 ... 179.0 179.0 180.0\n lat_bnds (lat, bnds) float64 3kB -90.0 -89.0 -89.0 -88.0 ... 89.0 89.0 90.0\n crs int64 8B 0\n height float64 8B 2.0\n tn (time, lat, lon) float32 3MB dask.array\n ... ...\n evspsbl (time, lat, lon) float32 3MB dask.array\n psl (time, lat, lon) float32 3MB dask.array\n sfcwind (time, lat, lon) float32 3MB dask.array\n clt (time, lat, lon) float32 3MB dask.array\n rsds (time, lat, lon) float32 3MB dask.array\n rlds (time, lat, lon) float32 3MB dask.array\n```", "text_with_prefix": "EQC Quality Assessment: \"Consistency between the C3S Atlas dataset and its origins: Multiple indicators\"\nDataset: multi-origin-c3s-atlas [CDS]\nAspect: consistency_q02 | Category: Derived_Datasets\nSection: Consistency between the C3S Atlas dataset and its origins: Multiple indicators > Analysis and results > 2. Calculate indicators from the origin dataset > Regrid to C3S Atlas grid\n---\nThis notebook uses [xESMF](https://github.com/pangeo-data/xESMF) for regridding data.\nxESMF is most easily installed using mamba/conda as explained in its documentation.\nUsers who cannot or do not wish to use mamba/conda can manually compile and install [ESMF](https://earthsystemmodeling.org/docs/release/latest/ESMF_usrdoc/node10.html) on their machines.\nIn future, this notebook will use [earthkit-regrid](https://github.com/ecmwf/earthkit-regrid) instead, once it reaches suitable maturity.\n```\n\nThe final step in the processing is regridding to the standardised grid used in the C3S Atlas dataset (Figure {numref}`{number} `).\nThis is performed through a custom function in the [User-tools for the C3S Atlas](https://github.com/ecmwf-projects/c3s-atlas/tree/main/c3s_atlas),\nspecifically `c3s_atlas.interpolation`.\nThis function is based on ESMF, as noted above.\n\nNote that the C3S Atlas workflow calculates indicators first, then regrids.\nFor operations that involve averaging, like smoothing and regridding, the order of operations can affect the result, especially in areas with steep gradients [[Avila+15](https://doi.org/10.1016/j.wace.2015.06.003)].\nExamples of such areas for a temperature index are coastlines and mountain ranges.\nIn the case of C3S Atlas, this order of operations was a conscious choice to preserve the \"raw\" signals,\ne.g. preventing extreme temperatures from being smoothed out.\nHowever, it can affect the indicator values and therefore must be considered when using the C3S Atlas application or dataset,\nparticularly when intercomparing it with another dataset.\n\nDue to the number of variables involved, this section can take several minutes to run.\n\n```text\n Size: 106MB\nDimensions: (lon: 360, lat: 180, time: 12, bnds: 2)\nCoordinates:\n * lon (lon) float64 3kB -179.5 -178.5 -177.5 ... 177.5 178.5 179.5\n * lat (lat) float64 1kB -89.5 -88.5 -87.5 -86.5 ... 86.5 87.5 88.5 89.5\n * time (time) datetime64[ns] 96B 2080-01-01 2080-02-01 ... 2080-12-01\nDimensions without coordinates: bnds\nData variables: (12/29)\n t (time, lat, lon) float32 3MB dask.array\n lon_bnds (lon, bnds) float64 6kB -180.0 -179.0 -179.0 ... 179.0 179.0 180.0\n lat_bnds (lat, bnds) float64 3kB -90.0 -89.0 -89.0 -88.0 ... 89.0 89.0 90.0\n crs int64 8B 0\n height float64 8B 2.0\n tn (time, lat, lon) float32 3MB dask.array\n ... ...\n evspsbl (time, lat, lon) float32 3MB dask.array\n psl (time, lat, lon) float32 3MB dask.array\n sfcwind (time, lat, lon) float32 3MB dask.array\n clt (time, lat, lon) float32 3MB dask.array\n rsds (time, lat, lon) float32 3MB dask.array\n rlds (time, lat, lon) float32 3MB dask.array\n```"} {"chunk_id": "derived_multi-origin-c3s-atlas_consistency_q02__ae115c791447", "report_id": "derived_multi-origin-c3s-atlas_consistency_q02", "dataset_id": "multi-origin-c3s-atlas", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q02", "aspect_base": "consistency", "category": "Derived_Datasets", "match_confidence": "exact", "section": "Consistency between the C3S Atlas dataset and its origins: Multiple indicators > Analysis and results > 2. Calculate indicators from the origin dataset > Regrid to C3S Atlas grid", "title": "Consistency between the C3S Atlas dataset and its origins: Multiple indicators", "chunk_index": 17, "token_count": 187, "text_raw": "=(1, 180, 360), meta=np.ndarray>\n rlds (time, lat, lon) float32 3MB dask.array\n```\n\nAnnual indicators\nindicators_cmip6_annual_regridded = regrid_multiple_variables(indicators_cmip6_annual)\nindicators_cmip6_annual_regridded\n\n(section-c3s-atlas)=", "text_with_prefix": "EQC Quality Assessment: \"Consistency between the C3S Atlas dataset and its origins: Multiple indicators\"\nDataset: multi-origin-c3s-atlas [CDS]\nAspect: consistency_q02 | Category: Derived_Datasets\nSection: Consistency between the C3S Atlas dataset and its origins: Multiple indicators > Analysis and results > 2. Calculate indicators from the origin dataset > Regrid to C3S Atlas grid\n---\n=(1, 180, 360), meta=np.ndarray>\n rlds (time, lat, lon) float32 3MB dask.array\n```\n\nAnnual indicators\nindicators_cmip6_annual_regridded = regrid_multiple_variables(indicators_cmip6_annual)\nindicators_cmip6_annual_regridded\n\n(section-c3s-atlas)="} {"chunk_id": "derived_multi-origin-c3s-atlas_consistency_q02__3d1bed2976f3", "report_id": "derived_multi-origin-c3s-atlas_consistency_q02", "dataset_id": "multi-origin-c3s-atlas", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q02", "aspect_base": "consistency", "category": "Derived_Datasets", "match_confidence": "exact", "section": "Consistency between the C3S Atlas dataset and its origins: Multiple indicators > Analysis and results > 3. Retrieve indicators from the C3S Atlas dataset", "title": "Consistency between the C3S Atlas dataset and its origins: Multiple indicators", "chunk_index": 18, "token_count": 372, "text_raw": "Here, we download the same indicators as above directly from the [Gridded dataset underpinning the Copernicus Interactive Climate Atlas](https://doi.org/10.24381/cds.h35hb680) so the values can be compared.\n\nWhen downloading the C3S Atlas dataset from the CDS, it is not possible to specify a specific year or model like it is for e.g. CMIP6.\nInstead, data are downloaded in blocks of several decades and include all model members.\nTo prevent memory issues from loading so many data at once,\na custom function is used that downloads the data for each indicator and pulls out only the year and model member of interest.\nBecause this is done in sequence\n(one indicator after another)\nrather than in parallel (all at the same time, like a normal earthkit-data download),\nthis can result in the following cells taking relatively long (>1 hour) to run the first time.\nAfter the first time, caching in earthkit-data will (if enabled) speed this section up significantly.\n\nSetup: General request\n\nDownload C3S Atlas data\nMonthly indicators\nAnnual indicators\nrequests_c3s_atlas_annual = create_requests_for_variables(request_c3s_atlas, indicators_annual_names)\nindicators_c3s_atlas_annual = download_c3s_atlas_data_onemember(requests_c3s_atlas_annual, YEARS, MODEL)", "text_with_prefix": "EQC Quality Assessment: \"Consistency between the C3S Atlas dataset and its origins: Multiple indicators\"\nDataset: multi-origin-c3s-atlas [CDS]\nAspect: consistency_q02 | Category: Derived_Datasets\nSection: Consistency between the C3S Atlas dataset and its origins: Multiple indicators > Analysis and results > 3. Retrieve indicators from the C3S Atlas dataset\n---\nHere, we download the same indicators as above directly from the [Gridded dataset underpinning the Copernicus Interactive Climate Atlas](https://doi.org/10.24381/cds.h35hb680) so the values can be compared.\n\nWhen downloading the C3S Atlas dataset from the CDS, it is not possible to specify a specific year or model like it is for e.g. CMIP6.\nInstead, data are downloaded in blocks of several decades and include all model members.\nTo prevent memory issues from loading so many data at once,\na custom function is used that downloads the data for each indicator and pulls out only the year and model member of interest.\nBecause this is done in sequence\n(one indicator after another)\nrather than in parallel (all at the same time, like a normal earthkit-data download),\nthis can result in the following cells taking relatively long (>1 hour) to run the first time.\nAfter the first time, caching in earthkit-data will (if enabled) speed this section up significantly.\n\nSetup: General request\n\nDownload C3S Atlas data\nMonthly indicators\nAnnual indicators\nrequests_c3s_atlas_annual = create_requests_for_variables(request_c3s_atlas, indicators_annual_names)\nindicators_c3s_atlas_annual = download_c3s_atlas_data_onemember(requests_c3s_atlas_annual, YEARS, MODEL)"} {"chunk_id": "derived_multi-origin-c3s-atlas_consistency_q02__c30eaecc9a07", "report_id": "derived_multi-origin-c3s-atlas_consistency_q02", "dataset_id": "multi-origin-c3s-atlas", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q02", "aspect_base": "consistency", "category": "Derived_Datasets", "match_confidence": "exact", "section": "Consistency between the C3S Atlas dataset and its origins: Multiple indicators > Analysis and results > 3. Retrieve indicators from the C3S Atlas dataset", "title": "Consistency between the C3S Atlas dataset and its origins: Multiple indicators", "chunk_index": 19, "token_count": 991, "text_raw": "(requests_c3s_atlas_annual, YEARS, MODEL)\n\n```text\n Size: 78MB\nDimensions: (time: 12, lat: 180, lon: 360)\nCoordinates: (12/16)\n * lat (lat) float64 1kB -89.5 -88.5 -87.5 ... 87.5 88.5 89.5\n * lon (lon) float64 3kB -179.5 -178.5 -177.5 ... 178.5 179.5\n * time (time) datetime64[ns] 96B 2080-01-01 ... 2080-12-01\n member_id \n crs int32 4B -2147483647\n tn (time, lat, lon) float32 3MB dask.array\n tx (time, lat, lon) float32 3MB dask.array\n dtr (time, lat, lon) float32 3MB dask.array\n tnn (time, lat, lon) float32 3MB dask.array\n ... ...\n huss (time, lat, lon) float32 3MB dask.array\n psl (time, lat, lon) float32 3MB dask.array\n sfcwind (time, lat, lon) float32 3MB dask.array\n clt (time, lat, lon) float32 3MB dask.array\n rsds (time, lat, lon) float32 3MB dask.array\n rlds (time, lat, lon) float32 3MB dask.array\nAttributes: (12/26)\n Conventions: CF-1.9 ACDD-1.3\n title: Copernicus Interactive Climate Atlas: gridded...\n summary: Monthly/annual gridded data from observations...\n institution: Copernicus Climate Change Service (C3S)\n producers: Institute of Physics of Cantabria (IFCA, CSIC...\n license: CC-BY 4.0, https://creativecommons.org/licens...\n ... ...\n geospatial_lon_min: -180.0\n geospatial_lon_max: 180.0\n geospatial_lon_resolution: 1.0\n geospatial_lon_units: degrees_east\n date_created: 2024-12-05 15:51:48.916442+01:00\n tracking_id: 89974d82-56b6-4b47-958e-5954716013d3\n```\n\nindicators_c3s_atlas_annual\n\n(section-results)=", "text_with_prefix": "EQC Quality Assessment: \"Consistency between the C3S Atlas dataset and its origins: Multiple indicators\"\nDataset: multi-origin-c3s-atlas [CDS]\nAspect: consistency_q02 | Category: Derived_Datasets\nSection: Consistency between the C3S Atlas dataset and its origins: Multiple indicators > Analysis and results > 3. Retrieve indicators from the C3S Atlas dataset\n---\n(requests_c3s_atlas_annual, YEARS, MODEL)\n\n```text\n Size: 78MB\nDimensions: (time: 12, lat: 180, lon: 360)\nCoordinates: (12/16)\n * lat (lat) float64 1kB -89.5 -88.5 -87.5 ... 87.5 88.5 89.5\n * lon (lon) float64 3kB -179.5 -178.5 -177.5 ... 178.5 179.5\n * time (time) datetime64[ns] 96B 2080-01-01 ... 2080-12-01\n member_id \n crs int32 4B -2147483647\n tn (time, lat, lon) float32 3MB dask.array\n tx (time, lat, lon) float32 3MB dask.array\n dtr (time, lat, lon) float32 3MB dask.array\n tnn (time, lat, lon) float32 3MB dask.array\n ... ...\n huss (time, lat, lon) float32 3MB dask.array\n psl (time, lat, lon) float32 3MB dask.array\n sfcwind (time, lat, lon) float32 3MB dask.array\n clt (time, lat, lon) float32 3MB dask.array\n rsds (time, lat, lon) float32 3MB dask.array\n rlds (time, lat, lon) float32 3MB dask.array\nAttributes: (12/26)\n Conventions: CF-1.9 ACDD-1.3\n title: Copernicus Interactive Climate Atlas: gridded...\n summary: Monthly/annual gridded data from observations...\n institution: Copernicus Climate Change Service (C3S)\n producers: Institute of Physics of Cantabria (IFCA, CSIC...\n license: CC-BY 4.0, https://creativecommons.org/licens...\n ... ...\n geospatial_lon_min: -180.0\n geospatial_lon_max: 180.0\n geospatial_lon_resolution: 1.0\n geospatial_lon_units: degrees_east\n date_created: 2024-12-05 15:51:48.916442+01:00\n tracking_id: 89974d82-56b6-4b47-958e-5954716013d3\n```\n\nindicators_c3s_atlas_annual\n\n(section-results)="} {"chunk_id": "derived_multi-origin-c3s-atlas_consistency_q02__67d12203dbe1", "report_id": "derived_multi-origin-c3s-atlas_consistency_q02", "dataset_id": "multi-origin-c3s-atlas", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q02", "aspect_base": "consistency", "category": "Derived_Datasets", "match_confidence": "exact", "section": "Consistency between the C3S Atlas dataset and its origins: Multiple indicators > Analysis and results > 4. Results", "title": "Consistency between the C3S Atlas dataset and its origins: Multiple indicators", "chunk_index": 20, "token_count": 447, "text_raw": "This section contains the comparison between the indicator values retrieved from the C3S Atlas dataset vs those reproduced from the origin dataset.\n\nThe datasets are first compared on their native grids.\nThis means a point-by-point comparison is not possible\n(because the points are not equivalent),\nbut the distributions can be compared geospatially and overall.\nThis qualitative comparison probes the consistency quality attribute:\nAre the climate indicators in the dataset underpinning the Copernicus Interactive Climate Atlas consistent with their origin datasets?\n\nSecond, the C3S Atlas dataset is compared to the indicators derived from the origin dataset and regridded to the C3S Atlas grid.\nThis makes a quantitative point-by-point comparison possible.\nThis second comparison probes how well the dataset underpinning the Copernicus Interactive Climate Atlas can be reproduced from its origin datasets,\nbased on the workflow (Figure {numref}`{number} `).\n\nFor the geospatial comparison, \nwe display the values of the indicators for one month,\nacross one region and globally.\nAs an example, we display the results across\nEurope\nin\nJune,\nwhich should provide significant spatial variation.\nThis region can easily be modified in the following code cell using the [domains provided by earthkit-plots](https://earthkit-plots.readthedocs.io/en/latest/examples/examples/introduction/07-domains.html).\nSome examples are provided in the cell (commented out using `#`).\n\nSetup: Choose a month to display\nSetup: Pick domain using earthkit-plots\ndomain = \"Mediterranean\"\nOther examples -- uncomment where desired\ndomain = ekp.geo.domains.union([\"Portugal\", \"Spain\"], name=\"Iberia\")\ndomain = \"Italy\"\ndomain = \"South America\"", "text_with_prefix": "EQC Quality Assessment: \"Consistency between the C3S Atlas dataset and its origins: Multiple indicators\"\nDataset: multi-origin-c3s-atlas [CDS]\nAspect: consistency_q02 | Category: Derived_Datasets\nSection: Consistency between the C3S Atlas dataset and its origins: Multiple indicators > Analysis and results > 4. Results\n---\nThis section contains the comparison between the indicator values retrieved from the C3S Atlas dataset vs those reproduced from the origin dataset.\n\nThe datasets are first compared on their native grids.\nThis means a point-by-point comparison is not possible\n(because the points are not equivalent),\nbut the distributions can be compared geospatially and overall.\nThis qualitative comparison probes the consistency quality attribute:\nAre the climate indicators in the dataset underpinning the Copernicus Interactive Climate Atlas consistent with their origin datasets?\n\nSecond, the C3S Atlas dataset is compared to the indicators derived from the origin dataset and regridded to the C3S Atlas grid.\nThis makes a quantitative point-by-point comparison possible.\nThis second comparison probes how well the dataset underpinning the Copernicus Interactive Climate Atlas can be reproduced from its origin datasets,\nbased on the workflow (Figure {numref}`{number} `).\n\nFor the geospatial comparison, \nwe display the values of the indicators for one month,\nacross one region and globally.\nAs an example, we display the results across\nEurope\nin\nJune,\nwhich should provide significant spatial variation.\nThis region can easily be modified in the following code cell using the [domains provided by earthkit-plots](https://earthkit-plots.readthedocs.io/en/latest/examples/examples/introduction/07-domains.html).\nSome examples are provided in the cell (commented out using `#`).\n\nSetup: Choose a month to display\nSetup: Pick domain using earthkit-plots\ndomain = \"Mediterranean\"\nOther examples -- uncomment where desired\ndomain = ekp.geo.domains.union([\"Portugal\", \"Spain\"], name=\"Iberia\")\ndomain = \"Italy\"\ndomain = \"South America\""} {"chunk_id": "derived_multi-origin-c3s-atlas_consistency_q02__9e48c7b2ce39", "report_id": "derived_multi-origin-c3s-atlas_consistency_q02", "dataset_id": "multi-origin-c3s-atlas", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q02", "aspect_base": "consistency", "category": "Derived_Datasets", "match_confidence": "exact", "section": "Consistency between the C3S Atlas dataset and its origins: Multiple indicators > Analysis and results > 4. Results > Consistency: Comparison on native grids", "title": "Consistency between the C3S Atlas dataset and its origins: Multiple indicators", "chunk_index": 21, "token_count": 805, "text_raw": "As in the [case study for `tx35`](./derived_multi-origin-c3s-atlas_consistency_q01),\nit is clear from the geospatial comparison\n(Figure {numref}`{number} `)\nthat the C3S Atlas dataset closely resembles a manual reproduction from its origin datasets.\nThe general distribution of indicator values is the same for\nall indicators.\nIn a few cases,\nsuch as\n`sdii` (northern Africa, north of the Caspian Sea)\nand\n`pet` (northern Africa, seas, within the Arctic circle),\nthe two datasets differ in terms of coverage.\nFor `pet`,\nthis difference is due to a deliberate decision in the C3S Atlas workflow \nto mask areas where agricultural studies do not make sense,\nsuch as arid regions, areas permanently covered by ice, and regions with no vegetation.\nThis masking is explained in a [Jupyter notebook written by the data provider](https://github.com/ecmwf-projects/c3s-atlas/blob/main/auxiliar/barrean_mask.ipynb).\n\nThis pattern is also visible in the overall distributions\n(Figure {numref}`{number} `),\nwhich are again very similar for almost all indicators.\nThe distributions only differ clearly for\n`sdii`\nand\n`pet`,\ndue to the aforementioned gaps and masks.\n\nThe overall conclusion from this comparison is the same as in the [case study for `tx35`](./derived_multi-origin-c3s-atlas_consistency_q01).\nThe C3S Atlas dataset and its origins are highly consistent,\nbut small differences exist due to the difference in grid.\nThere are also differences in data availability for some indicators,\ndue to deliberate masking of specific areas.\nUsers of the C3S Atlas dataset\n– and thus users of the C3S Atlas application –\nshould be aware that the indicator values retrieved for a specific location may differ slightly from a manual analysis of the origin dataset.\n\nGeospatial plot: Monthly indicators\nGeospatial plot: Annual indicators\ngeospatial_comparison_multiple_indicators(indicators_c3s_atlas_annual, indicators_cmip6_annual, indicators_annual_names, 1, domain=domain)\n\nHistogram: Monthly indicators\nHistogram: Annual indicators\nfig_hist = histogram_comparison_by_indicator(indicators_c3s_atlas_annual, indicators_cmip6_annual, indicators_annual_names)\n\n::::{tab-set}\n:::{tab-item} Geospatial comparison\n:sync: geo\n derived_multi-origin-c3s-atlas_consistency_q02_fig-geo-regional\n:figwidth: 700px\n:name: \"derived_multi-origin-c3s-atlas_consistency_q02_fig-geo-regional\"\n\nComparison between C3S Atlas dataset and reproduction for multiple indicators in one month,\nacross Europe,\non the native grid of each dataset.\n```\n:::\n:::{tab-item} Overall comparison\n:sync: hist\n derived_multi-origin-c3s-atlas_consistency_q02_fig-hist-native\n:figwidth: 700px\n:name: \"derived_multi-origin-c3s-atlas_consistency_q02_fig-hist-native\"\n\nComparison between overall distributions of indicator values in the C3S Atlas dataset and its reproduction,\nacross all spatial and temporal dimensions,\non the native grid of each dataset.\n```\n:::\n::::", "text_with_prefix": "EQC Quality Assessment: \"Consistency between the C3S Atlas dataset and its origins: Multiple indicators\"\nDataset: multi-origin-c3s-atlas [CDS]\nAspect: consistency_q02 | Category: Derived_Datasets\nSection: Consistency between the C3S Atlas dataset and its origins: Multiple indicators > Analysis and results > 4. Results > Consistency: Comparison on native grids\n---\nAs in the [case study for `tx35`](./derived_multi-origin-c3s-atlas_consistency_q01),\nit is clear from the geospatial comparison\n(Figure {numref}`{number} `)\nthat the C3S Atlas dataset closely resembles a manual reproduction from its origin datasets.\nThe general distribution of indicator values is the same for\nall indicators.\nIn a few cases,\nsuch as\n`sdii` (northern Africa, north of the Caspian Sea)\nand\n`pet` (northern Africa, seas, within the Arctic circle),\nthe two datasets differ in terms of coverage.\nFor `pet`,\nthis difference is due to a deliberate decision in the C3S Atlas workflow \nto mask areas where agricultural studies do not make sense,\nsuch as arid regions, areas permanently covered by ice, and regions with no vegetation.\nThis masking is explained in a [Jupyter notebook written by the data provider](https://github.com/ecmwf-projects/c3s-atlas/blob/main/auxiliar/barrean_mask.ipynb).\n\nThis pattern is also visible in the overall distributions\n(Figure {numref}`{number} `),\nwhich are again very similar for almost all indicators.\nThe distributions only differ clearly for\n`sdii`\nand\n`pet`,\ndue to the aforementioned gaps and masks.\n\nThe overall conclusion from this comparison is the same as in the [case study for `tx35`](./derived_multi-origin-c3s-atlas_consistency_q01).\nThe C3S Atlas dataset and its origins are highly consistent,\nbut small differences exist due to the difference in grid.\nThere are also differences in data availability for some indicators,\ndue to deliberate masking of specific areas.\nUsers of the C3S Atlas dataset\n– and thus users of the C3S Atlas application –\nshould be aware that the indicator values retrieved for a specific location may differ slightly from a manual analysis of the origin dataset.\n\nGeospatial plot: Monthly indicators\nGeospatial plot: Annual indicators\ngeospatial_comparison_multiple_indicators(indicators_c3s_atlas_annual, indicators_cmip6_annual, indicators_annual_names, 1, domain=domain)\n\nHistogram: Monthly indicators\nHistogram: Annual indicators\nfig_hist = histogram_comparison_by_indicator(indicators_c3s_atlas_annual, indicators_cmip6_annual, indicators_annual_names)\n\n::::{tab-set}\n:::{tab-item} Geospatial comparison\n:sync: geo\n derived_multi-origin-c3s-atlas_consistency_q02_fig-geo-regional\n:figwidth: 700px\n:name: \"derived_multi-origin-c3s-atlas_consistency_q02_fig-geo-regional\"\n\nComparison between C3S Atlas dataset and reproduction for multiple indicators in one month,\nacross Europe,\non the native grid of each dataset.\n```\n:::\n:::{tab-item} Overall comparison\n:sync: hist\n derived_multi-origin-c3s-atlas_consistency_q02_fig-hist-native\n:figwidth: 700px\n:name: \"derived_multi-origin-c3s-atlas_consistency_q02_fig-hist-native\"\n\nComparison between overall distributions of indicator values in the C3S Atlas dataset and its reproduction,\nacross all spatial and temporal dimensions,\non the native grid of each dataset.\n```\n:::\n::::"} {"chunk_id": "derived_multi-origin-c3s-atlas_consistency_q02__e6eb5f9ee60a", "report_id": "derived_multi-origin-c3s-atlas_consistency_q02", "dataset_id": "multi-origin-c3s-atlas", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q02", "aspect_base": "consistency", "category": "Derived_Datasets", "match_confidence": "exact", "section": "Consistency between the C3S Atlas dataset and its origins: Multiple indicators > Analysis and results > 4. Results > Reproducibility: Comparison on C3S Atlas grid", "title": "Consistency between the C3S Atlas dataset and its origins: Multiple indicators", "chunk_index": 22, "token_count": 1004, "text_raw": "After regridding to the C3S Atlas grid,\nthe indicator values reproduced from the origin dataset\ncan be compared point-by-point to the values retrieved from the C3S Atlas dataset.\nWe first examine some metrics that describe the difference Δ between corresponding pixels:\n\nAs in the [case study for `tx35`](./derived_multi-origin-c3s-atlas_consistency_q01),\nit is clear that\nthe C3S Atlas dataset and its manual reproduction\nare very similar.\nThe median difference, median absolute difference, and median absolute percentage difference\nare all close to 0 for\nalmost all indicators.\n\nThe percentage of pixels with a non-zero difference\n(defined here as |Δ| ≥ ε with ε = 10{sup}`–5` to avoid floating-point errors)\nis a less useful metric here since most of the indicators are\nreal numbers,\nrather than integers like `tx35`.\nThis leaves more room for small differences in the data to appear due to\nsmall changes in the origin dataset and/or the C3S Atlas workflow software,\nwhich for an integer indicator like `tx35` would tend to be rounded off.\nFor most indicators,\nthe relative difference between datasets is ≤0.1% for\nthe vast majority of point-by-point comparisons.\n\nThe distributions of indicator values and differences\n(Figure {numref}`{number} `)\nconfirm that the differences between the two datasets tend to be very small,\nwith distributions centered around 0 and few outliers.\n\nIt is not clear why some indicators show larger differences.\nFor `pet`, the differences are due to masking in a specific region\n(Northern Fennoscandia; Figure {numref}`{number} `),\nas discussed before.\nHowever, the same is not true for the daily precipitation indicators\n`r01`,\n`r10`,\nand\n`r20`.\nThese particular indicators show high values and significant spread across Europe in June\n(and presumably in other months)\nmeaning the differences may simply be due to regridding.\nThis also explains why there are non-integer differences:\nsince the C3S Atlas workflow calculates indicators and _then_ regrids,\nthe regridded indicators can have non-integer values.\nSimilar effects can be seen in other indicators when converting between integer and floating-point data types.\n\nWe can extend the conclusion from the [`tx35` case study](./derived_multi-origin-c3s-atlas_consistency_q01),\nnamely that the C3S Atlas dataset can be considered practically reproducible.\nHowever,\nsome of the indicators show non-zero differences,\neither\nin specific locations\n(e.g. `pet`)\nor\ndistributed randomly\n(e.g. `r01`).\nThese differences are caused\neither\nby deliberate decisions to mask certain regions\nor\n(likely)\nby small differences in the data processing.\nThey are small and rare enough to be negligible for most users using the C3S Atlas dataset,\nespecially since they will typically use it through the application.\nFor further analysis,\nit is generally best to manually process the origin dataset,\nif only to remove the amount of steps that may affect the result.\n\nGeospatial plot: Monthly indicators\n\nHistogram: Monthly indicators\n\n::::{tab-set}\n:::{tab-item} Geospatial comparison\n:sync: geo\n derived_multi-origin-c3s-atlas_consistency_q02_fig-geocomp-regional\n:figwidth: 700px\n:name: \"derived_multi-origin-c3s-atlas_consistency_q02_fig-geocomp-regional\"\n\nComparison between C3S Atlas dataset and reproduction for multiple indicators in one month,\nacross Europe,\non the C3S Atlas dataset grid,\nincluding the per-pixel difference.\n```\n:::\n:::{tab-item} Overall comparison\n:sync: hist\n derived_multi-origin-c3s-atlas_consistency_q02_fig-hist-c3s-atlas\n:figwidth: 700px\n:name: \"derived_multi-origin-c3s-atlas_consistency_q02_fig-hist-c3s-atlas\"\n\nComparison between overall distributions of indicator values in the C3S Atlas dataset and its reproduction\non the C3S Atlas grid,\nacross all spatial and temporal dimensions,\nincluding the per-pixel difference.\n```\n:::\n::::", "text_with_prefix": "EQC Quality Assessment: \"Consistency between the C3S Atlas dataset and its origins: Multiple indicators\"\nDataset: multi-origin-c3s-atlas [CDS]\nAspect: consistency_q02 | Category: Derived_Datasets\nSection: Consistency between the C3S Atlas dataset and its origins: Multiple indicators > Analysis and results > 4. Results > Reproducibility: Comparison on C3S Atlas grid\n---\nAfter regridding to the C3S Atlas grid,\nthe indicator values reproduced from the origin dataset\ncan be compared point-by-point to the values retrieved from the C3S Atlas dataset.\nWe first examine some metrics that describe the difference Δ between corresponding pixels:\n\nAs in the [case study for `tx35`](./derived_multi-origin-c3s-atlas_consistency_q01),\nit is clear that\nthe C3S Atlas dataset and its manual reproduction\nare very similar.\nThe median difference, median absolute difference, and median absolute percentage difference\nare all close to 0 for\nalmost all indicators.\n\nThe percentage of pixels with a non-zero difference\n(defined here as |Δ| ≥ ε with ε = 10{sup}`–5` to avoid floating-point errors)\nis a less useful metric here since most of the indicators are\nreal numbers,\nrather than integers like `tx35`.\nThis leaves more room for small differences in the data to appear due to\nsmall changes in the origin dataset and/or the C3S Atlas workflow software,\nwhich for an integer indicator like `tx35` would tend to be rounded off.\nFor most indicators,\nthe relative difference between datasets is ≤0.1% for\nthe vast majority of point-by-point comparisons.\n\nThe distributions of indicator values and differences\n(Figure {numref}`{number} `)\nconfirm that the differences between the two datasets tend to be very small,\nwith distributions centered around 0 and few outliers.\n\nIt is not clear why some indicators show larger differences.\nFor `pet`, the differences are due to masking in a specific region\n(Northern Fennoscandia; Figure {numref}`{number} `),\nas discussed before.\nHowever, the same is not true for the daily precipitation indicators\n`r01`,\n`r10`,\nand\n`r20`.\nThese particular indicators show high values and significant spread across Europe in June\n(and presumably in other months)\nmeaning the differences may simply be due to regridding.\nThis also explains why there are non-integer differences:\nsince the C3S Atlas workflow calculates indicators and _then_ regrids,\nthe regridded indicators can have non-integer values.\nSimilar effects can be seen in other indicators when converting between integer and floating-point data types.\n\nWe can extend the conclusion from the [`tx35` case study](./derived_multi-origin-c3s-atlas_consistency_q01),\nnamely that the C3S Atlas dataset can be considered practically reproducible.\nHowever,\nsome of the indicators show non-zero differences,\neither\nin specific locations\n(e.g. `pet`)\nor\ndistributed randomly\n(e.g. `r01`).\nThese differences are caused\neither\nby deliberate decisions to mask certain regions\nor\n(likely)\nby small differences in the data processing.\nThey are small and rare enough to be negligible for most users using the C3S Atlas dataset,\nespecially since they will typically use it through the application.\nFor further analysis,\nit is generally best to manually process the origin dataset,\nif only to remove the amount of steps that may affect the result.\n\nGeospatial plot: Monthly indicators\n\nHistogram: Monthly indicators\n\n::::{tab-set}\n:::{tab-item} Geospatial comparison\n:sync: geo\n derived_multi-origin-c3s-atlas_consistency_q02_fig-geocomp-regional\n:figwidth: 700px\n:name: \"derived_multi-origin-c3s-atlas_consistency_q02_fig-geocomp-regional\"\n\nComparison between C3S Atlas dataset and reproduction for multiple indicators in one month,\nacross Europe,\non the C3S Atlas dataset grid,\nincluding the per-pixel difference.\n```\n:::\n:::{tab-item} Overall comparison\n:sync: hist\n derived_multi-origin-c3s-atlas_consistency_q02_fig-hist-c3s-atlas\n:figwidth: 700px\n:name: \"derived_multi-origin-c3s-atlas_consistency_q02_fig-hist-c3s-atlas\"\n\nComparison between overall distributions of indicator values in the C3S Atlas dataset and its reproduction\non the C3S Atlas grid,\nacross all spatial and temporal dimensions,\nincluding the per-pixel difference.\n```\n:::\n::::"} {"chunk_id": "derived_multi-origin-c3s-atlas_consistency_q02__596b51fd4bb9", "report_id": "derived_multi-origin-c3s-atlas_consistency_q02", "dataset_id": "multi-origin-c3s-atlas", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q02", "aspect_base": "consistency", "category": "Derived_Datasets", "match_confidence": "exact", "section": "Consistency between the C3S Atlas dataset and its origins: Multiple indicators > ℹ️ If you want to know more > Key resources", "title": "Consistency between the C3S Atlas dataset and its origins: Multiple indicators", "chunk_index": 23, "token_count": 618, "text_raw": "The CDS catalogue entries for the data used were:\n* Gridded dataset underpinning the Copernicus Interactive Climate Atlas: [multi-origin-c3s-atlas](https://doi.org/10.24381/cds.h35hb680)\n * [](./derived_multi-origin-c3s-atlas_consistency_q01)\n * **Consistency between the C3S Atlas dataset and its origins: Multiple indicators**\n * [](./derived_multi-origin-c3s-atlas_consistency_q03)\n* CMIP6 climate projections: [projections-cmip6](https://doi.org/10.24381/cds.c866074c)\n * [Quality assessments for CMIP6](../Climate_Projections/CMIP6/CMIP6.md)\n\nCode libraries used:\n* [earthkit](https://github.com/ecmwf/earthkit)\n * [earthkit-data](https://github.com/ecmwf/earthkit-data)\n * [earthkit-plots](https://github.com/ecmwf/earthkit-plots)\n* [User-tools for the C3S Atlas](https://github.com/ecmwf-projects/c3s-atlas)\n* [xclim](https://xclim.readthedocs.io/en/stable/) climate indicator tools\n\nMore about the Copernicus Interactive Climate Atlas and its IPCC predecessor:\n* [Copernicus Interactive Climate Atlas application](https://atlas.climate.copernicus.eu/)\n* [Gridded data underpinning the Copernicus Interactive Climate Atlas: Description of the datasets and variables](https://confluence.ecmwf.int/display/CKB/Gridded+data+underpinning+the+Copernicus+Interactive+Climate+Atlas%3A+Description+of+the+datasets+and+variables)\n* [The Copernicus Interactive Climate Atlas: a tool to explore regional climate change](https://doi.org/10.21957/ah52ufc369)\n* [Copernicus Interactive Climate Atlas: a new tool to visualise climate variability and change](https://www.ecmwf.int/en/newsletter/179/news/copernicus-interactive-climate-atlas-new-tool-visualise-climate-variability)\n* [Implementation of FAIR principles in the IPCC: the WGI AR6 Atlas repository](https://doi.org/10.1038/s41597-022-01739-y)\n* [Climate Change 2021 – The Physical Science Basis: Atlas](https://doi.org/10.1017/9781009157896.021)", "text_with_prefix": "EQC Quality Assessment: \"Consistency between the C3S Atlas dataset and its origins: Multiple indicators\"\nDataset: multi-origin-c3s-atlas [CDS]\nAspect: consistency_q02 | Category: Derived_Datasets\nSection: Consistency between the C3S Atlas dataset and its origins: Multiple indicators > ℹ️ If you want to know more > Key resources\n---\nThe CDS catalogue entries for the data used were:\n* Gridded dataset underpinning the Copernicus Interactive Climate Atlas: [multi-origin-c3s-atlas](https://doi.org/10.24381/cds.h35hb680)\n * [](./derived_multi-origin-c3s-atlas_consistency_q01)\n * **Consistency between the C3S Atlas dataset and its origins: Multiple indicators**\n * [](./derived_multi-origin-c3s-atlas_consistency_q03)\n* CMIP6 climate projections: [projections-cmip6](https://doi.org/10.24381/cds.c866074c)\n * [Quality assessments for CMIP6](../Climate_Projections/CMIP6/CMIP6.md)\n\nCode libraries used:\n* [earthkit](https://github.com/ecmwf/earthkit)\n * [earthkit-data](https://github.com/ecmwf/earthkit-data)\n * [earthkit-plots](https://github.com/ecmwf/earthkit-plots)\n* [User-tools for the C3S Atlas](https://github.com/ecmwf-projects/c3s-atlas)\n* [xclim](https://xclim.readthedocs.io/en/stable/) climate indicator tools\n\nMore about the Copernicus Interactive Climate Atlas and its IPCC predecessor:\n* [Copernicus Interactive Climate Atlas application](https://atlas.climate.copernicus.eu/)\n* [Gridded data underpinning the Copernicus Interactive Climate Atlas: Description of the datasets and variables](https://confluence.ecmwf.int/display/CKB/Gridded+data+underpinning+the+Copernicus+Interactive+Climate+Atlas%3A+Description+of+the+datasets+and+variables)\n* [The Copernicus Interactive Climate Atlas: a tool to explore regional climate change](https://doi.org/10.21957/ah52ufc369)\n* [Copernicus Interactive Climate Atlas: a new tool to visualise climate variability and change](https://www.ecmwf.int/en/newsletter/179/news/copernicus-interactive-climate-atlas-new-tool-visualise-climate-variability)\n* [Implementation of FAIR principles in the IPCC: the WGI AR6 Atlas repository](https://doi.org/10.1038/s41597-022-01739-y)\n* [Climate Change 2021 – The Physical Science Basis: Atlas](https://doi.org/10.1017/9781009157896.021)"} {"chunk_id": "derived_multi-origin-c3s-atlas_consistency_q02__cc0c0f0a560b", "report_id": "derived_multi-origin-c3s-atlas_consistency_q02", "dataset_id": "multi-origin-c3s-atlas", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q02", "aspect_base": "consistency", "category": "Derived_Datasets", "match_confidence": "exact", "section": "Consistency between the C3S Atlas dataset and its origins: Multiple indicators > ℹ️ If you want to know more > References", "title": "Consistency between the C3S Atlas dataset and its origins: Multiple indicators", "chunk_index": 24, "token_count": 439, "text_raw": "[[Avila+15](https://doi.org/10.1016/j.wace.2015.06.003)] F. B. Avila et al., ‘Systematic investigation of gridding-related scaling effects on annual statistics of daily temperature and precipitation maxima: A case study for south-east Australia’, Weather and Climate Extremes, vol. 9, pp. 6–16, Aug. 2015, doi: 10.1016/j.wace.2015.06.003.\n\n[[C3S Atlas dataset](https://doi.org/10.24381/cds.h35hb680)] Copernicus Climate Change Service, ‘Gridded dataset underpinning the Copernicus Interactive Climate Atlas’. Copernicus Climate Change Service (C3S) Climate Data Store (CDS), Jun. 17, 2024. doi: 10.24381/cds.h35hb680.\n\n[[CMIP6 dataset](https://doi.org/10.24381/cds.c866074c)] Copernicus Climate Change Service, ‘CMIP6 climate projections’. Copernicus Climate Change Service (C3S) Climate Data Store (CDS), Mar. 23, 2021. doi: 10.24381/cds.c866074c.\n\n[[Gutiérrez+24](https://doi.org/10.21957/ah52ufc369)] J. M. Gutiérrez et al., ‘The Copernicus Interactive Climate Atlas: a tool to explore regional climate change’, ECMWF Newsletter, vol. 181, pp. 38–45, Oct. 2024, doi: 10.21957/ah52ufc369.", "text_with_prefix": "EQC Quality Assessment: \"Consistency between the C3S Atlas dataset and its origins: Multiple indicators\"\nDataset: multi-origin-c3s-atlas [CDS]\nAspect: consistency_q02 | Category: Derived_Datasets\nSection: Consistency between the C3S Atlas dataset and its origins: Multiple indicators > ℹ️ If you want to know more > References\n---\n[[Avila+15](https://doi.org/10.1016/j.wace.2015.06.003)] F. B. Avila et al., ‘Systematic investigation of gridding-related scaling effects on annual statistics of daily temperature and precipitation maxima: A case study for south-east Australia’, Weather and Climate Extremes, vol. 9, pp. 6–16, Aug. 2015, doi: 10.1016/j.wace.2015.06.003.\n\n[[C3S Atlas dataset](https://doi.org/10.24381/cds.h35hb680)] Copernicus Climate Change Service, ‘Gridded dataset underpinning the Copernicus Interactive Climate Atlas’. Copernicus Climate Change Service (C3S) Climate Data Store (CDS), Jun. 17, 2024. doi: 10.24381/cds.h35hb680.\n\n[[CMIP6 dataset](https://doi.org/10.24381/cds.c866074c)] Copernicus Climate Change Service, ‘CMIP6 climate projections’. Copernicus Climate Change Service (C3S) Climate Data Store (CDS), Mar. 23, 2021. doi: 10.24381/cds.c866074c.\n\n[[Gutiérrez+24](https://doi.org/10.21957/ah52ufc369)] J. M. Gutiérrez et al., ‘The Copernicus Interactive Climate Atlas: a tool to explore regional climate change’, ECMWF Newsletter, vol. 181, pp. 38–45, Oct. 2024, doi: 10.21957/ah52ufc369."} {"chunk_id": "derived_multi-origin-c3s-atlas_consistency_q03__91c634bc3208", "report_id": "derived_multi-origin-c3s-atlas_consistency_q03", "dataset_id": "multi-origin-c3s-atlas", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q03", "aspect_base": "consistency", "category": "Derived_Datasets", "match_confidence": "exact", "section": "Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets", "title": "Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets", "chunk_index": 0, "token_count": 105, "text_raw": "Production date: 2025-10-14.\n\nDataset version: 2.0.\n\nProduced by: Olivier Burggraaff, Nicole Reynolds (National Physical Laboratory).", "text_with_prefix": "EQC Quality Assessment: \"Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets\"\nDataset: multi-origin-c3s-atlas [CDS]\nAspect: consistency_q03 | Category: Derived_Datasets\nSection: Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets\n---\nProduction date: 2025-10-14.\n\nDataset version: 2.0.\n\nProduced by: Olivier Burggraaff, Nicole Reynolds (National Physical Laboratory)."} {"chunk_id": "derived_multi-origin-c3s-atlas_consistency_q03__935d1253b082", "report_id": "derived_multi-origin-c3s-atlas_consistency_q03", "dataset_id": "multi-origin-c3s-atlas", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q03", "aspect_base": "consistency", "category": "Derived_Datasets", "match_confidence": "exact", "section": "Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets > Quality assessment question", "title": "Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets", "chunk_index": 1, "token_count": 884, "text_raw": "* **Are the climate indicators in the dataset underpinning the Copernicus Interactive Climate Atlas consistent with their origin datasets?**\n* **Can the dataset underpinning the Copernicus Interactive Climate Atlas be reproduced from its origin datasets?**\n\nThe [_Copernicus Interactive Climate Atlas_](https://atlas.climate.copernicus.eu/atlas), or _C3S Atlas_ for short, is a C3S web application providing an easy-to-access tool for exploring climate projections, reanalyses, and observational data [[Gutiérrez+24](https://doi.org/10.21957/ah52ufc369)].\nVersion 2.0 of the application allows the user to interact with 12 datasets:\n\n| Type | Dataset |\n|--------------------|---------------|\n| Climate Projection | CMIP6 |\n| Climate Projection | CMIP5 |\n| Climate Projection | CORDEX-CORE |\n| Climate Projection | CORDEX-EUR-11 |\n| Reanalysis | ERA5 |\n| Reanalysis | ERA5-Land |\n| Reanalysis | ORAS5 |\n| Reanalysis | CERRA |\n| Observations | E-OBS |\n| Observations | BERKEARTH |\n| Observations | CPC |\n| Observations | SST-CCI |\n\nThese datasets are provided through an intermediary dataset, the [_Gridded dataset underpinning the Copernicus Interactive Climate Atlas_](https://doi.org/10.24381/cds.h35hb680) or _C3S Atlas dataset_ for short [[C3S Atlas dataset](https://doi.org/10.24381/cds.h35hb680)].\nCompared to their origins, the versions of the climate datasets within the C3S Atlas dataset have been processed following the workflow in Figure {numref}`{number} `.\n\nattachment:c3s_atlas_dataset_workflow.png\n---\nheight: 360px\nname: multi-origin-c3s-atlas_consistency_q03_workflow-fig\n---\nSchematic representation of the workflow for the production of the C3S Atlas dataset from its origin datasets, from the [User-tools for the C3S Atlas](https://ecmwf-projects.github.io/c3s-atlas/chapter01.html).\n```\n\nBecause a wide range of users interact with climate data through the C3S Atlas application, it is crucial that the underpinning dataset represent its origins correctly.\nIn other words, the C3S Atlas dataset must be consistent with and reproducible from its origins.\nHere, we assess this consistency and reproducibility by comparing climate indicators retrieved from the C3S Atlas dataset with their equivalents calculated from the origin dataset, mirroring the workflow from Figure {numref}`{number} `.\nWhile a full analysis and reproduction of every record within the C3S Atlas dataset is outside the scope of quality assessment\n(and would require high-performance computing infrastructure),\na case study with a narrower scope probes these quality attributes of the dataset\nand can be a jumping-off point for further analysis by the reader.\n\nThis notebook is part of a series:\n| Notebook | Contents |\n|---|---|\n| [](./derived_multi-origin-c3s-atlas_consistency_q01) | Comparison between C3S Atlas dataset and one origin dataset (CMIP6) for one indicator (`tx35`), including detailed setup. |\n| [](./derived_multi-origin-c3s-atlas_consistency_q02) | Comparison between C3S Atlas dataset and one origin dataset (CMIP6) for multiple indicators. |\n| **Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets** | Comparison between C3S Atlas dataset and multiple origin datasets for one indicator. |", "text_with_prefix": "EQC Quality Assessment: \"Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets\"\nDataset: multi-origin-c3s-atlas [CDS]\nAspect: consistency_q03 | Category: Derived_Datasets\nSection: Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets > Quality assessment question\n---\n* **Are the climate indicators in the dataset underpinning the Copernicus Interactive Climate Atlas consistent with their origin datasets?**\n* **Can the dataset underpinning the Copernicus Interactive Climate Atlas be reproduced from its origin datasets?**\n\nThe [_Copernicus Interactive Climate Atlas_](https://atlas.climate.copernicus.eu/atlas), or _C3S Atlas_ for short, is a C3S web application providing an easy-to-access tool for exploring climate projections, reanalyses, and observational data [[Gutiérrez+24](https://doi.org/10.21957/ah52ufc369)].\nVersion 2.0 of the application allows the user to interact with 12 datasets:\n\n| Type | Dataset |\n|--------------------|---------------|\n| Climate Projection | CMIP6 |\n| Climate Projection | CMIP5 |\n| Climate Projection | CORDEX-CORE |\n| Climate Projection | CORDEX-EUR-11 |\n| Reanalysis | ERA5 |\n| Reanalysis | ERA5-Land |\n| Reanalysis | ORAS5 |\n| Reanalysis | CERRA |\n| Observations | E-OBS |\n| Observations | BERKEARTH |\n| Observations | CPC |\n| Observations | SST-CCI |\n\nThese datasets are provided through an intermediary dataset, the [_Gridded dataset underpinning the Copernicus Interactive Climate Atlas_](https://doi.org/10.24381/cds.h35hb680) or _C3S Atlas dataset_ for short [[C3S Atlas dataset](https://doi.org/10.24381/cds.h35hb680)].\nCompared to their origins, the versions of the climate datasets within the C3S Atlas dataset have been processed following the workflow in Figure {numref}`{number} `.\n\nattachment:c3s_atlas_dataset_workflow.png\n---\nheight: 360px\nname: multi-origin-c3s-atlas_consistency_q03_workflow-fig\n---\nSchematic representation of the workflow for the production of the C3S Atlas dataset from its origin datasets, from the [User-tools for the C3S Atlas](https://ecmwf-projects.github.io/c3s-atlas/chapter01.html).\n```\n\nBecause a wide range of users interact with climate data through the C3S Atlas application, it is crucial that the underpinning dataset represent its origins correctly.\nIn other words, the C3S Atlas dataset must be consistent with and reproducible from its origins.\nHere, we assess this consistency and reproducibility by comparing climate indicators retrieved from the C3S Atlas dataset with their equivalents calculated from the origin dataset, mirroring the workflow from Figure {numref}`{number} `.\nWhile a full analysis and reproduction of every record within the C3S Atlas dataset is outside the scope of quality assessment\n(and would require high-performance computing infrastructure),\na case study with a narrower scope probes these quality attributes of the dataset\nand can be a jumping-off point for further analysis by the reader.\n\nThis notebook is part of a series:\n| Notebook | Contents |\n|---|---|\n| [](./derived_multi-origin-c3s-atlas_consistency_q01) | Comparison between C3S Atlas dataset and one origin dataset (CMIP6) for one indicator (`tx35`), including detailed setup. |\n| [](./derived_multi-origin-c3s-atlas_consistency_q02) | Comparison between C3S Atlas dataset and one origin dataset (CMIP6) for multiple indicators. |\n| **Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets** | Comparison between C3S Atlas dataset and multiple origin datasets for one indicator. |"} {"chunk_id": "derived_multi-origin-c3s-atlas_consistency_q03__ee325aab05d6", "report_id": "derived_multi-origin-c3s-atlas_consistency_q03", "dataset_id": "multi-origin-c3s-atlas", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q03", "aspect_base": "consistency", "category": "Derived_Datasets", "match_confidence": "exact", "section": "Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets > Quality assessment statement", "title": "Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets", "chunk_index": 2, "token_count": 334, "text_raw": "These are the key outcomes of this assessment\n\n* Climate indicators (here 3 monthly indicators) provided by the C3S Atlas dataset are highly consistent with values calculated from its origin datasets (here 10 climate projection, observation, and reanalysis datasets).\n\n* There are some differences in coverage between the C3S Atlas dataset and its origins due to differences in the version used, e.g. E-OBS temperature indicators at the edges of Europe, but these do not affect the vast majority of use cases for the C3S Atlas.\n\n* Differences between the C3S Atlas dataset and a manual reproduction are rare and generally negligible (median absolute difference of 0 for `tx35` and `r01`, ≤0.0003 °C for SST). Where differences occur, they can be explained by dataset versions or details of the workflow implementation.\n\n* The C3S Atlas is traceable and reproducible, and can be confidently used to view, analyse, and download climate data.\n\n* For specific use cases, like scientific papers, it is recommended to manually process the origin dataset instead of using the C3S Atlas. This is not necessary for other use cases, such as climate risk assessments and climate reports, in which case the C3S Atlas can be used as is.\n```", "text_with_prefix": "EQC Quality Assessment: \"Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets\"\nDataset: multi-origin-c3s-atlas [CDS]\nAspect: consistency_q03 | Category: Derived_Datasets\nSection: Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets > Quality assessment statement\n---\nThese are the key outcomes of this assessment\n\n* Climate indicators (here 3 monthly indicators) provided by the C3S Atlas dataset are highly consistent with values calculated from its origin datasets (here 10 climate projection, observation, and reanalysis datasets).\n\n* There are some differences in coverage between the C3S Atlas dataset and its origins due to differences in the version used, e.g. E-OBS temperature indicators at the edges of Europe, but these do not affect the vast majority of use cases for the C3S Atlas.\n\n* Differences between the C3S Atlas dataset and a manual reproduction are rare and generally negligible (median absolute difference of 0 for `tx35` and `r01`, ≤0.0003 °C for SST). Where differences occur, they can be explained by dataset versions or details of the workflow implementation.\n\n* The C3S Atlas is traceable and reproducible, and can be confidently used to view, analyse, and download climate data.\n\n* For specific use cases, like scientific papers, it is recommended to manually process the origin dataset instead of using the C3S Atlas. This is not necessary for other use cases, such as climate risk assessments and climate reports, in which case the C3S Atlas can be used as is.\n```"} {"chunk_id": "derived_multi-origin-c3s-atlas_consistency_q03__55a92ca5b9da", "report_id": "derived_multi-origin-c3s-atlas_consistency_q03", "dataset_id": "multi-origin-c3s-atlas", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q03", "aspect_base": "consistency", "category": "Derived_Datasets", "match_confidence": "exact", "section": "Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets > Methodology", "title": "Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets", "chunk_index": 3, "token_count": 641, "text_raw": "This quality assessment tests the consistency between climate indicators retrieved from the [_Gridded dataset underpinning the Copernicus Interactive Climate Atlas_](https://doi.org/10.24381/cds.h35hb680) [[C3S Atlas dataset](https://doi.org/10.24381/cds.h35hb680)] and their equivalents calculated from the origin datasets,\nas well as the reproducibility of said dataset.\n\nThis notebook probes the consistency between the C3S Atlas dataset and multiple origin datasets at the same time.\nDue to differences in scope (e.g. atmosphere / land / sea), not every indicator is available in every origin dataset or its C3S Atlas derivative.\nFurthermore, some origin datasets are historical while others are future projections.\nFor this reason, we will examine the following indicators in the following origin datasets:\n\n**Monthly count of days with maximum near-surface (2-metre) air temperature above 35 °C (`tx35`)**\n| Type | Dataset |\n|--------------------|--------------|\n| Climate Projection | CMIP6 |\n| Climate Projection | CMIP5 |\n| Climate Projection | CORDEX-EUR-11|\n| Reanalysis | ERA5 |\n| Reanalysis | ERA5-Land |\n| Observations | E-OBS |\n| Observations | BERKEARTH |\n\nNote that CORDEX-CORE has been left out of this assessment because its mosaicking workflow is out of scope.\nCERRA has been left out because the C3S User-tools package is currently not fully compatible with this dataset.\n\n**Monthly mean temperature of sea water near the surface (`sst`)**\n| Type | Dataset |\n|--------------------|--------------|\n| Reanalysis | ORAS5 |\n| Observations | SST-CCI |\n\n**Monthly count of days with daily accumulated precipitation of liquid water equivalent from all phases above 1 mm (`r01`)**\n| Type | Dataset |\n|--------------------|--------------|\n| Observations | CPC |\n\nThe analysis and results are organised in the following steps, which are detailed in the sections below:\n\n**[](section-codesetup)**\n * Install User-tools for the C3S Atlas.\n * Import all required libraries.\n * Define indicators.\n * Define helper functions.\n\n**[](section-indicators)**\n * Download data from the origin datasets.\n * Homogenise data.\n * Calculate indicators.\n * Regrid the origin data to the C3S Atlas grid.\n * Download corresponding data from the C3S Atlas dataset.\n\n**[](section-results)**\n * Consistency: Compare the C3S Atlas and reproduced datasets on native grids.\n * Reproducibility: Compare the C3S Atlas and reproduced datasets on the C3S Atlas grid.", "text_with_prefix": "EQC Quality Assessment: \"Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets\"\nDataset: multi-origin-c3s-atlas [CDS]\nAspect: consistency_q03 | Category: Derived_Datasets\nSection: Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets > Methodology\n---\nThis quality assessment tests the consistency between climate indicators retrieved from the [_Gridded dataset underpinning the Copernicus Interactive Climate Atlas_](https://doi.org/10.24381/cds.h35hb680) [[C3S Atlas dataset](https://doi.org/10.24381/cds.h35hb680)] and their equivalents calculated from the origin datasets,\nas well as the reproducibility of said dataset.\n\nThis notebook probes the consistency between the C3S Atlas dataset and multiple origin datasets at the same time.\nDue to differences in scope (e.g. atmosphere / land / sea), not every indicator is available in every origin dataset or its C3S Atlas derivative.\nFurthermore, some origin datasets are historical while others are future projections.\nFor this reason, we will examine the following indicators in the following origin datasets:\n\n**Monthly count of days with maximum near-surface (2-metre) air temperature above 35 °C (`tx35`)**\n| Type | Dataset |\n|--------------------|--------------|\n| Climate Projection | CMIP6 |\n| Climate Projection | CMIP5 |\n| Climate Projection | CORDEX-EUR-11|\n| Reanalysis | ERA5 |\n| Reanalysis | ERA5-Land |\n| Observations | E-OBS |\n| Observations | BERKEARTH |\n\nNote that CORDEX-CORE has been left out of this assessment because its mosaicking workflow is out of scope.\nCERRA has been left out because the C3S User-tools package is currently not fully compatible with this dataset.\n\n**Monthly mean temperature of sea water near the surface (`sst`)**\n| Type | Dataset |\n|--------------------|--------------|\n| Reanalysis | ORAS5 |\n| Observations | SST-CCI |\n\n**Monthly count of days with daily accumulated precipitation of liquid water equivalent from all phases above 1 mm (`r01`)**\n| Type | Dataset |\n|--------------------|--------------|\n| Observations | CPC |\n\nThe analysis and results are organised in the following steps, which are detailed in the sections below:\n\n**[](section-codesetup)**\n * Install User-tools for the C3S Atlas.\n * Import all required libraries.\n * Define indicators.\n * Define helper functions.\n\n**[](section-indicators)**\n * Download data from the origin datasets.\n * Homogenise data.\n * Calculate indicators.\n * Regrid the origin data to the C3S Atlas grid.\n * Download corresponding data from the C3S Atlas dataset.\n\n**[](section-results)**\n * Consistency: Compare the C3S Atlas and reproduced datasets on native grids.\n * Reproducibility: Compare the C3S Atlas and reproduced datasets on the C3S Atlas grid."} {"chunk_id": "derived_multi-origin-c3s-atlas_consistency_q03__4cbfff22f59a", "report_id": "derived_multi-origin-c3s-atlas_consistency_q03", "dataset_id": "multi-origin-c3s-atlas", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q03", "aspect_base": "consistency", "category": "Derived_Datasets", "match_confidence": "exact", "section": "Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets > Analysis and results > 1. Code setup", "title": "Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets", "chunk_index": 4, "token_count": 185, "text_raw": "This notebook uses [earthkit](https://github.com/ecmwf/earthkit) for\ndownloading ([earthkit-data](https://github.com/ecmwf/earthkit-data))\nand visualising ([earthkit-plots](https://github.com/ecmwf/earthkit-plots)) data.\nBecause earthkit is in active development, some functionality may change after this notebook is published.\nIf any part of the code stops functioning, please raise an issue on our GitHub repository so it can be fixed.\n```", "text_with_prefix": "EQC Quality Assessment: \"Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets\"\nDataset: multi-origin-c3s-atlas [CDS]\nAspect: consistency_q03 | Category: Derived_Datasets\nSection: Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets > Analysis and results > 1. Code setup\n---\nThis notebook uses [earthkit](https://github.com/ecmwf/earthkit) for\ndownloading ([earthkit-data](https://github.com/ecmwf/earthkit-data))\nand visualising ([earthkit-plots](https://github.com/ecmwf/earthkit-plots)) data.\nBecause earthkit is in active development, some functionality may change after this notebook is published.\nIf any part of the code stops functioning, please raise an issue on our GitHub repository so it can be fixed.\n```"} {"chunk_id": "derived_multi-origin-c3s-atlas_consistency_q03__846b80c9a95d", "report_id": "derived_multi-origin-c3s-atlas_consistency_q03", "dataset_id": "multi-origin-c3s-atlas", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q03", "aspect_base": "consistency", "category": "Derived_Datasets", "match_confidence": "exact", "section": "Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets > Analysis and results > 1. Code setup > Install the User-tools for the C3S Atlas", "title": "Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets", "chunk_index": 5, "token_count": 184, "text_raw": "This notebook uses the [User-tools for the C3S Atlas](https://github.com/ecmwf-projects/c3s-atlas), which can be installed from GitHub using `pip`.\nFor convenience, the following cell can do this from within the notebook.\nFurther details and alternative options for installing this library are available in its [documentation](https://github.com/ecmwf-projects/c3s-atlas?tab=readme-ov-file#requirements).", "text_with_prefix": "EQC Quality Assessment: \"Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets\"\nDataset: multi-origin-c3s-atlas [CDS]\nAspect: consistency_q03 | Category: Derived_Datasets\nSection: Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets > Analysis and results > 1. Code setup > Install the User-tools for the C3S Atlas\n---\nThis notebook uses the [User-tools for the C3S Atlas](https://github.com/ecmwf-projects/c3s-atlas), which can be installed from GitHub using `pip`.\nFor convenience, the following cell can do this from within the notebook.\nFurther details and alternative options for installing this library are available in its [documentation](https://github.com/ecmwf-projects/c3s-atlas?tab=readme-ov-file#requirements)."} {"chunk_id": "derived_multi-origin-c3s-atlas_consistency_q03__a88b66f280b2", "report_id": "derived_multi-origin-c3s-atlas_consistency_q03", "dataset_id": "multi-origin-c3s-atlas", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q03", "aspect_base": "consistency", "category": "Derived_Datasets", "match_confidence": "exact", "section": "Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets > Analysis and results > 1. Code setup > Import required libraries", "title": "Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets", "chunk_index": 6, "token_count": 128, "text_raw": "In this section, we import all the relevant packages needed for running the notebook.\n\nInput / Output\nGeneral data handling\nData pre-processing\nClimate indicators\nVisualisation\nVisualisation in Jupyter book -- automatically ignored otherwise", "text_with_prefix": "EQC Quality Assessment: \"Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets\"\nDataset: multi-origin-c3s-atlas [CDS]\nAspect: consistency_q03 | Category: Derived_Datasets\nSection: Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets > Analysis and results > 1. Code setup > Import required libraries\n---\nIn this section, we import all the relevant packages needed for running the notebook.\n\nInput / Output\nGeneral data handling\nData pre-processing\nClimate indicators\nVisualisation\nVisualisation in Jupyter book -- automatically ignored otherwise"} {"chunk_id": "derived_multi-origin-c3s-atlas_consistency_q03__ace01afabcf7", "report_id": "derived_multi-origin-c3s-atlas_consistency_q03", "dataset_id": "multi-origin-c3s-atlas", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q03", "aspect_base": "consistency", "category": "Derived_Datasets", "match_confidence": "exact", "section": "Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets > Analysis and results > 1. Code setup > Define indicators", "title": "Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets", "chunk_index": 7, "token_count": 293, "text_raw": "This section defines functions and variables for calculating and using the climate indicators.\n\nThe following cell includes a helper function (in the form of a decorator) to correctly propagate NaNs, which is important for datasets using land/sea masks:\n\nCalculate indicator as normal\nIf propagation is not desired (e.g. you know there are no NaNs), simply return the result\nThis preserves integer data types\nInfer input (var-iable), output (ind-icator) keys\nResample data to frequency of indicator, generate mask\nApply mask and return result\n\nThe following cell contains functions for calculating the indicators described in the introduction:\n\nThe following cell defines [earthkit-plots styles](https://earthkit-plots.readthedocs.io/en/latest/examples/examples/introduction/05-styles.html) for the indicators.\nThese styles define the colour maps and colour bar ranges for each quantity.\n\nStyles for indicators\ntx35 indicator\nsst indicator\nr01 indicator\nIndividual styles\nSet up like this so they can still be edited individually\nApply general settings", "text_with_prefix": "EQC Quality Assessment: \"Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets\"\nDataset: multi-origin-c3s-atlas [CDS]\nAspect: consistency_q03 | Category: Derived_Datasets\nSection: Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets > Analysis and results > 1. Code setup > Define indicators\n---\nThis section defines functions and variables for calculating and using the climate indicators.\n\nThe following cell includes a helper function (in the form of a decorator) to correctly propagate NaNs, which is important for datasets using land/sea masks:\n\nCalculate indicator as normal\nIf propagation is not desired (e.g. you know there are no NaNs), simply return the result\nThis preserves integer data types\nInfer input (var-iable), output (ind-icator) keys\nResample data to frequency of indicator, generate mask\nApply mask and return result\n\nThe following cell contains functions for calculating the indicators described in the introduction:\n\nThe following cell defines [earthkit-plots styles](https://earthkit-plots.readthedocs.io/en/latest/examples/examples/introduction/05-styles.html) for the indicators.\nThese styles define the colour maps and colour bar ranges for each quantity.\n\nStyles for indicators\ntx35 indicator\nsst indicator\nr01 indicator\nIndividual styles\nSet up like this so they can still be edited individually\nApply general settings"} {"chunk_id": "derived_multi-origin-c3s-atlas_consistency_q03__32a2091a5ae8", "report_id": "derived_multi-origin-c3s-atlas_consistency_q03", "dataset_id": "multi-origin-c3s-atlas", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q03", "aspect_base": "consistency", "category": "Derived_Datasets", "match_confidence": "exact", "section": "Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets > Analysis and results > 1. Code setup > Helper functions", "title": "Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets", "chunk_index": 8, "token_count": 948, "text_raw": "This section defines some functions and variables used in the following analysis, allowing code cells in later sections to be shorter and ensuring consistency.\n\n##### Data downloading & (pre-)processing\n\nThe following functions help with downloading data from datasets with limits, by generating multiple CDS requests with similar parameters:\n\nSplit requests into smaller parts\nSST-CCI: Generate monthly batches of requests\nGenerate requests\n\nThe following functions aid in sub-selecting data, e.g. selecting one model from the ensemble included in the C3S Atlas dataset or selecting data for a specific time frame:\n\nSelect (multiple) years in a dataset\nC3S Atlas dataset: Individual model selection\nEnsure the model ID is provided in the right format\nFind the corresponding model ID in the list of models\nThis cannot use .sel because the coordinate is not indexed\nFind the corresponding data and return those\nC3S Atlas dataset: Common pre-processing\n\nThe following functions handle [data chunking in dask](https://docs.xarray.dev/en/latest/user-guide/dask.html) for computational efficiency:\n\nRechunking of data to speed up dask calculations\nFind coordinates (usually lat/lon, sometimes x/y)\nAssign new chunk sizes\n\nThe following functions handle the homogenisation of origin data to a consistent format using the [User-tools for the C3S Atlas](https://ecmwf-projects.github.io/c3s-atlas/intro.html):\n\nHomogenisation of origin dataset\n\nThe following functions handle regridding data based on ESMF as implemented in the [User-tools for the C3S Atlas](https://ecmwf-projects.github.io/c3s-atlas/intro.html).\nThis step is explained in more detail in the relevant section below.\n\nRegridding from native grid to C3S Atlas grid\n\n##### Statistics\nThe following functions calculate the difference (absolute / relative) between datasets, handling metadata etc.:\n\nDifference between datasets\nSelect and calculate\nReplace 0/0 with 0\n\nThe following functions calculate and display metrics for the difference between two datasets, e.g. mean and median deviation:\n\nCalculate differences\nCalculate relative difference if desired\nCalculate aggregate statistics\n(n.b. in dask, this only queues them up, doesn't actually compute)\nCalculate correlation coefficients\nPerform the queued-up calculations\nCombine statistics into one dataframe\nSetup: origins\nCalculate statistics per dataset, combine result, return\n\n##### Visualisation\nThe following cells contain functions for plotting results, starting with some base helper functions (e.g. displaying in Jupyter Notebook or Jupyter Book style, adding textboxes with consistent formatting, etc.):\n\nVisualisation: Helper functions, general\nGet the plt.Axes for each ekp.Subplot\nSet up location\nAdd the text\n\nThe following functions are also base helper functions, but specific to geospatial plots:\n\nVisualisation: Helper functions for geospatial plots\nCreate subplots\n\nThe following cell contains functions for geospatial comparisons between datasets on their native grids or on a common grid (the latter also showing the per-pixel difference):\n\nSetup: origins\nCreate figure\nPlot individual datasets\nDecorate: Text\nColour bar at the bottom\nTitles on top\nDecorate figure\nVisualisation: Plot indicators geospatially\nSetup: origins\nCreate figure\nPlot individual datasets\nPlot difference\nDecorate: Text\nColour bar at the bottom\nTitles on top\nDecorate figure\n\nThe following cell contains functions for histogram comparisons between datasets on their native grids or on a common grid (the latter also showing the per-pixel difference):\n\nVisualisation: Plot data in histograms\nSetup: origins\nCreate figure\nPlot histograms of data\nLoop over rows / origins\nFlatten data\nCreate histogram\nIdentify panel\nTitles on top\nVisualisation: Plot data + difference in histograms\nSetup: origins\nCreate figure\nSetup x/y share -- cannot be done in plt.subplots because of difference panel not sharing these\n_sharexy(axs[:, -1]) # Leave out for now -- ALlow differences between origin datasets\nPlot histograms of data\nLoop over rows / origins\nLoop over columns / data\nFlatten data\nCreate histogram\nPlot difference\nIdentify panel\nTitles on top\n\n(section-indicators)=", "text_with_prefix": "EQC Quality Assessment: \"Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets\"\nDataset: multi-origin-c3s-atlas [CDS]\nAspect: consistency_q03 | Category: Derived_Datasets\nSection: Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets > Analysis and results > 1. Code setup > Helper functions\n---\nThis section defines some functions and variables used in the following analysis, allowing code cells in later sections to be shorter and ensuring consistency.\n\n##### Data downloading & (pre-)processing\n\nThe following functions help with downloading data from datasets with limits, by generating multiple CDS requests with similar parameters:\n\nSplit requests into smaller parts\nSST-CCI: Generate monthly batches of requests\nGenerate requests\n\nThe following functions aid in sub-selecting data, e.g. selecting one model from the ensemble included in the C3S Atlas dataset or selecting data for a specific time frame:\n\nSelect (multiple) years in a dataset\nC3S Atlas dataset: Individual model selection\nEnsure the model ID is provided in the right format\nFind the corresponding model ID in the list of models\nThis cannot use .sel because the coordinate is not indexed\nFind the corresponding data and return those\nC3S Atlas dataset: Common pre-processing\n\nThe following functions handle [data chunking in dask](https://docs.xarray.dev/en/latest/user-guide/dask.html) for computational efficiency:\n\nRechunking of data to speed up dask calculations\nFind coordinates (usually lat/lon, sometimes x/y)\nAssign new chunk sizes\n\nThe following functions handle the homogenisation of origin data to a consistent format using the [User-tools for the C3S Atlas](https://ecmwf-projects.github.io/c3s-atlas/intro.html):\n\nHomogenisation of origin dataset\n\nThe following functions handle regridding data based on ESMF as implemented in the [User-tools for the C3S Atlas](https://ecmwf-projects.github.io/c3s-atlas/intro.html).\nThis step is explained in more detail in the relevant section below.\n\nRegridding from native grid to C3S Atlas grid\n\n##### Statistics\nThe following functions calculate the difference (absolute / relative) between datasets, handling metadata etc.:\n\nDifference between datasets\nSelect and calculate\nReplace 0/0 with 0\n\nThe following functions calculate and display metrics for the difference between two datasets, e.g. mean and median deviation:\n\nCalculate differences\nCalculate relative difference if desired\nCalculate aggregate statistics\n(n.b. in dask, this only queues them up, doesn't actually compute)\nCalculate correlation coefficients\nPerform the queued-up calculations\nCombine statistics into one dataframe\nSetup: origins\nCalculate statistics per dataset, combine result, return\n\n##### Visualisation\nThe following cells contain functions for plotting results, starting with some base helper functions (e.g. displaying in Jupyter Notebook or Jupyter Book style, adding textboxes with consistent formatting, etc.):\n\nVisualisation: Helper functions, general\nGet the plt.Axes for each ekp.Subplot\nSet up location\nAdd the text\n\nThe following functions are also base helper functions, but specific to geospatial plots:\n\nVisualisation: Helper functions for geospatial plots\nCreate subplots\n\nThe following cell contains functions for geospatial comparisons between datasets on their native grids or on a common grid (the latter also showing the per-pixel difference):\n\nSetup: origins\nCreate figure\nPlot individual datasets\nDecorate: Text\nColour bar at the bottom\nTitles on top\nDecorate figure\nVisualisation: Plot indicators geospatially\nSetup: origins\nCreate figure\nPlot individual datasets\nPlot difference\nDecorate: Text\nColour bar at the bottom\nTitles on top\nDecorate figure\n\nThe following cell contains functions for histogram comparisons between datasets on their native grids or on a common grid (the latter also showing the per-pixel difference):\n\nVisualisation: Plot data in histograms\nSetup: origins\nCreate figure\nPlot histograms of data\nLoop over rows / origins\nFlatten data\nCreate histogram\nIdentify panel\nTitles on top\nVisualisation: Plot data + difference in histograms\nSetup: origins\nCreate figure\nSetup x/y share -- cannot be done in plt.subplots because of difference panel not sharing these\n_sharexy(axs[:, -1]) # Leave out for now -- ALlow differences between origin datasets\nPlot histograms of data\nLoop over rows / origins\nLoop over columns / data\nFlatten data\nCreate histogram\nPlot difference\nIdentify panel\nTitles on top\n\n(section-indicators)="} {"chunk_id": "derived_multi-origin-c3s-atlas_consistency_q03__f4d47fa1769c", "report_id": "derived_multi-origin-c3s-atlas_consistency_q03", "dataset_id": "multi-origin-c3s-atlas", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q03", "aspect_base": "consistency", "category": "Derived_Datasets", "match_confidence": "exact", "section": "Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets > Analysis and results > 2. Calculate and retrieve indicators", "title": "Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets", "chunk_index": 9, "token_count": 802, "text_raw": "In the previous two notebooks in this assessment,\nthe origin data were downloaded,\npre-processed,\nused to calculate the relevant indicator(s),\nand regridded;\nafter which the [Atlas dataset](https://doi.org/10.24381/cds.h35hb680) was downloaded.\nThis notebook follows the same structure but for each origin in turn,\nfor clarity and to preserve memory when loading multiple datasets at the same time.\nAs such,\nthe individual steps are described in less detail,\nbecause this information is available in the previous notebooks.\n\nIf you are only interested in specific origin datasets,\nor want to limit your bandwidth or memory usage,\nyou can choose to only run specific subsections below.\n\nThis assessment examines the dataset in one year (2080),\nfor one ensemble member\nand\nfor one climate scenario\n– which is not how climate projection data are normally used.\nGood practice in climate science is to look at multi-year statistics and trends,\nacross multiple ensemble members.\nHowever, since the purpose of this assessment is to assess the consistency and reproducibility of the post-processing performed to produce the C3S Atlas dataset,\nit is valid to use a subset of the data here.\nThe specific subset used can be easily tweaked by changing the `EXPERIMENT`, `MODEL`, and `YEARS` variables in the relevant sections.\n\nThis notebook uses [earthkit-data](https://github.com/ecmwf/earthkit-data) to download files from the CDS.\nIf you intend to run this notebook multiple times, it is highly recommended that you [enable caching](https://earthkit-data.readthedocs.io/en/latest/concepts/caching.html) to prevent having to download the same files multiple times.\nIf you prefer not to use earthkit, the following requests can also be used with the [cdsapi module](https://cds.climate.copernicus.eu/how-to-api#linux-use-client-step).\nIn either case (earthkit-data or cdsapi), it is required to set up a CDS account and API key as explained [on the CDS website](https://cds.climate.copernicus.eu/how-to-api).\n\nThis notebook uses [xESMF](https://github.com/pangeo-data/xESMF) for regridding data.\nxESMF is most easily installed using mamba/conda as explained in its documentation.\nUsers who cannot or do not wish to use mamba/conda can manually compile and install [ESMF](https://earthsystemmodeling.org/docs/release/latest/ESMF_usrdoc/node10.html) on their machines.\nIn future, this notebook will use [earthkit-regrid](https://github.com/ecmwf/earthkit-regrid) instead, once it reaches suitable maturity.\n```\n\nNote that the C3S Atlas workflow calculates indicators first, then regrids.\nFor operations that involve averaging, like smoothing and regridding, the order of operations can affect the result, especially in areas with steep gradients [[Avila+15](https://doi.org/10.1016/j.wace.2015.06.003)].\nExamples of such areas for a temperature index are coastlines and mountain ranges.\nIn the case of C3S Atlas, this order of operations was a conscious choice to preserve the \"raw\" signals,\ne.g. preventing extreme temperatures from being smoothed out.\nHowever, it can affect the indicator values and therefore must be considered when using the C3S Atlas application or dataset,\nparticularly when intercomparing it with another dataset.", "text_with_prefix": "EQC Quality Assessment: \"Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets\"\nDataset: multi-origin-c3s-atlas [CDS]\nAspect: consistency_q03 | Category: Derived_Datasets\nSection: Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets > Analysis and results > 2. Calculate and retrieve indicators\n---\nIn the previous two notebooks in this assessment,\nthe origin data were downloaded,\npre-processed,\nused to calculate the relevant indicator(s),\nand regridded;\nafter which the [Atlas dataset](https://doi.org/10.24381/cds.h35hb680) was downloaded.\nThis notebook follows the same structure but for each origin in turn,\nfor clarity and to preserve memory when loading multiple datasets at the same time.\nAs such,\nthe individual steps are described in less detail,\nbecause this information is available in the previous notebooks.\n\nIf you are only interested in specific origin datasets,\nor want to limit your bandwidth or memory usage,\nyou can choose to only run specific subsections below.\n\nThis assessment examines the dataset in one year (2080),\nfor one ensemble member\nand\nfor one climate scenario\n– which is not how climate projection data are normally used.\nGood practice in climate science is to look at multi-year statistics and trends,\nacross multiple ensemble members.\nHowever, since the purpose of this assessment is to assess the consistency and reproducibility of the post-processing performed to produce the C3S Atlas dataset,\nit is valid to use a subset of the data here.\nThe specific subset used can be easily tweaked by changing the `EXPERIMENT`, `MODEL`, and `YEARS` variables in the relevant sections.\n\nThis notebook uses [earthkit-data](https://github.com/ecmwf/earthkit-data) to download files from the CDS.\nIf you intend to run this notebook multiple times, it is highly recommended that you [enable caching](https://earthkit-data.readthedocs.io/en/latest/concepts/caching.html) to prevent having to download the same files multiple times.\nIf you prefer not to use earthkit, the following requests can also be used with the [cdsapi module](https://cds.climate.copernicus.eu/how-to-api#linux-use-client-step).\nIn either case (earthkit-data or cdsapi), it is required to set up a CDS account and API key as explained [on the CDS website](https://cds.climate.copernicus.eu/how-to-api).\n\nThis notebook uses [xESMF](https://github.com/pangeo-data/xESMF) for regridding data.\nxESMF is most easily installed using mamba/conda as explained in its documentation.\nUsers who cannot or do not wish to use mamba/conda can manually compile and install [ESMF](https://earthsystemmodeling.org/docs/release/latest/ESMF_usrdoc/node10.html) on their machines.\nIn future, this notebook will use [earthkit-regrid](https://github.com/ecmwf/earthkit-regrid) instead, once it reaches suitable maturity.\n```\n\nNote that the C3S Atlas workflow calculates indicators first, then regrids.\nFor operations that involve averaging, like smoothing and regridding, the order of operations can affect the result, especially in areas with steep gradients [[Avila+15](https://doi.org/10.1016/j.wace.2015.06.003)].\nExamples of such areas for a temperature index are coastlines and mountain ranges.\nIn the case of C3S Atlas, this order of operations was a conscious choice to preserve the \"raw\" signals,\ne.g. preventing extreme temperatures from being smoothed out.\nHowever, it can affect the indicator values and therefore must be considered when using the C3S Atlas application or dataset,\nparticularly when intercomparing it with another dataset."} {"chunk_id": "derived_multi-origin-c3s-atlas_consistency_q03__99911b4d7016", "report_id": "derived_multi-origin-c3s-atlas_consistency_q03", "dataset_id": "multi-origin-c3s-atlas", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q03", "aspect_base": "consistency", "category": "Derived_Datasets", "match_confidence": "exact", "section": "Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets > Analysis and results > 2. Calculate and retrieve indicators > General setup", "title": "Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets", "chunk_index": 10, "token_count": 172, "text_raw": "Throughout this section,\nwe combine the downloaded datasets into dictionaries for easy access.\nThey cannot be combined into a single xarray object because of differing grids.\nEach dataset is added in its own subsection,\nmeaning any datasets not downloaded will automatically be skipped in the analysis.\n\nConstants to ensure consistency between origins\nChange these to customise the analysis\nAll days and months\n\nSetup: C3S Atlas request\nTemplates for C3S Atlas requests", "text_with_prefix": "EQC Quality Assessment: \"Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets\"\nDataset: multi-origin-c3s-atlas [CDS]\nAspect: consistency_q03 | Category: Derived_Datasets\nSection: Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets > Analysis and results > 2. Calculate and retrieve indicators > General setup\n---\nThroughout this section,\nwe combine the downloaded datasets into dictionaries for easy access.\nThey cannot be combined into a single xarray object because of differing grids.\nEach dataset is added in its own subsection,\nmeaning any datasets not downloaded will automatically be skipped in the analysis.\n\nConstants to ensure consistency between origins\nChange these to customise the analysis\nAll days and months\n\nSetup: C3S Atlas request\nTemplates for C3S Atlas requests"} {"chunk_id": "derived_multi-origin-c3s-atlas_consistency_q03__89659025b3f2", "report_id": "derived_multi-origin-c3s-atlas_consistency_q03", "dataset_id": "multi-origin-c3s-atlas", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q03", "aspect_base": "consistency", "category": "Derived_Datasets", "match_confidence": "exact", "section": "Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets > Analysis and results > 2. Calculate and retrieve indicators > CMIP6", "title": "Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets", "chunk_index": 11, "token_count": 936, "text_raw": "Setup: Customisable options for this dataset\n\nSetup: Origin request\nDownload data, pre-process, calculate indicator, regrid\n\n```text\nUnknown file type, no reader available. path=/home/ob2/cds_data/cds-67885792c67c919c6adef8f9452ca11327458aa2fc79e661bd6835af4bbb1791.d/provenance.png magic=b'\\x89PNG\\r\\n\\x1a\\n\\x00\\x00\\x00\\rIHDR\\x00\\x00\\n\\xd3\\x00\\x00\\x03\"\\x08\\x02\\x00\\x00\\x00\\x99\\xec9+\\x00\\x00\\x00\\x06bKGD\\x00\\xff\\x00\\xff\\x00\\xff\\xa0\\xbd\\xa7\\x93\\x00\\x00 \\x00IDATx\\x9c\\xec\\xddw' content_type=None\n2025-10-13 16:44:59,142 — Homogenization-fixers — INFO — Dataset has already the correct names for its coordinates\n2025-10-13 16:44:59,155 — Homogenization-fixers — INFO — Fixing calendar for Size: 81MB\nDimensions: (time: 365, bnds: 2, lat: 192, lon: 288)\nCoordinates:\n * time (time) object 3kB 2080-01-01 12:00:00 ... 2080-12-31 12:00:00\n * lat (lat) float64 2kB -90.0 -89.06 -88.12 -87.17 ... 88.12 89.06 90.0\n * lon (lon) float64 2kB 0.0 1.25 2.5 3.75 ... 355.0 356.2 357.5 358.8\n height float64 8B ...\nDimensions without coordinates: bnds\nData variables:\n time_bnds (time, bnds) object 6kB dask.array\n lat_bnds (lat, bnds) float64 3kB dask.array\n lon_bnds (lon, bnds) float64 5kB dask.array\n tasmax (time, lat, lon) float32 81MB dask.array\nAttributes: (12/48)\n Conventions: CF-1.7 CMIP-6.2\n activity_id: ScenarioMIP\n branch_method: standard\n branch_time_in_child: 60225.0\n branch_time_in_parent: 60225.0\n comment: none\n ... ...\n title: CMCC-ESM2 output prepared for CMIP6\n variable_id: tasmax\n variant_label: r1i1p1f1\n license: CMIP6 model data produced by CMCC is licensed und...\n cmor_version: 3.6.0\n tracking_id: hdl:21.14100/ba2e335b-8bac-45ec-abbe-f1f16299d2d4\n2025-10-13 16:44:59,332 — UNITS_TRANSFORM — INFO — The dataset tasmax units are not in the correct magnitude. A conversion from K to Celsius will be performed.\n2025-10-13 16:44:59,405 — Homogenization-fixers — INFO — The dataset is in daily or monthly resolution, we don't need to resample it from hourly frequency\n```\n\nSetup: C3S Atlas request\nDownload data, pre-process\n\nAdd to collection", "text_with_prefix": "EQC Quality Assessment: \"Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets\"\nDataset: multi-origin-c3s-atlas [CDS]\nAspect: consistency_q03 | Category: Derived_Datasets\nSection: Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets > Analysis and results > 2. Calculate and retrieve indicators > CMIP6\n---\nSetup: Customisable options for this dataset\n\nSetup: Origin request\nDownload data, pre-process, calculate indicator, regrid\n\n```text\nUnknown file type, no reader available. path=/home/ob2/cds_data/cds-67885792c67c919c6adef8f9452ca11327458aa2fc79e661bd6835af4bbb1791.d/provenance.png magic=b'\\x89PNG\\r\\n\\x1a\\n\\x00\\x00\\x00\\rIHDR\\x00\\x00\\n\\xd3\\x00\\x00\\x03\"\\x08\\x02\\x00\\x00\\x00\\x99\\xec9+\\x00\\x00\\x00\\x06bKGD\\x00\\xff\\x00\\xff\\x00\\xff\\xa0\\xbd\\xa7\\x93\\x00\\x00 \\x00IDATx\\x9c\\xec\\xddw' content_type=None\n2025-10-13 16:44:59,142 — Homogenization-fixers — INFO — Dataset has already the correct names for its coordinates\n2025-10-13 16:44:59,155 — Homogenization-fixers — INFO — Fixing calendar for Size: 81MB\nDimensions: (time: 365, bnds: 2, lat: 192, lon: 288)\nCoordinates:\n * time (time) object 3kB 2080-01-01 12:00:00 ... 2080-12-31 12:00:00\n * lat (lat) float64 2kB -90.0 -89.06 -88.12 -87.17 ... 88.12 89.06 90.0\n * lon (lon) float64 2kB 0.0 1.25 2.5 3.75 ... 355.0 356.2 357.5 358.8\n height float64 8B ...\nDimensions without coordinates: bnds\nData variables:\n time_bnds (time, bnds) object 6kB dask.array\n lat_bnds (lat, bnds) float64 3kB dask.array\n lon_bnds (lon, bnds) float64 5kB dask.array\n tasmax (time, lat, lon) float32 81MB dask.array\nAttributes: (12/48)\n Conventions: CF-1.7 CMIP-6.2\n activity_id: ScenarioMIP\n branch_method: standard\n branch_time_in_child: 60225.0\n branch_time_in_parent: 60225.0\n comment: none\n ... ...\n title: CMCC-ESM2 output prepared for CMIP6\n variable_id: tasmax\n variant_label: r1i1p1f1\n license: CMIP6 model data produced by CMCC is licensed und...\n cmor_version: 3.6.0\n tracking_id: hdl:21.14100/ba2e335b-8bac-45ec-abbe-f1f16299d2d4\n2025-10-13 16:44:59,332 — UNITS_TRANSFORM — INFO — The dataset tasmax units are not in the correct magnitude. A conversion from K to Celsius will be performed.\n2025-10-13 16:44:59,405 — Homogenization-fixers — INFO — The dataset is in daily or monthly resolution, we don't need to resample it from hourly frequency\n```\n\nSetup: C3S Atlas request\nDownload data, pre-process\n\nAdd to collection"} {"chunk_id": "derived_multi-origin-c3s-atlas_consistency_q03__f866aaf91b19", "report_id": "derived_multi-origin-c3s-atlas_consistency_q03", "dataset_id": "multi-origin-c3s-atlas", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q03", "aspect_base": "consistency", "category": "Derived_Datasets", "match_confidence": "exact", "section": "Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets > Analysis and results > 2. Calculate and retrieve indicators > CMIP5", "title": "Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets", "chunk_index": 12, "token_count": 839, "text_raw": "Setup: Customisable options for this dataset\n\nSetup: Origin request\nNote: CMIP5 \"period\" cannot be [YEAR_FUTURE,]\nYou may have to change the following line if you want to change\nYEAR_FUTURE, CMIP5_MODEL, or CMIP5_EXPERIMENT\nDownload data, pre-process, calculate indicator, regrid\n\n```text\n2025-10-13 16:45:09,478 — Homogenization-fixers — INFO — Dataset has already the correct names for its coordinates\n2025-10-13 16:45:09,481 — Homogenization-fixers — INFO — Fixing calendar for Size: 41MB\nDimensions: (time: 366, bnds: 2, lat: 145, lon: 192)\nCoordinates:\n * time (time) datetime64[ns] 3kB 2080-01-01T12:00:00 ... 2080-12-31T1...\n * lat (lat) float64 1kB -90.0 -88.75 -87.5 -86.25 ... 87.5 88.75 90.0\n * lon (lon) float64 2kB 0.0 1.875 3.75 5.625 ... 354.4 356.2 358.1\n height float64 8B ...\nDimensions without coordinates: bnds\nData variables:\n time_bnds (time, bnds) datetime64[ns] 6kB dask.array\n lat_bnds (lat, bnds) float64 2kB dask.array\n lon_bnds (lon, bnds) float64 3kB dask.array\n tasmax (time, lat, lon) float32 41MB dask.array\nAttributes: (12/28)\n institution: CSIRO (Commonwealth Scientific and Industrial Res...\n institute_id: CSIRO-BOM\n experiment_id: rcp85\n source: ACCESS1-0 2011. Atmosphere: AGCM v1.0 (N96 grid-p...\n model_id: ACCESS1-0\n forcing: GHG, Oz, SA, Sl, Vl, BC, OC, (GHG = CO2, N2O, CH4...\n ... ...\n table_id: Table day (01 February 2012) b6353e9919862612c81d...\n title: ACCESS1-0 model output prepared for CMIP5 RCP8.5\n parent_experiment: historical\n modeling_realm: atmos\n realization: 1\n cmor_version: 2.8.0\n2025-10-13 16:45:09,492 — UNITS_TRANSFORM — INFO — The dataset tasmax units are not in the correct magnitude. A conversion from K to Celsius will be performed.\n2025-10-13 16:45:09,543 — Homogenization-fixers — INFO — The dataset is in daily or monthly resolution, we don't need to resample it from hourly frequency\n```\n\nSetup: C3S Atlas request\nDownload data, pre-process\n\nAdd to collection", "text_with_prefix": "EQC Quality Assessment: \"Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets\"\nDataset: multi-origin-c3s-atlas [CDS]\nAspect: consistency_q03 | Category: Derived_Datasets\nSection: Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets > Analysis and results > 2. Calculate and retrieve indicators > CMIP5\n---\nSetup: Customisable options for this dataset\n\nSetup: Origin request\nNote: CMIP5 \"period\" cannot be [YEAR_FUTURE,]\nYou may have to change the following line if you want to change\nYEAR_FUTURE, CMIP5_MODEL, or CMIP5_EXPERIMENT\nDownload data, pre-process, calculate indicator, regrid\n\n```text\n2025-10-13 16:45:09,478 — Homogenization-fixers — INFO — Dataset has already the correct names for its coordinates\n2025-10-13 16:45:09,481 — Homogenization-fixers — INFO — Fixing calendar for Size: 41MB\nDimensions: (time: 366, bnds: 2, lat: 145, lon: 192)\nCoordinates:\n * time (time) datetime64[ns] 3kB 2080-01-01T12:00:00 ... 2080-12-31T1...\n * lat (lat) float64 1kB -90.0 -88.75 -87.5 -86.25 ... 87.5 88.75 90.0\n * lon (lon) float64 2kB 0.0 1.875 3.75 5.625 ... 354.4 356.2 358.1\n height float64 8B ...\nDimensions without coordinates: bnds\nData variables:\n time_bnds (time, bnds) datetime64[ns] 6kB dask.array\n lat_bnds (lat, bnds) float64 2kB dask.array\n lon_bnds (lon, bnds) float64 3kB dask.array\n tasmax (time, lat, lon) float32 41MB dask.array\nAttributes: (12/28)\n institution: CSIRO (Commonwealth Scientific and Industrial Res...\n institute_id: CSIRO-BOM\n experiment_id: rcp85\n source: ACCESS1-0 2011. Atmosphere: AGCM v1.0 (N96 grid-p...\n model_id: ACCESS1-0\n forcing: GHG, Oz, SA, Sl, Vl, BC, OC, (GHG = CO2, N2O, CH4...\n ... ...\n table_id: Table day (01 February 2012) b6353e9919862612c81d...\n title: ACCESS1-0 model output prepared for CMIP5 RCP8.5\n parent_experiment: historical\n modeling_realm: atmos\n realization: 1\n cmor_version: 2.8.0\n2025-10-13 16:45:09,492 — UNITS_TRANSFORM — INFO — The dataset tasmax units are not in the correct magnitude. A conversion from K to Celsius will be performed.\n2025-10-13 16:45:09,543 — Homogenization-fixers — INFO — The dataset is in daily or monthly resolution, we don't need to resample it from hourly frequency\n```\n\nSetup: C3S Atlas request\nDownload data, pre-process\n\nAdd to collection"} {"chunk_id": "derived_multi-origin-c3s-atlas_consistency_q03__22d281711e04", "report_id": "derived_multi-origin-c3s-atlas_consistency_q03", "dataset_id": "multi-origin-c3s-atlas", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q03", "aspect_base": "consistency", "category": "Derived_Datasets", "match_confidence": "exact", "section": "Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets > Analysis and results > 2. Calculate and retrieve indicators > CORDEX-EUR-11", "title": "Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets", "chunk_index": 13, "token_count": 937, "text_raw": "Setup: Customisable options for this dataset\nNote that the model nomenclature is different for the origin dataset (GCM, RCM separate) vs C3S Atlas (combined)\nIf you want to change the model, be sure to edit all of the relevant lines below\n\nSetup: Origin request\nNote: CORDEX \"start_year\" and \"end_year\" cannot be YEAR_FUTURE\nand cannot be determined generically for the various model members\nYou may have to change the following line if you want to change\nYEAR_FUTURE, CORDEX_EUR_11_MODEL, or CORDEX_EUR_11_EXPERIMENT\nDownload data, pre-process, calculate indicator, regrid\n\n```text\n2025-10-13 16:45:13,404 — Homogenization-fixers — INFO — Fixing coordinates names: {'rlon': 'x', 'rlat': 'y'}\n2025-10-13 16:45:13,407 — Homogenization-fixers — INFO — Fixing calendar for Size: 263MB\nDimensions: (time: 366, bnds: 2, y: 412, x: 424, vertices: 4)\nCoordinates:\n * time (time) datetime64[ns] 3kB 2080-01-01T12:00:00...\n * y (y) float64 3kB -23.38 -23.27 ... 21.72 21.84\n * x (x) float64 3kB -28.38 -28.27 ... 18.04 18.16\n lat (y, x) float32 699kB dask.array\n lon (y, x) float32 699kB dask.array\n height float64 8B ...\nDimensions without coordinates: bnds, vertices\nData variables:\n time_bnds (time, bnds) datetime64[ns] 6kB dask.array\n rotated_latitude_longitude int32 4B ...\n lat_vertices (y, x, vertices) float32 3MB dask.array\n lon_vertices (y, x, vertices) float32 3MB dask.array\n tasmax (time, y, x) float32 256MB dask.array\nAttributes: (12/35)\n institution: Helmholtz-Zentrum Geesthacht, Climate Ser...\n institute_id: GERICS\n experiment_id: rcp85\n source: GERICS-REMO2015\n model_id: GERICS-REMO2015\n forcing: N/A\n ... ...\n parent_experiment: N/A\n modeling_realm: atmos\n realization: 1\n cmor_version: 2.9.1\n tracking_id: hdl:21.14103/c3a4e035-9dcd-429f-a86d-8d53...\n c3s_disclaimer: This data has been produced in the contex...\n2025-10-13 16:45:13,425 — UNITS_TRANSFORM — INFO — The dataset tasmax units are not in the correct magnitude. A conversion from K to Celsius will be performed.\n2025-10-13 16:45:13,436 — Homogenization-fixers — INFO — The dataset is in daily or monthly resolution, we don't need to resample it from hourly frequency\n```\n\nSetup: C3S Atlas request\nDownload data, pre-process\n\nAdd to collection", "text_with_prefix": "EQC Quality Assessment: \"Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets\"\nDataset: multi-origin-c3s-atlas [CDS]\nAspect: consistency_q03 | Category: Derived_Datasets\nSection: Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets > Analysis and results > 2. Calculate and retrieve indicators > CORDEX-EUR-11\n---\nSetup: Customisable options for this dataset\nNote that the model nomenclature is different for the origin dataset (GCM, RCM separate) vs C3S Atlas (combined)\nIf you want to change the model, be sure to edit all of the relevant lines below\n\nSetup: Origin request\nNote: CORDEX \"start_year\" and \"end_year\" cannot be YEAR_FUTURE\nand cannot be determined generically for the various model members\nYou may have to change the following line if you want to change\nYEAR_FUTURE, CORDEX_EUR_11_MODEL, or CORDEX_EUR_11_EXPERIMENT\nDownload data, pre-process, calculate indicator, regrid\n\n```text\n2025-10-13 16:45:13,404 — Homogenization-fixers — INFO — Fixing coordinates names: {'rlon': 'x', 'rlat': 'y'}\n2025-10-13 16:45:13,407 — Homogenization-fixers — INFO — Fixing calendar for Size: 263MB\nDimensions: (time: 366, bnds: 2, y: 412, x: 424, vertices: 4)\nCoordinates:\n * time (time) datetime64[ns] 3kB 2080-01-01T12:00:00...\n * y (y) float64 3kB -23.38 -23.27 ... 21.72 21.84\n * x (x) float64 3kB -28.38 -28.27 ... 18.04 18.16\n lat (y, x) float32 699kB dask.array\n lon (y, x) float32 699kB dask.array\n height float64 8B ...\nDimensions without coordinates: bnds, vertices\nData variables:\n time_bnds (time, bnds) datetime64[ns] 6kB dask.array\n rotated_latitude_longitude int32 4B ...\n lat_vertices (y, x, vertices) float32 3MB dask.array\n lon_vertices (y, x, vertices) float32 3MB dask.array\n tasmax (time, y, x) float32 256MB dask.array\nAttributes: (12/35)\n institution: Helmholtz-Zentrum Geesthacht, Climate Ser...\n institute_id: GERICS\n experiment_id: rcp85\n source: GERICS-REMO2015\n model_id: GERICS-REMO2015\n forcing: N/A\n ... ...\n parent_experiment: N/A\n modeling_realm: atmos\n realization: 1\n cmor_version: 2.9.1\n tracking_id: hdl:21.14103/c3a4e035-9dcd-429f-a86d-8d53...\n c3s_disclaimer: This data has been produced in the contex...\n2025-10-13 16:45:13,425 — UNITS_TRANSFORM — INFO — The dataset tasmax units are not in the correct magnitude. A conversion from K to Celsius will be performed.\n2025-10-13 16:45:13,436 — Homogenization-fixers — INFO — The dataset is in daily or monthly resolution, we don't need to resample it from hourly frequency\n```\n\nSetup: C3S Atlas request\nDownload data, pre-process\n\nAdd to collection"} {"chunk_id": "derived_multi-origin-c3s-atlas_consistency_q03__2abdbf1d70c8", "report_id": "derived_multi-origin-c3s-atlas_consistency_q03", "dataset_id": "multi-origin-c3s-atlas", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q03", "aspect_base": "consistency", "category": "Derived_Datasets", "match_confidence": "exact", "section": "Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets > Analysis and results > 2. Calculate and retrieve indicators > ERA5", "title": "Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets", "chunk_index": 14, "token_count": 615, "text_raw": "Note that ERA5 data for a year are >30 GB in size.\nThese data may take up to several hours to download\nand\nrequire sufficient storage to download and cache.\n\nSetup: Origin request\nDownload data, pre-process, calculate indicator\n\n```text\n2025-10-13 16:45:32,294 — Homogenization-fixers — INFO — Fixing coordinates names: {'longitude': 'lon', 'latitude': 'lat'}\n2025-10-13 16:45:32,298 — Homogenization-fixers — INFO — Fixing calendar for Size: 2GB\nDimensions: (time: 365, lat: 721, lon: 1440)\nCoordinates:\n number int64 8B ...\n * lat (lat) float64 6kB 90.0 89.75 89.5 89.25 ... -89.5 -89.75 -90.0\n * lon (lon) float64 12kB 0.0 0.25 0.5 0.75 ... 359.0 359.2 359.5 359.8\n * time (time) datetime64[ns] 3kB 2010-01-01 2010-01-02 ... 2010-12-31\nData variables:\n tasmax (time, lat, lon) float32 2GB dask.array\nAttributes:\n GRIB_centre: ecmf\n GRIB_centreDescription: European Centre for Medium-Range Weather Forecasts\n GRIB_subCentre: 0\n Conventions: CF-1.7\n institution: European Centre for Medium-Range Weather Forecasts\n history: 2025-10-07T20:34 GRIB to CDM+CF via cfgrib-0.9.1...\n2025-10-13 16:45:32,381 — UNITS_TRANSFORM — INFO — The dataset tasmax units are not in the correct magnitude. A conversion from K to Celsius will be performed.\n2025-10-13 16:45:34,651 — Homogenization-fixers — INFO — The dataset is in daily or monthly resolution, we don't need to resample it from hourly frequency\n```\n\nSetup: C3S Atlas request\nDownload data, pre-process\n\nAdd to collection", "text_with_prefix": "EQC Quality Assessment: \"Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets\"\nDataset: multi-origin-c3s-atlas [CDS]\nAspect: consistency_q03 | Category: Derived_Datasets\nSection: Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets > Analysis and results > 2. Calculate and retrieve indicators > ERA5\n---\nNote that ERA5 data for a year are >30 GB in size.\nThese data may take up to several hours to download\nand\nrequire sufficient storage to download and cache.\n\nSetup: Origin request\nDownload data, pre-process, calculate indicator\n\n```text\n2025-10-13 16:45:32,294 — Homogenization-fixers — INFO — Fixing coordinates names: {'longitude': 'lon', 'latitude': 'lat'}\n2025-10-13 16:45:32,298 — Homogenization-fixers — INFO — Fixing calendar for Size: 2GB\nDimensions: (time: 365, lat: 721, lon: 1440)\nCoordinates:\n number int64 8B ...\n * lat (lat) float64 6kB 90.0 89.75 89.5 89.25 ... -89.5 -89.75 -90.0\n * lon (lon) float64 12kB 0.0 0.25 0.5 0.75 ... 359.0 359.2 359.5 359.8\n * time (time) datetime64[ns] 3kB 2010-01-01 2010-01-02 ... 2010-12-31\nData variables:\n tasmax (time, lat, lon) float32 2GB dask.array\nAttributes:\n GRIB_centre: ecmf\n GRIB_centreDescription: European Centre for Medium-Range Weather Forecasts\n GRIB_subCentre: 0\n Conventions: CF-1.7\n institution: European Centre for Medium-Range Weather Forecasts\n history: 2025-10-07T20:34 GRIB to CDM+CF via cfgrib-0.9.1...\n2025-10-13 16:45:32,381 — UNITS_TRANSFORM — INFO — The dataset tasmax units are not in the correct magnitude. A conversion from K to Celsius will be performed.\n2025-10-13 16:45:34,651 — Homogenization-fixers — INFO — The dataset is in daily or monthly resolution, we don't need to resample it from hourly frequency\n```\n\nSetup: C3S Atlas request\nDownload data, pre-process\n\nAdd to collection"} {"chunk_id": "derived_multi-origin-c3s-atlas_consistency_q03__f12e4a847908", "report_id": "derived_multi-origin-c3s-atlas_consistency_q03", "dataset_id": "multi-origin-c3s-atlas", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q03", "aspect_base": "consistency", "category": "Derived_Datasets", "match_confidence": "exact", "section": "Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets > Analysis and results > 2. Calculate and retrieve indicators > ERA5-Land", "title": "Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets", "chunk_index": 15, "token_count": 767, "text_raw": "Note that ERA5-Land data for a year are >200 GB in size.\nThese data may take up to several hours to download\nand\nrequire sufficient storage to download and cache.\n\nSetup: Origin request\nERA5-Land doesn't allow us to download a single year in one go, it has to be split into multiple requests\nDownload data, pre-process, calculate indicator\n\n```text\n100%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 12/12 [00:00<00:00, 247.68it/s]\n2025-10-13 16:45:38,419 — Homogenization-fixers — INFO — Fixing coordinates names: {'longitude': 'lon', 'latitude': 'lat'}\n2025-10-13 16:45:38,422 — Homogenization-fixers — INFO — Fixing calendar for Size: 9GB\nDimensions: (time: 365, lat: 1801, lon: 3600)\nCoordinates:\n number int64 8B 0\n * lat (lat) float64 14kB 90.0 89.9 89.8 89.7 ... -89.7 -89.8 -89.9 -90.0\n * lon (lon) float64 29kB 0.0 0.1 0.2 0.3 0.4 ... 359.6 359.7 359.8 359.9\n * time (time) datetime64[ns] 3kB 2010-01-01 2010-01-02 ... 2010-12-31\nData variables:\n tasmax (time, lat, lon) float32 9GB dask.array\nAttributes:\n GRIB_centre: ecmf\n GRIB_centreDescription: European Centre for Medium-Range Weather Forecasts\n GRIB_subCentre: 0\n Conventions: CF-1.7\n institution: European Centre for Medium-Range Weather Forecasts\n history: 2025-10-08T09:44 GRIB to CDM+CF via cfgrib-0.9.1...\n2025-10-13 16:45:38,476 — UNITS_TRANSFORM — INFO — The dataset tasmax units are not in the correct magnitude. A conversion from K to Celsius will be performed.\n2025-10-13 16:45:40,453 — Homogenization-fixers — INFO — The dataset is in daily or monthly resolution, we don't need to resample it from hourly frequency\n```\n\nSetup: C3S Atlas request\nDownload data, pre-process\n\nRegridding for ERA5-Land\nThe coordinates for C3S Atlas/ERA5-Land are irregular, likely due to floating-point errors.\nTo mitigate this, we round the coords to one digit.\n\nAdd to collection", "text_with_prefix": "EQC Quality Assessment: \"Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets\"\nDataset: multi-origin-c3s-atlas [CDS]\nAspect: consistency_q03 | Category: Derived_Datasets\nSection: Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets > Analysis and results > 2. Calculate and retrieve indicators > ERA5-Land\n---\nNote that ERA5-Land data for a year are >200 GB in size.\nThese data may take up to several hours to download\nand\nrequire sufficient storage to download and cache.\n\nSetup: Origin request\nERA5-Land doesn't allow us to download a single year in one go, it has to be split into multiple requests\nDownload data, pre-process, calculate indicator\n\n```text\n100%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 12/12 [00:00<00:00, 247.68it/s]\n2025-10-13 16:45:38,419 — Homogenization-fixers — INFO — Fixing coordinates names: {'longitude': 'lon', 'latitude': 'lat'}\n2025-10-13 16:45:38,422 — Homogenization-fixers — INFO — Fixing calendar for Size: 9GB\nDimensions: (time: 365, lat: 1801, lon: 3600)\nCoordinates:\n number int64 8B 0\n * lat (lat) float64 14kB 90.0 89.9 89.8 89.7 ... -89.7 -89.8 -89.9 -90.0\n * lon (lon) float64 29kB 0.0 0.1 0.2 0.3 0.4 ... 359.6 359.7 359.8 359.9\n * time (time) datetime64[ns] 3kB 2010-01-01 2010-01-02 ... 2010-12-31\nData variables:\n tasmax (time, lat, lon) float32 9GB dask.array\nAttributes:\n GRIB_centre: ecmf\n GRIB_centreDescription: European Centre for Medium-Range Weather Forecasts\n GRIB_subCentre: 0\n Conventions: CF-1.7\n institution: European Centre for Medium-Range Weather Forecasts\n history: 2025-10-08T09:44 GRIB to CDM+CF via cfgrib-0.9.1...\n2025-10-13 16:45:38,476 — UNITS_TRANSFORM — INFO — The dataset tasmax units are not in the correct magnitude. A conversion from K to Celsius will be performed.\n2025-10-13 16:45:40,453 — Homogenization-fixers — INFO — The dataset is in daily or monthly resolution, we don't need to resample it from hourly frequency\n```\n\nSetup: C3S Atlas request\nDownload data, pre-process\n\nRegridding for ERA5-Land\nThe coordinates for C3S Atlas/ERA5-Land are irregular, likely due to floating-point errors.\nTo mitigate this, we round the coords to one digit.\n\nAdd to collection"} {"chunk_id": "derived_multi-origin-c3s-atlas_consistency_q03__0606c8f14f49", "report_id": "derived_multi-origin-c3s-atlas_consistency_q03", "dataset_id": "multi-origin-c3s-atlas", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q03", "aspect_base": "consistency", "category": "Derived_Datasets", "match_confidence": "exact", "section": "Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets > Analysis and results > 2. Calculate and retrieve indicators > E-OBS", "title": "Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets", "chunk_index": 16, "token_count": 640, "text_raw": "Note that E-OBS data for the full period are >30 GB in size.\nThese data may take up to several hours to download\nand\nrequire sufficient storage to download and cache.\n\nSetup: Origin request\nIt is not possible to provide specific years to \"period\",\nbut a function that picks the shortest period covering all\nYEAR_HISTORICAL would be useful to reduce download size\nDownload data, pre-process, calculate indicator, regrid\n\n```text\n2025-10-13 16:45:43,266 — Homogenization-fixers — INFO — Fixing coordinates names: {'longitude': 'lon', 'latitude': 'lat'}\n2025-10-13 16:45:43,269 — Homogenization-fixers — INFO — Fixing calendar for Size: 479MB\nDimensions: (time: 365, lat: 465, lon: 705)\nCoordinates:\n * lat (lat) float64 4kB 25.05 25.15 25.25 25.35 ... 71.25 71.35 71.45\n * lon (lon) float64 6kB -24.95 -24.85 -24.75 -24.65 ... 45.25 45.35 45.45\n * time (time) datetime64[ns] 3kB 2010-01-01 2010-01-02 ... 2010-12-31\nData variables:\n tasmax (time, lat, lon) float32 479MB dask.array\nAttributes:\n E-OBS_version: 30.0e\n Conventions: CF-1.4\n References: http://surfobs.climate.copernicus.eu/dataaccess/access_eo...\n history: Fri Aug 30 11:43:24 2024: ncks --no-abc -d time,0,27209 /...\n NCO: netCDF Operators version 5.1.8 (Homepage = http://nco.sf....\n2025-10-13 16:45:43,279 — UNITS_TRANSFORM — INFO — The dataset tasmax units are already in the correct magnitude\n2025-10-13 16:45:43,422 — Homogenization-fixers — INFO — The dataset is in daily or monthly resolution, we don't need to resample it from hourly frequency\n```\n\nSetup: C3S Atlas request\nDownload data, pre-process\n\nAdd to collection", "text_with_prefix": "EQC Quality Assessment: \"Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets\"\nDataset: multi-origin-c3s-atlas [CDS]\nAspect: consistency_q03 | Category: Derived_Datasets\nSection: Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets > Analysis and results > 2. Calculate and retrieve indicators > E-OBS\n---\nNote that E-OBS data for the full period are >30 GB in size.\nThese data may take up to several hours to download\nand\nrequire sufficient storage to download and cache.\n\nSetup: Origin request\nIt is not possible to provide specific years to \"period\",\nbut a function that picks the shortest period covering all\nYEAR_HISTORICAL would be useful to reduce download size\nDownload data, pre-process, calculate indicator, regrid\n\n```text\n2025-10-13 16:45:43,266 — Homogenization-fixers — INFO — Fixing coordinates names: {'longitude': 'lon', 'latitude': 'lat'}\n2025-10-13 16:45:43,269 — Homogenization-fixers — INFO — Fixing calendar for Size: 479MB\nDimensions: (time: 365, lat: 465, lon: 705)\nCoordinates:\n * lat (lat) float64 4kB 25.05 25.15 25.25 25.35 ... 71.25 71.35 71.45\n * lon (lon) float64 6kB -24.95 -24.85 -24.75 -24.65 ... 45.25 45.35 45.45\n * time (time) datetime64[ns] 3kB 2010-01-01 2010-01-02 ... 2010-12-31\nData variables:\n tasmax (time, lat, lon) float32 479MB dask.array\nAttributes:\n E-OBS_version: 30.0e\n Conventions: CF-1.4\n References: http://surfobs.climate.copernicus.eu/dataaccess/access_eo...\n history: Fri Aug 30 11:43:24 2024: ncks --no-abc -d time,0,27209 /...\n NCO: netCDF Operators version 5.1.8 (Homepage = http://nco.sf....\n2025-10-13 16:45:43,279 — UNITS_TRANSFORM — INFO — The dataset tasmax units are already in the correct magnitude\n2025-10-13 16:45:43,422 — Homogenization-fixers — INFO — The dataset is in daily or monthly resolution, we don't need to resample it from hourly frequency\n```\n\nSetup: C3S Atlas request\nDownload data, pre-process\n\nAdd to collection"} {"chunk_id": "derived_multi-origin-c3s-atlas_consistency_q03__fd16b79a59e5", "report_id": "derived_multi-origin-c3s-atlas_consistency_q03", "dataset_id": "multi-origin-c3s-atlas", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q03", "aspect_base": "consistency", "category": "Derived_Datasets", "match_confidence": "exact", "section": "Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets > Analysis and results > 2. Calculate and retrieve indicators > BERKEARTH", "title": "Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets", "chunk_index": 17, "token_count": 679, "text_raw": "Setup: Origin request\nDownload data, pre-process, calculate indicator\n\n```text\n2025-10-13 16:45:55,010 — Homogenization-fixers — INFO — Fixing coordinates names: {'longitude': 'lon', 'latitude': 'lat'}\n2025-10-13 16:45:55,015 — Homogenization-fixers — INFO — Fixing calendar for Size: 95MB\nDimensions: (time: 365, lat: 180, lon: 360)\nCoordinates:\n * time (time) datetime64[ns] 3kB 2010-01-01 2010-01-02 ... 2010-12-31\n * lon (lon) float32 1kB -179.5 -178.5 -177.5 -176.5 ... 177.5 178.5 179.5\n * lat (lat) float32 720B -89.5 -88.5 -87.5 -86.5 ... 86.5 87.5 88.5 89.5\nData variables:\n tasmax (time, lat, lon) float32 95MB dask.array\nAttributes: (12/21)\n CDI: Climate Data Interface version 1.9.10 (https:...\n institution: Berkeley Earth Surface Temperature Project\n Conventions: Berkeley Earth Internal Convention (based on ...\n title: Gridded Berkeley Earth Surface Temperature An...\n source_history: 14-Sep-2020 16:15:39\n comment: This file contains surface temperature anomal...\n ... ...\n geospatial_lon_resolution: 1.0\n climexp_url: https://climexp.knmi.nl/select.cgi?berkeley_t...\n history: Fri Jun 18 07:44:32 2021: cdo -O -splityear /...\n time_coverage_start: 1880-01-01 00:00:00\n time_coverage_end: 2019-12-31 00:00:00\n CDO: Climate Data Operators version 1.9.10 (https:...\n2025-10-13 16:45:55,027 — UNITS_TRANSFORM — INFO — The dataset tasmax units are already in the correct magnitude\n2025-10-13 16:45:55,109 — Homogenization-fixers — INFO — The dataset is in daily or monthly resolution, we don't need to resample it from hourly frequency\n```\n\nSetup: C3S Atlas request\nDownload data, pre-process\n\nAdd to collection", "text_with_prefix": "EQC Quality Assessment: \"Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets\"\nDataset: multi-origin-c3s-atlas [CDS]\nAspect: consistency_q03 | Category: Derived_Datasets\nSection: Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets > Analysis and results > 2. Calculate and retrieve indicators > BERKEARTH\n---\nSetup: Origin request\nDownload data, pre-process, calculate indicator\n\n```text\n2025-10-13 16:45:55,010 — Homogenization-fixers — INFO — Fixing coordinates names: {'longitude': 'lon', 'latitude': 'lat'}\n2025-10-13 16:45:55,015 — Homogenization-fixers — INFO — Fixing calendar for Size: 95MB\nDimensions: (time: 365, lat: 180, lon: 360)\nCoordinates:\n * time (time) datetime64[ns] 3kB 2010-01-01 2010-01-02 ... 2010-12-31\n * lon (lon) float32 1kB -179.5 -178.5 -177.5 -176.5 ... 177.5 178.5 179.5\n * lat (lat) float32 720B -89.5 -88.5 -87.5 -86.5 ... 86.5 87.5 88.5 89.5\nData variables:\n tasmax (time, lat, lon) float32 95MB dask.array\nAttributes: (12/21)\n CDI: Climate Data Interface version 1.9.10 (https:...\n institution: Berkeley Earth Surface Temperature Project\n Conventions: Berkeley Earth Internal Convention (based on ...\n title: Gridded Berkeley Earth Surface Temperature An...\n source_history: 14-Sep-2020 16:15:39\n comment: This file contains surface temperature anomal...\n ... ...\n geospatial_lon_resolution: 1.0\n climexp_url: https://climexp.knmi.nl/select.cgi?berkeley_t...\n history: Fri Jun 18 07:44:32 2021: cdo -O -splityear /...\n time_coverage_start: 1880-01-01 00:00:00\n time_coverage_end: 2019-12-31 00:00:00\n CDO: Climate Data Operators version 1.9.10 (https:...\n2025-10-13 16:45:55,027 — UNITS_TRANSFORM — INFO — The dataset tasmax units are already in the correct magnitude\n2025-10-13 16:45:55,109 — Homogenization-fixers — INFO — The dataset is in daily or monthly resolution, we don't need to resample it from hourly frequency\n```\n\nSetup: C3S Atlas request\nDownload data, pre-process\n\nAdd to collection"} {"chunk_id": "derived_multi-origin-c3s-atlas_consistency_q03__584adf07710d", "report_id": "derived_multi-origin-c3s-atlas_consistency_q03", "dataset_id": "multi-origin-c3s-atlas", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q03", "aspect_base": "consistency", "category": "Derived_Datasets", "match_confidence": "exact", "section": "Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets > Analysis and results > 2. Calculate and retrieve indicators > ORAS5", "title": "Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets", "chunk_index": 18, "token_count": 244, "text_raw": "Setup: Origin request\nDownload data, pre-process, calculate indicator, regrid\n\n```text\n2025-10-14 00:08:31,387 — Homogenization-fixers — INFO — Fixing coordinates names: {'nav_lon': 'lon', 'nav_lat': 'lat'}\n2025-10-14 00:08:31,416 — UNITS_TRANSFORM — INFO — The dataset sst units are already in the correct magnitude\n2025-10-14 00:08:31,451 — Homogenization-fixers — INFO — The dataset is in daily or monthly resolution, we don't need to resample it from hourly frequency\n```\n\nSetup: C3S Atlas request\nDownload data, pre-process\n\nAdd to collection", "text_with_prefix": "EQC Quality Assessment: \"Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets\"\nDataset: multi-origin-c3s-atlas [CDS]\nAspect: consistency_q03 | Category: Derived_Datasets\nSection: Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets > Analysis and results > 2. Calculate and retrieve indicators > ORAS5\n---\nSetup: Origin request\nDownload data, pre-process, calculate indicator, regrid\n\n```text\n2025-10-14 00:08:31,387 — Homogenization-fixers — INFO — Fixing coordinates names: {'nav_lon': 'lon', 'nav_lat': 'lat'}\n2025-10-14 00:08:31,416 — UNITS_TRANSFORM — INFO — The dataset sst units are already in the correct magnitude\n2025-10-14 00:08:31,451 — Homogenization-fixers — INFO — The dataset is in daily or monthly resolution, we don't need to resample it from hourly frequency\n```\n\nSetup: C3S Atlas request\nDownload data, pre-process\n\nAdd to collection"} {"chunk_id": "derived_multi-origin-c3s-atlas_consistency_q03__f70053e69519", "report_id": "derived_multi-origin-c3s-atlas_consistency_q03", "dataset_id": "multi-origin-c3s-atlas", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q03", "aspect_base": "consistency", "category": "Derived_Datasets", "match_confidence": "exact", "section": "Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets > Analysis and results > 2. Calculate and retrieve indicators > SST-CCI", "title": "Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets", "chunk_index": 19, "token_count": 1088, "text_raw": "Note that SST-CCI data for a year are >150 GB in size.\nThese data may take up to several hours to download\nand\nrequire sufficient storage to download and cache.\n\nSetup: Origin request\nSST-CCI doesn't allow us to download a single year in one go, it has to be split into multiple requests\nDownload data, pre-process, calculate indicator\n\n```text\n100%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 36/36 [00:00<00:00, 165.43it/s]\n2025-11-03 16:20:40,236 — Homogenization-fixers — INFO — Dataset has already the correct names for its coordinates\n2025-11-03 16:20:40,240 — Homogenization-fixers — INFO — Fixing calendar for Size: 151GB\nDimensions: (time: 365, lat: 3600, lon: 7200, bnds: 2)\nCoordinates:\n * lat (lat) float32 14kB -89.97 -89.93 ... 89.93 89.97\n * lon (lon) float32 29kB -180.0 -179.9 ... 179.9 180.0\n * time (time) datetime64[ns] 3kB 2010-01-01T12:00:00 ....\nDimensions without coordinates: bnds\nData variables:\n sst (time, lat, lon) float32 38GB dask.array\n mask (time, lat, lon) float32 38GB dask.array\n sea_ice_fraction (time, lat, lon) float32 38GB dask.array\n analysed_sst_uncertainty (time, lat, lon) float32 38GB dask.array\n time_bnds (time, bnds) datetime64[ns] 6kB dask.array\n lat_bnds (time, lat, bnds) float32 11MB dask.array\n lon_bnds (time, lon, bnds) float32 21MB dask.array\nAttributes: (12/58)\n Conventions: CF-1.5, Unidata Observation Dataset v1.0\n title: ESA SST CCI OSTIA L4 product\n summary: OSTIA L4 product from the ESA SST CCI pr...\n references: http://www.esa-sst-cci.org\n institution: ESACCI\n history: Created using OSTIA reanalysis system v3.0\n ... ...\n source: ATSR<1,2>-ESACCI-L3U-v2.0, AATSR-ESACCI-...\n platform: ERS-<1,2>, Envisat, NOAA-<07,09,11,12,14...\n creator_name: ESA SST CCI\n product_specification_version: SST_CCI-PSD-UKMO-201-Issue-2\n id: OSTIA-ESACCI-L4-GLOB-v2.1\n product_version: 2.1\n2025-11-03 16:20:40,294 — UNITS_TRANSFORM — INFO — The dataset sst units are not in the correct magnitude. A conversion from Kelvin to Celsius will be performed.\n2025-11-03 16:20:41,695 — Homogenization-fixers — INFO — The dataset is in daily or monthly resolution, we don't need to resample it from hourly frequency\n```\n\nSetup: C3S Atlas request\nDownload data, pre-process\nThe coordinates for C3S Atlas/SST-CCI are irregular, likely due to floating-point errors.\nTo mitigate this, we force it to use the same coordinates as the SST-CCI origin.\nPlease note that this is NOT generally good practice, and is only done here for the sake of the assessment.\n\nAdd to collection", "text_with_prefix": "EQC Quality Assessment: \"Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets\"\nDataset: multi-origin-c3s-atlas [CDS]\nAspect: consistency_q03 | Category: Derived_Datasets\nSection: Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets > Analysis and results > 2. Calculate and retrieve indicators > SST-CCI\n---\nNote that SST-CCI data for a year are >150 GB in size.\nThese data may take up to several hours to download\nand\nrequire sufficient storage to download and cache.\n\nSetup: Origin request\nSST-CCI doesn't allow us to download a single year in one go, it has to be split into multiple requests\nDownload data, pre-process, calculate indicator\n\n```text\n100%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 36/36 [00:00<00:00, 165.43it/s]\n2025-11-03 16:20:40,236 — Homogenization-fixers — INFO — Dataset has already the correct names for its coordinates\n2025-11-03 16:20:40,240 — Homogenization-fixers — INFO — Fixing calendar for Size: 151GB\nDimensions: (time: 365, lat: 3600, lon: 7200, bnds: 2)\nCoordinates:\n * lat (lat) float32 14kB -89.97 -89.93 ... 89.93 89.97\n * lon (lon) float32 29kB -180.0 -179.9 ... 179.9 180.0\n * time (time) datetime64[ns] 3kB 2010-01-01T12:00:00 ....\nDimensions without coordinates: bnds\nData variables:\n sst (time, lat, lon) float32 38GB dask.array\n mask (time, lat, lon) float32 38GB dask.array\n sea_ice_fraction (time, lat, lon) float32 38GB dask.array\n analysed_sst_uncertainty (time, lat, lon) float32 38GB dask.array\n time_bnds (time, bnds) datetime64[ns] 6kB dask.array\n lat_bnds (time, lat, bnds) float32 11MB dask.array\n lon_bnds (time, lon, bnds) float32 21MB dask.array\nAttributes: (12/58)\n Conventions: CF-1.5, Unidata Observation Dataset v1.0\n title: ESA SST CCI OSTIA L4 product\n summary: OSTIA L4 product from the ESA SST CCI pr...\n references: http://www.esa-sst-cci.org\n institution: ESACCI\n history: Created using OSTIA reanalysis system v3.0\n ... ...\n source: ATSR<1,2>-ESACCI-L3U-v2.0, AATSR-ESACCI-...\n platform: ERS-<1,2>, Envisat, NOAA-<07,09,11,12,14...\n creator_name: ESA SST CCI\n product_specification_version: SST_CCI-PSD-UKMO-201-Issue-2\n id: OSTIA-ESACCI-L4-GLOB-v2.1\n product_version: 2.1\n2025-11-03 16:20:40,294 — UNITS_TRANSFORM — INFO — The dataset sst units are not in the correct magnitude. A conversion from Kelvin to Celsius will be performed.\n2025-11-03 16:20:41,695 — Homogenization-fixers — INFO — The dataset is in daily or monthly resolution, we don't need to resample it from hourly frequency\n```\n\nSetup: C3S Atlas request\nDownload data, pre-process\nThe coordinates for C3S Atlas/SST-CCI are irregular, likely due to floating-point errors.\nTo mitigate this, we force it to use the same coordinates as the SST-CCI origin.\nPlease note that this is NOT generally good practice, and is only done here for the sake of the assessment.\n\nAdd to collection"} {"chunk_id": "derived_multi-origin-c3s-atlas_consistency_q03__7076b6b63768", "report_id": "derived_multi-origin-c3s-atlas_consistency_q03", "dataset_id": "multi-origin-c3s-atlas", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q03", "aspect_base": "consistency", "category": "Derived_Datasets", "match_confidence": "exact", "section": "Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets > Analysis and results > 2. Calculate and retrieve indicators > CPC", "title": "Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets", "chunk_index": 20, "token_count": 708, "text_raw": "Setup: Origin request\nDownload data, pre-process, calculate indicator\n\n```text\n2025-10-13 16:23:32,652 — Homogenization-fixers — INFO — Dataset has already the correct names for its coordinates\n2025-10-13 16:23:32,661 — Homogenization-fixers — INFO — Fixing calendar for Size: 757MB\nDimensions: (time: 365, lat: 360, lon: 720)\nCoordinates:\n * time (time) datetime64[ns] 3kB 2010-01-01T12:00:00 ... 2010-12-31T12:...\n * lon (lon) float64 6kB 0.25 0.75 1.25 1.75 ... 358.2 358.8 359.2 359.8\n * lat (lat) float64 3kB -89.75 -89.25 -88.75 -88.25 ... 88.75 89.25 89.75\nData variables:\n pr (time, lat, lon) float64 757MB dask.array\nAttributes: (12/19)\n CDI: Climate Data Interface version 1.9.10 (https:...\n institution: NOAA/NCEP/CPC (converted to netcdf at KNMI)\n Conventions: CF-1.4\n NCO: netCDF Operators version 4.9.0 (Homepage = ht...\n title: CPC unified (gauge-based) precipitation\n source_url: ftp://ftp.cpc.ncep.noaa.gov/precip/CPC_UNI_PR...\n ... ...\n geospatial_lon_resolution: 0.5\n climexp_url: https://climexp.knmi.nl/select.cgi?prcp_cpc_d...\n history: Fri Jun 18 07:52:31 2021: cdo -O -splityear /...\n time_coverage_start: 1979-01-01 12:00:00\n time_coverage_end: 2021-12-31 12:00:00\n CDO: Climate Data Operators version 1.9.10 (https:...\n2025-10-13 16:23:32,818 — UNITS_TRANSFORM — INFO — The dataset pr units are not in the correct magnitude. A conversion from mm day**-1 to mm will be performed.\n2025-10-13 16:23:32,972 — Homogenization-fixers — INFO — The dataset is in daily or monthly resolution, we don't need to resample it from hourly frequency\n```\n\nSetup: C3S Atlas request\nDownload data, pre-process\n\nAdd to collection", "text_with_prefix": "EQC Quality Assessment: \"Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets\"\nDataset: multi-origin-c3s-atlas [CDS]\nAspect: consistency_q03 | Category: Derived_Datasets\nSection: Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets > Analysis and results > 2. Calculate and retrieve indicators > CPC\n---\nSetup: Origin request\nDownload data, pre-process, calculate indicator\n\n```text\n2025-10-13 16:23:32,652 — Homogenization-fixers — INFO — Dataset has already the correct names for its coordinates\n2025-10-13 16:23:32,661 — Homogenization-fixers — INFO — Fixing calendar for Size: 757MB\nDimensions: (time: 365, lat: 360, lon: 720)\nCoordinates:\n * time (time) datetime64[ns] 3kB 2010-01-01T12:00:00 ... 2010-12-31T12:...\n * lon (lon) float64 6kB 0.25 0.75 1.25 1.75 ... 358.2 358.8 359.2 359.8\n * lat (lat) float64 3kB -89.75 -89.25 -88.75 -88.25 ... 88.75 89.25 89.75\nData variables:\n pr (time, lat, lon) float64 757MB dask.array\nAttributes: (12/19)\n CDI: Climate Data Interface version 1.9.10 (https:...\n institution: NOAA/NCEP/CPC (converted to netcdf at KNMI)\n Conventions: CF-1.4\n NCO: netCDF Operators version 4.9.0 (Homepage = ht...\n title: CPC unified (gauge-based) precipitation\n source_url: ftp://ftp.cpc.ncep.noaa.gov/precip/CPC_UNI_PR...\n ... ...\n geospatial_lon_resolution: 0.5\n climexp_url: https://climexp.knmi.nl/select.cgi?prcp_cpc_d...\n history: Fri Jun 18 07:52:31 2021: cdo -O -splityear /...\n time_coverage_start: 1979-01-01 12:00:00\n time_coverage_end: 2021-12-31 12:00:00\n CDO: Climate Data Operators version 1.9.10 (https:...\n2025-10-13 16:23:32,818 — UNITS_TRANSFORM — INFO — The dataset pr units are not in the correct magnitude. A conversion from mm day**-1 to mm will be performed.\n2025-10-13 16:23:32,972 — Homogenization-fixers — INFO — The dataset is in daily or monthly resolution, we don't need to resample it from hourly frequency\n```\n\nSetup: C3S Atlas request\nDownload data, pre-process\n\nAdd to collection"} {"chunk_id": "derived_multi-origin-c3s-atlas_consistency_q03__a325135107a4", "report_id": "derived_multi-origin-c3s-atlas_consistency_q03", "dataset_id": "multi-origin-c3s-atlas", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q03", "aspect_base": "consistency", "category": "Derived_Datasets", "match_confidence": "exact", "section": "Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets > Analysis and results > 2. Calculate and retrieve indicators > Cleanup", "title": "Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets", "chunk_index": 21, "token_count": 133, "text_raw": "Lastly,\nwe manually clear out some memory-intensive objects that are no longer necessary.\n\nWe also [re-chunk](https://docs.xarray.dev/en/latest/user-guide/dask.html) the datasets to be more computationally efficient:\n\n(section-results)=", "text_with_prefix": "EQC Quality Assessment: \"Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets\"\nDataset: multi-origin-c3s-atlas [CDS]\nAspect: consistency_q03 | Category: Derived_Datasets\nSection: Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets > Analysis and results > 2. Calculate and retrieve indicators > Cleanup\n---\nLastly,\nwe manually clear out some memory-intensive objects that are no longer necessary.\n\nWe also [re-chunk](https://docs.xarray.dev/en/latest/user-guide/dask.html) the datasets to be more computationally efficient:\n\n(section-results)="} {"chunk_id": "derived_multi-origin-c3s-atlas_consistency_q03__a9fafebb65c3", "report_id": "derived_multi-origin-c3s-atlas_consistency_q03", "dataset_id": "multi-origin-c3s-atlas", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q03", "aspect_base": "consistency", "category": "Derived_Datasets", "match_confidence": "exact", "section": "Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets > Analysis and results > 3. Results", "title": "Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets", "chunk_index": 22, "token_count": 448, "text_raw": "This section contains the comparison between the indicator values retrieved from the C3S Atlas dataset vs those reproduced from the origin datasets.\n\nThe datasets are first compared on their native grids.\nThis means a point-by-point comparison is not possible\n(because the points are not equivalent),\nbut the distributions can be compared geospatially and overall.\nThis qualitative comparison probes the consistency quality attribute:\nAre the climate indicators in the dataset underpinning the Copernicus Interactive Climate Atlas consistent with their origin datasets?\n\nSecond, the C3S Atlas dataset is compared to the indicators derived from the origin dataset and regridded to the C3S Atlas grid.\nThis makes a quantitative point-by-point comparison possible.\nThis second comparison probes how well the dataset underpinning the Copernicus Interactive Climate Atlas can be reproduced from its origin datasets,\nbased on the workflow (Figure {numref}`{number} `).\n\nFor the geospatial comparison,\nwe display the values of the indicators for one month,\nacross one region and globally.\nAs an example, we display the results across\nEurope\nin\nJune,\nwhich should provide significant spatial variation.\nThis region can easily be modified in the following code cell using the [domains provided by earthkit-plots](https://earthkit-plots.readthedocs.io/en/latest/examples/examples/introduction/07-domains.html).\nSome examples are provided in the cell (commented out using `#`).\n\nSetup: Choose a month to display\nSetup: Pick domain using earthkit-plots\ndomain = \"Mediterranean\"\nOther examples -- uncomment where desired\ndomain = ekp.geo.domains.union([\"Portugal\", \"Spain\"], name=\"Iberia\")\ndomain = \"Italy\"\ndomain = \"South America\"", "text_with_prefix": "EQC Quality Assessment: \"Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets\"\nDataset: multi-origin-c3s-atlas [CDS]\nAspect: consistency_q03 | Category: Derived_Datasets\nSection: Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets > Analysis and results > 3. Results\n---\nThis section contains the comparison between the indicator values retrieved from the C3S Atlas dataset vs those reproduced from the origin datasets.\n\nThe datasets are first compared on their native grids.\nThis means a point-by-point comparison is not possible\n(because the points are not equivalent),\nbut the distributions can be compared geospatially and overall.\nThis qualitative comparison probes the consistency quality attribute:\nAre the climate indicators in the dataset underpinning the Copernicus Interactive Climate Atlas consistent with their origin datasets?\n\nSecond, the C3S Atlas dataset is compared to the indicators derived from the origin dataset and regridded to the C3S Atlas grid.\nThis makes a quantitative point-by-point comparison possible.\nThis second comparison probes how well the dataset underpinning the Copernicus Interactive Climate Atlas can be reproduced from its origin datasets,\nbased on the workflow (Figure {numref}`{number} `).\n\nFor the geospatial comparison,\nwe display the values of the indicators for one month,\nacross one region and globally.\nAs an example, we display the results across\nEurope\nin\nJune,\nwhich should provide significant spatial variation.\nThis region can easily be modified in the following code cell using the [domains provided by earthkit-plots](https://earthkit-plots.readthedocs.io/en/latest/examples/examples/introduction/07-domains.html).\nSome examples are provided in the cell (commented out using `#`).\n\nSetup: Choose a month to display\nSetup: Pick domain using earthkit-plots\ndomain = \"Mediterranean\"\nOther examples -- uncomment where desired\ndomain = ekp.geo.domains.union([\"Portugal\", \"Spain\"], name=\"Iberia\")\ndomain = \"Italy\"\ndomain = \"South America\""} {"chunk_id": "derived_multi-origin-c3s-atlas_consistency_q03__d6c7a695b152", "report_id": "derived_multi-origin-c3s-atlas_consistency_q03", "dataset_id": "multi-origin-c3s-atlas", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q03", "aspect_base": "consistency", "category": "Derived_Datasets", "match_confidence": "exact", "section": "Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets > Analysis and results > 3. Results > Consistency: Comparison on native grids", "title": "Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets", "chunk_index": 23, "token_count": 1029, "text_raw": "As in the\n[previous](./derived_multi-origin-c3s-atlas_consistency_q01)\n[notebooks](./derived_multi-origin-c3s-atlas_consistency_q02),\nit is clear from the geospatial comparisons\n(Figures {numref}`{number} `–{numref}`{number} `)\nthat the C3S Atlas dataset closely resembles a manual reproduction from its origin datasets.\nThe general distribution of indicator values is the same for\nall comparisons.\nThe\nE-OBS `tx35` (Figure {numref}`{number} `)\ncomparison shows clear differences in terms of coverage,\nwhich is explained by differences in versions used\n(26.0e in the production of the C3S Atlas dataset, 30.0e here).\nAdditionally,\nthe CMIP5 comparison\n(Figure {numref}`{number} `)\nshows clear differences here due to the fact that the C3S Atlas version of CMIP5 is regridded to a coarser resolution\nto ensure consistency between the different model members of the ensemble.\n\nThis pattern is also visible in the overall distributions\n(Figures {numref}`{number} `–{numref}`{number} `),\nwhich are again very similar for almost all comparisons.\n\nThe overall conclusion from this comparison is the same as in the\nprevious\nnotebooks.\nThe C3S Atlas dataset and its origins are highly consistent,\nbut small differences exist due to the difference in grid\nand differences in the origin dataset version used and workflow.\nLarge differences were observed\nonly\nin the availability or masking of data in specific areas,\nsuch as\non the edges of the E-OBS dataset.\nUsers of the C3S Atlas dataset\n– and thus users of the C3S Atlas application –\nshould be aware that the indicator values retrieved for a specific location may differ slightly from a manual analysis of the origin dataset.\n\nSet the plotting dates based on user inputs\n\n::::{tab-set}\n:::{tab-item} Projected `tx35`\n:sync: tx35future\n derived_multi-origin-c3s-atlas_consistency_q03_fig-geo-tx35future\n:figwidth: 700px\n:name: \"derived_multi-origin-c3s-atlas_consistency_q03_fig-geo-tx35future\"\n\nComparison between C3S Atlas dataset and reproduction for\nprojected\n`tx35` in one month,\nacross Europe,\non the native grid of each dataset.\n```\n:::\n:::{tab-item} Historical `tx35`\n:sync: tx35historical\n derived_multi-origin-c3s-atlas_consistency_q03_fig-geo-tx35historical\n:figwidth: 700px\n:name: \"derived_multi-origin-c3s-atlas_consistency_q03_fig-geo-tx35historical\"\n\nComparison between C3S Atlas dataset and reproduction for\nhistorical\n`tx35` in one month,\nacross Europe,\non the native grid of each dataset.\n```\n:::\n:::{tab-item} `sst`\n:sync: sst\n derived_multi-origin-c3s-atlas_consistency_q03_fig-geo-sst\n:figwidth: 700px\n:name: \"derived_multi-origin-c3s-atlas_consistency_q03_fig-geo-sst\"\n\nComparison between C3S Atlas dataset and reproduction for\nhistorical\n`sst` in one month,\nacross Europe,\non the native grid of each dataset.\n```\n:::\n:::{tab-item} `r01`\n:sync: r01\n derived_multi-origin-c3s-atlas_consistency_q03_fig-geo-r01\n:figwidth: 700px\n:name: \"derived_multi-origin-c3s-atlas_consistency_q03_fig-geo-r01\"\n\nComparison between C3S Atlas dataset and reproduction for\nhistorical\n`r01` in one month,\nacross Europe,\non the native grid of each dataset.\n```\n:::\n::::", "text_with_prefix": "EQC Quality Assessment: \"Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets\"\nDataset: multi-origin-c3s-atlas [CDS]\nAspect: consistency_q03 | Category: Derived_Datasets\nSection: Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets > Analysis and results > 3. Results > Consistency: Comparison on native grids\n---\nAs in the\n[previous](./derived_multi-origin-c3s-atlas_consistency_q01)\n[notebooks](./derived_multi-origin-c3s-atlas_consistency_q02),\nit is clear from the geospatial comparisons\n(Figures {numref}`{number} `–{numref}`{number} `)\nthat the C3S Atlas dataset closely resembles a manual reproduction from its origin datasets.\nThe general distribution of indicator values is the same for\nall comparisons.\nThe\nE-OBS `tx35` (Figure {numref}`{number} `)\ncomparison shows clear differences in terms of coverage,\nwhich is explained by differences in versions used\n(26.0e in the production of the C3S Atlas dataset, 30.0e here).\nAdditionally,\nthe CMIP5 comparison\n(Figure {numref}`{number} `)\nshows clear differences here due to the fact that the C3S Atlas version of CMIP5 is regridded to a coarser resolution\nto ensure consistency between the different model members of the ensemble.\n\nThis pattern is also visible in the overall distributions\n(Figures {numref}`{number} `–{numref}`{number} `),\nwhich are again very similar for almost all comparisons.\n\nThe overall conclusion from this comparison is the same as in the\nprevious\nnotebooks.\nThe C3S Atlas dataset and its origins are highly consistent,\nbut small differences exist due to the difference in grid\nand differences in the origin dataset version used and workflow.\nLarge differences were observed\nonly\nin the availability or masking of data in specific areas,\nsuch as\non the edges of the E-OBS dataset.\nUsers of the C3S Atlas dataset\n– and thus users of the C3S Atlas application –\nshould be aware that the indicator values retrieved for a specific location may differ slightly from a manual analysis of the origin dataset.\n\nSet the plotting dates based on user inputs\n\n::::{tab-set}\n:::{tab-item} Projected `tx35`\n:sync: tx35future\n derived_multi-origin-c3s-atlas_consistency_q03_fig-geo-tx35future\n:figwidth: 700px\n:name: \"derived_multi-origin-c3s-atlas_consistency_q03_fig-geo-tx35future\"\n\nComparison between C3S Atlas dataset and reproduction for\nprojected\n`tx35` in one month,\nacross Europe,\non the native grid of each dataset.\n```\n:::\n:::{tab-item} Historical `tx35`\n:sync: tx35historical\n derived_multi-origin-c3s-atlas_consistency_q03_fig-geo-tx35historical\n:figwidth: 700px\n:name: \"derived_multi-origin-c3s-atlas_consistency_q03_fig-geo-tx35historical\"\n\nComparison between C3S Atlas dataset and reproduction for\nhistorical\n`tx35` in one month,\nacross Europe,\non the native grid of each dataset.\n```\n:::\n:::{tab-item} `sst`\n:sync: sst\n derived_multi-origin-c3s-atlas_consistency_q03_fig-geo-sst\n:figwidth: 700px\n:name: \"derived_multi-origin-c3s-atlas_consistency_q03_fig-geo-sst\"\n\nComparison between C3S Atlas dataset and reproduction for\nhistorical\n`sst` in one month,\nacross Europe,\non the native grid of each dataset.\n```\n:::\n:::{tab-item} `r01`\n:sync: r01\n derived_multi-origin-c3s-atlas_consistency_q03_fig-geo-r01\n:figwidth: 700px\n:name: \"derived_multi-origin-c3s-atlas_consistency_q03_fig-geo-r01\"\n\nComparison between C3S Atlas dataset and reproduction for\nhistorical\n`r01` in one month,\nacross Europe,\non the native grid of each dataset.\n```\n:::\n::::"} {"chunk_id": "derived_multi-origin-c3s-atlas_consistency_q03__eb90fad6c299", "report_id": "derived_multi-origin-c3s-atlas_consistency_q03", "dataset_id": "multi-origin-c3s-atlas", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q03", "aspect_base": "consistency", "category": "Derived_Datasets", "match_confidence": "exact", "section": "Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets > Analysis and results > 3. Results > Consistency: Comparison on native grids", "title": "Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets", "chunk_index": 24, "token_count": 591, "text_raw": "_q03_fig-geo-r01\"\n\nComparison between C3S Atlas dataset and reproduction for\nhistorical\n`r01` in one month,\nacross Europe,\non the native grid of each dataset.\n```\n:::\n::::\n\n::::{tab-set}\n:::{tab-item} Projected `tx35`\n:sync: tx35future\n derived_multi-origin-c3s-atlas_consistency_q03_fig-hist-tx35future-native\n:figwidth: 700px\n:name: \"derived_multi-origin-c3s-atlas_consistency_q03_fig-hist-tx35future-native\"\n\nComparison between overall distributions of\nprojected\n`tx35`\nvalues in the C3S Atlas dataset and its reproduction,\nacross all spatial and temporal dimensions,\non the native grid of each dataset.\n```\n:::\n:::{tab-item} Historical `tx35`\n:sync: tx35historical\n derived_multi-origin-c3s-atlas_consistency_q03_fig-hist-tx35historical-native\n:figwidth: 700px\n:name: \"derived_multi-origin-c3s-atlas_consistency_q03_fig-hist-tx35historical-native\"\n\nComparison between overall distributions of\nhistorical\n`tx35`\nvalues in the C3S Atlas dataset and its reproduction,\nacross all spatial and temporal dimensions,\non the native grid of each dataset.\n```\n:::\n:::{tab-item} `sst`\n:sync: sst\n derived_multi-origin-c3s-atlas_consistency_q03_fig-hist-sst-native\n:figwidth: 700px\n:name: \"derived_multi-origin-c3s-atlas_consistency_q03_fig-hist-sst-native\"\n\nComparison between overall distributions of\nhistorical\n`sst`\nvalues in the C3S Atlas dataset and its reproduction,\nacross all spatial and temporal dimensions,\non the native grid of each dataset.\n```\n:::\n:::{tab-item} `r01`\n:sync: r01\n derived_multi-origin-c3s-atlas_consistency_q03_fig-hist-r01-native\n:figwidth: 700px\n:name: \"derived_multi-origin-c3s-atlas_consistency_q03_fig-hist-r01-native\"\n\nComparison between overall distributions of\nhistorical\n`r01`\nvalues in the C3S Atlas dataset and its reproduction,\nacross all spatial and temporal dimensions,\non the native grid of each dataset.\n```\n:::\n::::", "text_with_prefix": "EQC Quality Assessment: \"Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets\"\nDataset: multi-origin-c3s-atlas [CDS]\nAspect: consistency_q03 | Category: Derived_Datasets\nSection: Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets > Analysis and results > 3. Results > Consistency: Comparison on native grids\n---\n_q03_fig-geo-r01\"\n\nComparison between C3S Atlas dataset and reproduction for\nhistorical\n`r01` in one month,\nacross Europe,\non the native grid of each dataset.\n```\n:::\n::::\n\n::::{tab-set}\n:::{tab-item} Projected `tx35`\n:sync: tx35future\n derived_multi-origin-c3s-atlas_consistency_q03_fig-hist-tx35future-native\n:figwidth: 700px\n:name: \"derived_multi-origin-c3s-atlas_consistency_q03_fig-hist-tx35future-native\"\n\nComparison between overall distributions of\nprojected\n`tx35`\nvalues in the C3S Atlas dataset and its reproduction,\nacross all spatial and temporal dimensions,\non the native grid of each dataset.\n```\n:::\n:::{tab-item} Historical `tx35`\n:sync: tx35historical\n derived_multi-origin-c3s-atlas_consistency_q03_fig-hist-tx35historical-native\n:figwidth: 700px\n:name: \"derived_multi-origin-c3s-atlas_consistency_q03_fig-hist-tx35historical-native\"\n\nComparison between overall distributions of\nhistorical\n`tx35`\nvalues in the C3S Atlas dataset and its reproduction,\nacross all spatial and temporal dimensions,\non the native grid of each dataset.\n```\n:::\n:::{tab-item} `sst`\n:sync: sst\n derived_multi-origin-c3s-atlas_consistency_q03_fig-hist-sst-native\n:figwidth: 700px\n:name: \"derived_multi-origin-c3s-atlas_consistency_q03_fig-hist-sst-native\"\n\nComparison between overall distributions of\nhistorical\n`sst`\nvalues in the C3S Atlas dataset and its reproduction,\nacross all spatial and temporal dimensions,\non the native grid of each dataset.\n```\n:::\n:::{tab-item} `r01`\n:sync: r01\n derived_multi-origin-c3s-atlas_consistency_q03_fig-hist-r01-native\n:figwidth: 700px\n:name: \"derived_multi-origin-c3s-atlas_consistency_q03_fig-hist-r01-native\"\n\nComparison between overall distributions of\nhistorical\n`r01`\nvalues in the C3S Atlas dataset and its reproduction,\nacross all spatial and temporal dimensions,\non the native grid of each dataset.\n```\n:::\n::::"} {"chunk_id": "derived_multi-origin-c3s-atlas_consistency_q03__cb79cad71440", "report_id": "derived_multi-origin-c3s-atlas_consistency_q03", "dataset_id": "multi-origin-c3s-atlas", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q03", "aspect_base": "consistency", "category": "Derived_Datasets", "match_confidence": "exact", "section": "Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets > Analysis and results > 3. Results > Reproducibility: Comparison on C3S Atlas grid", "title": "Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets", "chunk_index": 25, "token_count": 1003, "text_raw": "After regridding to the C3S Atlas grid,\nthe indicator values reproduced from the origin dataset\ncan be compared point-by-point to the values retrieved from the C3S Atlas dataset.\nWe first examine some metrics that describe the difference Δ between corresponding pixels:\n\nAs in the\n[previous](./derived_multi-origin-c3s-atlas_consistency_q01)\n[notebooks](./derived_multi-origin-c3s-atlas_consistency_q02),\nit is clear that\nthe C3S Atlas dataset and its manual reproduction\nare very similar.\nThe median difference, median absolute difference, and median absolute percentage difference\nare all close to 0\nand the vast majority of pixels show a near-zero difference\n(defined here as |Δ| ≥ ε with ε = 10{sup}`–5` to avoid floating-point errors)\nin all comparisons.\nE-OBS shows the most significant differences,\nlikely explained by differences in the version used\nas in the previous section.\n\nThese observations are confirmed by\nthe overall distributions\n(Figures {numref}`{number} `–{numref}`{number} `)\nand\nthe geospatial distributions\n(Figures {numref}`{number} `–{numref}`{number} `).\nNotably,\nwhile the difference between `tx35` in the C3S Atlas dataset and E-OBS is typically small,\nwith a median of 0 and with ≤4% of pixels showing non-zero differences,\nit has a long tail of large differences\n(Figure {numref}`{number} `).\nThese pixels are concentrated at the periphery of the domain\n(Figures {numref}`{number} `).\nThis is likely the result of a difference in\nthe underlying dataset version\n(as before).\n\nWe can extend the conclusion from the\n[previous](./derived_multi-origin-c3s-atlas_consistency_q01)\n[notebooks](./derived_multi-origin-c3s-atlas_consistency_q02),\nnamely that the C3S Atlas dataset can be considered practically reproducible.\nSome indicators in the C3S Atlas dataset\nare completely identical to\ntheir origin datasets,\ne.g. `r01` in the CPC comparison;\nothers\nshow non-zero differences,\ne.g. some of the pixels in the CMIP6 and E-OBS `tx35` comparisons.\nThe causes of these differences are unclear,\nbut they are generally small and rare enough to be negligible for most users using the C3S Atlas dataset,\nespecially since they will typically use it through the application.\nFor further analysis,\nit is generally best to manually process the origin dataset,\nif only to remove the amount of steps that may affect the result.\n\n::::{tab-set}\n:::{tab-item} Projected `tx35`\n:sync: tx35future\n derived_multi-origin-c3s-atlas_consistency_q03_fig-hist-tx35future-c3s-atlas\n:figwidth: 700px\n:name: \"derived_multi-origin-c3s-atlas_consistency_q03_fig-hist-tx35future-c3s-atlas\"\n\nComparison between overall distributions of\nprojected\n`tx35`\nvalues in the C3S Atlas dataset and its reproduction\non the C3S Atlas grid,\nacross all spatial and temporal dimensions,\nincluding the per-pixel difference.\n\n```\n:::\n:::{tab-item} Historical `tx35`\n:sync: tx35historical\n derived_multi-origin-c3s-atlas_consistency_q03_fig-hist-tx35historical-c3s-atlas\n:figwidth: 700px\n:name: \"derived_multi-origin-c3s-atlas_consistency_q03_fig-hist-tx35historical-c3s-atlas\"", "text_with_prefix": "EQC Quality Assessment: \"Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets\"\nDataset: multi-origin-c3s-atlas [CDS]\nAspect: consistency_q03 | Category: Derived_Datasets\nSection: Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets > Analysis and results > 3. Results > Reproducibility: Comparison on C3S Atlas grid\n---\nAfter regridding to the C3S Atlas grid,\nthe indicator values reproduced from the origin dataset\ncan be compared point-by-point to the values retrieved from the C3S Atlas dataset.\nWe first examine some metrics that describe the difference Δ between corresponding pixels:\n\nAs in the\n[previous](./derived_multi-origin-c3s-atlas_consistency_q01)\n[notebooks](./derived_multi-origin-c3s-atlas_consistency_q02),\nit is clear that\nthe C3S Atlas dataset and its manual reproduction\nare very similar.\nThe median difference, median absolute difference, and median absolute percentage difference\nare all close to 0\nand the vast majority of pixels show a near-zero difference\n(defined here as |Δ| ≥ ε with ε = 10{sup}`–5` to avoid floating-point errors)\nin all comparisons.\nE-OBS shows the most significant differences,\nlikely explained by differences in the version used\nas in the previous section.\n\nThese observations are confirmed by\nthe overall distributions\n(Figures {numref}`{number} `–{numref}`{number} `)\nand\nthe geospatial distributions\n(Figures {numref}`{number} `–{numref}`{number} `).\nNotably,\nwhile the difference between `tx35` in the C3S Atlas dataset and E-OBS is typically small,\nwith a median of 0 and with ≤4% of pixels showing non-zero differences,\nit has a long tail of large differences\n(Figure {numref}`{number} `).\nThese pixels are concentrated at the periphery of the domain\n(Figures {numref}`{number} `).\nThis is likely the result of a difference in\nthe underlying dataset version\n(as before).\n\nWe can extend the conclusion from the\n[previous](./derived_multi-origin-c3s-atlas_consistency_q01)\n[notebooks](./derived_multi-origin-c3s-atlas_consistency_q02),\nnamely that the C3S Atlas dataset can be considered practically reproducible.\nSome indicators in the C3S Atlas dataset\nare completely identical to\ntheir origin datasets,\ne.g. `r01` in the CPC comparison;\nothers\nshow non-zero differences,\ne.g. some of the pixels in the CMIP6 and E-OBS `tx35` comparisons.\nThe causes of these differences are unclear,\nbut they are generally small and rare enough to be negligible for most users using the C3S Atlas dataset,\nespecially since they will typically use it through the application.\nFor further analysis,\nit is generally best to manually process the origin dataset,\nif only to remove the amount of steps that may affect the result.\n\n::::{tab-set}\n:::{tab-item} Projected `tx35`\n:sync: tx35future\n derived_multi-origin-c3s-atlas_consistency_q03_fig-hist-tx35future-c3s-atlas\n:figwidth: 700px\n:name: \"derived_multi-origin-c3s-atlas_consistency_q03_fig-hist-tx35future-c3s-atlas\"\n\nComparison between overall distributions of\nprojected\n`tx35`\nvalues in the C3S Atlas dataset and its reproduction\non the C3S Atlas grid,\nacross all spatial and temporal dimensions,\nincluding the per-pixel difference.\n\n```\n:::\n:::{tab-item} Historical `tx35`\n:sync: tx35historical\n derived_multi-origin-c3s-atlas_consistency_q03_fig-hist-tx35historical-c3s-atlas\n:figwidth: 700px\n:name: \"derived_multi-origin-c3s-atlas_consistency_q03_fig-hist-tx35historical-c3s-atlas\""} {"chunk_id": "derived_multi-origin-c3s-atlas_consistency_q03__abb717099d2b", "report_id": "derived_multi-origin-c3s-atlas_consistency_q03", "dataset_id": "multi-origin-c3s-atlas", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q03", "aspect_base": "consistency", "category": "Derived_Datasets", "match_confidence": "exact", "section": "Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets > Analysis and results > 3. Results > Reproducibility: Comparison on C3S Atlas grid", "title": "Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets", "chunk_index": 26, "token_count": 905, "text_raw": "historical-c3s-atlas\n:figwidth: 700px\n:name: \"derived_multi-origin-c3s-atlas_consistency_q03_fig-hist-tx35historical-c3s-atlas\"\n\nComparison between overall distributions of\nhistorical\n`tx35`\nvalues in the C3S Atlas dataset and its reproduction\non the C3S Atlas grid,\nacross all spatial and temporal dimensions,\nincluding the per-pixel difference.\n```\n:::\n:::{tab-item} `sst`\n:sync: sst\n derived_multi-origin-c3s-atlas_consistency_q03_fig-hist-sst-c3s-atlas\n:figwidth: 700px\n:name: \"derived_multi-origin-c3s-atlas_consistency_q03_fig-hist-sst-c3s-atlas\"\n\nComparison between overall distributions of\nhistorical\n`sst`\nvalues in the C3S Atlas dataset and its reproduction\non the C3S Atlas grid,\nacross all spatial and temporal dimensions,\nincluding the per-pixel difference.\n```\n:::\n:::{tab-item} `r01`\n:sync: r01\n derived_multi-origin-c3s-atlas_consistency_q03_fig-hist-r01-c3s-atlas\n:figwidth: 700px\n:name: \"derived_multi-origin-c3s-atlas_consistency_q03_fig-hist-r01-c3s-atlas\"\n\nComparison between overall distributions of\nhistorical\n`r01`\nvalues in the C3S Atlas dataset and its reproduction\non the C3S Atlas grid,\nacross all spatial and temporal dimensions,\nincluding the per-pixel difference.\n```\n:::\n::::\n\n::::{tab-set}\n:::{tab-item} Projected `tx35`\n:sync: tx35future\n derived_multi-origin-c3s-atlas_consistency_q03_fig-geocomp-tx35future\n:figwidth: 700px\n:name: \"derived_multi-origin-c3s-atlas_consistency_q03_fig-geocomp-tx35future\"\n\nComparison between C3S Atlas dataset and reproduction for\nprojected\n`tx35` in one month,\nacross Europe,\non the C3S Atlas dataset grid,\nincluding the per-pixel difference.\n```\n:::\n:::{tab-item} Historical `tx35`\n:sync: tx35historical\n derived_multi-origin-c3s-atlas_consistency_q03_fig-geocomp-tx35historical\n:figwidth: 700px\n:name: \"derived_multi-origin-c3s-atlas_consistency_q03_fig-geocomp-tx35historical\"\n\nComparison between C3S Atlas dataset and reproduction for\nhistorical\n`tx35` in one month,\nacross Europe,\non the C3S Atlas dataset grid,\nincluding the per-pixel difference.\n```\n:::\n:::{tab-item} `sst`\n:sync: sst\n derived_multi-origin-c3s-atlas_consistency_q03_fig-geocomp-sst\n:figwidth: 700px\n:name: \"derived_multi-origin-c3s-atlas_consistency_q03_fig-geocomp-sst\"\n\nComparison between C3S Atlas dataset and reproduction for\nhistorical\n`sst` in one month,\nacross Europe,\non the C3S Atlas dataset grid,\nincluding the per-pixel difference.\n```\n:::\n:::{tab-item} `r01`\n:sync: r01\n derived_multi-origin-c3s-atlas_consistency_q03_fig-geocomp-r01\n:figwidth: 700px\n:name: \"derived_multi-origin-c3s-atlas_consistency_q03_fig-geocomp-r01\"\n\nComparison between C3S Atlas dataset and reproduction for\nhistorical\n`r01` in one month,\nacross Europe,\non the C3S Atlas dataset grid,\nincluding the per-pixel difference.\n```\n:::\n::::", "text_with_prefix": "EQC Quality Assessment: \"Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets\"\nDataset: multi-origin-c3s-atlas [CDS]\nAspect: consistency_q03 | Category: Derived_Datasets\nSection: Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets > Analysis and results > 3. Results > Reproducibility: Comparison on C3S Atlas grid\n---\nhistorical-c3s-atlas\n:figwidth: 700px\n:name: \"derived_multi-origin-c3s-atlas_consistency_q03_fig-hist-tx35historical-c3s-atlas\"\n\nComparison between overall distributions of\nhistorical\n`tx35`\nvalues in the C3S Atlas dataset and its reproduction\non the C3S Atlas grid,\nacross all spatial and temporal dimensions,\nincluding the per-pixel difference.\n```\n:::\n:::{tab-item} `sst`\n:sync: sst\n derived_multi-origin-c3s-atlas_consistency_q03_fig-hist-sst-c3s-atlas\n:figwidth: 700px\n:name: \"derived_multi-origin-c3s-atlas_consistency_q03_fig-hist-sst-c3s-atlas\"\n\nComparison between overall distributions of\nhistorical\n`sst`\nvalues in the C3S Atlas dataset and its reproduction\non the C3S Atlas grid,\nacross all spatial and temporal dimensions,\nincluding the per-pixel difference.\n```\n:::\n:::{tab-item} `r01`\n:sync: r01\n derived_multi-origin-c3s-atlas_consistency_q03_fig-hist-r01-c3s-atlas\n:figwidth: 700px\n:name: \"derived_multi-origin-c3s-atlas_consistency_q03_fig-hist-r01-c3s-atlas\"\n\nComparison between overall distributions of\nhistorical\n`r01`\nvalues in the C3S Atlas dataset and its reproduction\non the C3S Atlas grid,\nacross all spatial and temporal dimensions,\nincluding the per-pixel difference.\n```\n:::\n::::\n\n::::{tab-set}\n:::{tab-item} Projected `tx35`\n:sync: tx35future\n derived_multi-origin-c3s-atlas_consistency_q03_fig-geocomp-tx35future\n:figwidth: 700px\n:name: \"derived_multi-origin-c3s-atlas_consistency_q03_fig-geocomp-tx35future\"\n\nComparison between C3S Atlas dataset and reproduction for\nprojected\n`tx35` in one month,\nacross Europe,\non the C3S Atlas dataset grid,\nincluding the per-pixel difference.\n```\n:::\n:::{tab-item} Historical `tx35`\n:sync: tx35historical\n derived_multi-origin-c3s-atlas_consistency_q03_fig-geocomp-tx35historical\n:figwidth: 700px\n:name: \"derived_multi-origin-c3s-atlas_consistency_q03_fig-geocomp-tx35historical\"\n\nComparison between C3S Atlas dataset and reproduction for\nhistorical\n`tx35` in one month,\nacross Europe,\non the C3S Atlas dataset grid,\nincluding the per-pixel difference.\n```\n:::\n:::{tab-item} `sst`\n:sync: sst\n derived_multi-origin-c3s-atlas_consistency_q03_fig-geocomp-sst\n:figwidth: 700px\n:name: \"derived_multi-origin-c3s-atlas_consistency_q03_fig-geocomp-sst\"\n\nComparison between C3S Atlas dataset and reproduction for\nhistorical\n`sst` in one month,\nacross Europe,\non the C3S Atlas dataset grid,\nincluding the per-pixel difference.\n```\n:::\n:::{tab-item} `r01`\n:sync: r01\n derived_multi-origin-c3s-atlas_consistency_q03_fig-geocomp-r01\n:figwidth: 700px\n:name: \"derived_multi-origin-c3s-atlas_consistency_q03_fig-geocomp-r01\"\n\nComparison between C3S Atlas dataset and reproduction for\nhistorical\n`r01` in one month,\nacross Europe,\non the C3S Atlas dataset grid,\nincluding the per-pixel difference.\n```\n:::\n::::"} {"chunk_id": "derived_multi-origin-c3s-atlas_consistency_q03__c9d03babfe54", "report_id": "derived_multi-origin-c3s-atlas_consistency_q03", "dataset_id": "multi-origin-c3s-atlas", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q03", "aspect_base": "consistency", "category": "Derived_Datasets", "match_confidence": "exact", "section": "Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets > ℹ️ If you want to know more > Key resources", "title": "Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets", "chunk_index": 27, "token_count": 854, "text_raw": "The CDS catalogue entries for the data used were:\n* Gridded dataset underpinning the Copernicus Interactive Climate Atlas: [multi-origin-c3s-atlas](https://doi.org/10.24381/cds.h35hb680)\n * [](./derived_multi-origin-c3s-atlas_consistency_q01)\n * [](./derived_multi-origin-c3s-atlas_consistency_q02)\n * **Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets**\n* CMIP6 climate projections: [projections-cmip6](https://doi.org/10.24381/cds.c866074c)\n * [Quality assessments for CMIP6](../Climate_Projections/CMIP6/CMIP6.md)\n* CMIP5 daily data on single levels: [projections-cmip5-daily-single-levels](https://doi.org/10.24381/cds.d3513dbf)\n* CORDEX regional climate model data on single levels: [projections-cordex-domains-single-levels](https://doi.org/10.24381/cds.bc91edc3)\n * [Quality assessments for CORDEX](../Climate_Projections/CORDEX/CORDEX.md)\n* ERA5 hourly data on single levels from 1940 to present: [reanalysis-era5-single-levels](https://doi.org/10.24381/cds.adbb2d47)\n * [Quality assessments for Reanalyses](../Reanalyses/reanalysis.md)\n* ERA5-Land hourly data from 1950 to present: [reanalysis-era5-land](https://doi.org/10.24381/cds.e2161bac)\n * [Quality assessments for Reanalyses](../Reanalyses/reanalysis.md)\n* E-OBS daily gridded meteorological data for Europe from 1950 to present derived from in-situ observations: [insitu-gridded-observations-europe](https://doi.org/10.24381/cds.151d3ec6)\n * [Quality assessments for Insitu Observations](../In_Situ/insitu.md)\n* Temperature and precipitation gridded data for global and regional domains derived from in-situ and satellite observations (BERKEARTH, CPC) [insitu-gridded-observations-global-and-regional](https://doi.org/10.24381/cds.11dedf0c)\n * [Quality assessments for Insitu Observations](../In_Situ/insitu.md)\n* ORAS5 global ocean reanalysis monthly data from 1958 to present: [reanalysis-oras5](https://doi.org/10.24381/cds.67e8eeb7)\n * [Quality assessments for Reanalyses](../Reanalyses/reanalysis.md)\n* Sea surface temperature daily data from 1981 to present derived from satellite observations (SST-CCI): [satellite-sea-surface-temperature](https://doi.org/10.24381/cds.cf608234)\n * [Quality assessments for Satellite Observations: Ocean](../Satellite_ECVs/Ocean/Ocean.md)\n\nCode libraries used:\n* [earthkit](https://github.com/ecmwf/earthkit)\n * [earthkit-data](https://github.com/ecmwf/earthkit-data)\n * [earthkit-plots](https://github.com/ecmwf/earthkit-plots)\n* [User-tools for the C3S Atlas](https://github.com/ecmwf-projects/c3s-atlas)\n* [xclim](https://xclim.readthedocs.io/en/stable/) climate indicator tools", "text_with_prefix": "EQC Quality Assessment: \"Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets\"\nDataset: multi-origin-c3s-atlas [CDS]\nAspect: consistency_q03 | Category: Derived_Datasets\nSection: Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets > ℹ️ If you want to know more > Key resources\n---\nThe CDS catalogue entries for the data used were:\n* Gridded dataset underpinning the Copernicus Interactive Climate Atlas: [multi-origin-c3s-atlas](https://doi.org/10.24381/cds.h35hb680)\n * [](./derived_multi-origin-c3s-atlas_consistency_q01)\n * [](./derived_multi-origin-c3s-atlas_consistency_q02)\n * **Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets**\n* CMIP6 climate projections: [projections-cmip6](https://doi.org/10.24381/cds.c866074c)\n * [Quality assessments for CMIP6](../Climate_Projections/CMIP6/CMIP6.md)\n* CMIP5 daily data on single levels: [projections-cmip5-daily-single-levels](https://doi.org/10.24381/cds.d3513dbf)\n* CORDEX regional climate model data on single levels: [projections-cordex-domains-single-levels](https://doi.org/10.24381/cds.bc91edc3)\n * [Quality assessments for CORDEX](../Climate_Projections/CORDEX/CORDEX.md)\n* ERA5 hourly data on single levels from 1940 to present: [reanalysis-era5-single-levels](https://doi.org/10.24381/cds.adbb2d47)\n * [Quality assessments for Reanalyses](../Reanalyses/reanalysis.md)\n* ERA5-Land hourly data from 1950 to present: [reanalysis-era5-land](https://doi.org/10.24381/cds.e2161bac)\n * [Quality assessments for Reanalyses](../Reanalyses/reanalysis.md)\n* E-OBS daily gridded meteorological data for Europe from 1950 to present derived from in-situ observations: [insitu-gridded-observations-europe](https://doi.org/10.24381/cds.151d3ec6)\n * [Quality assessments for Insitu Observations](../In_Situ/insitu.md)\n* Temperature and precipitation gridded data for global and regional domains derived from in-situ and satellite observations (BERKEARTH, CPC) [insitu-gridded-observations-global-and-regional](https://doi.org/10.24381/cds.11dedf0c)\n * [Quality assessments for Insitu Observations](../In_Situ/insitu.md)\n* ORAS5 global ocean reanalysis monthly data from 1958 to present: [reanalysis-oras5](https://doi.org/10.24381/cds.67e8eeb7)\n * [Quality assessments for Reanalyses](../Reanalyses/reanalysis.md)\n* Sea surface temperature daily data from 1981 to present derived from satellite observations (SST-CCI): [satellite-sea-surface-temperature](https://doi.org/10.24381/cds.cf608234)\n * [Quality assessments for Satellite Observations: Ocean](../Satellite_ECVs/Ocean/Ocean.md)\n\nCode libraries used:\n* [earthkit](https://github.com/ecmwf/earthkit)\n * [earthkit-data](https://github.com/ecmwf/earthkit-data)\n * [earthkit-plots](https://github.com/ecmwf/earthkit-plots)\n* [User-tools for the C3S Atlas](https://github.com/ecmwf-projects/c3s-atlas)\n* [xclim](https://xclim.readthedocs.io/en/stable/) climate indicator tools"} {"chunk_id": "derived_multi-origin-c3s-atlas_consistency_q03__193097028691", "report_id": "derived_multi-origin-c3s-atlas_consistency_q03", "dataset_id": "multi-origin-c3s-atlas", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q03", "aspect_base": "consistency", "category": "Derived_Datasets", "match_confidence": "exact", "section": "Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets > ℹ️ If you want to know more > Key resources", "title": "Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets", "chunk_index": 28, "token_count": 390, "text_raw": "](https://github.com/ecmwf-projects/c3s-atlas)\n* [xclim](https://xclim.readthedocs.io/en/stable/) climate indicator tools\n\nMore about the Copernicus Interactive Climate Atlas and its IPCC predecessor:\n* [Copernicus Interactive Climate Atlas application](https://atlas.climate.copernicus.eu/)\n* [Gridded data underpinning the Copernicus Interactive Climate Atlas: Description of the datasets and variables](https://confluence.ecmwf.int/display/CKB/Gridded+data+underpinning+the+Copernicus+Interactive+Climate+Atlas%3A+Description+of+the+datasets+and+variables)\n* [The Copernicus Interactive Climate Atlas: a tool to explore regional climate change](https://doi.org/10.21957/ah52ufc369)\n* [Copernicus Interactive Climate Atlas: a new tool to visualise climate variability and change](https://www.ecmwf.int/en/newsletter/179/news/copernicus-interactive-climate-atlas-new-tool-visualise-climate-variability)\n* [Implementation of FAIR principles in the IPCC: the WGI AR6 Atlas repository](https://doi.org/10.1038/s41597-022-01739-y)\n* [Climate Change 2021 – The Physical Science Basis: Atlas](https://doi.org/10.1017/9781009157896.021)", "text_with_prefix": "EQC Quality Assessment: \"Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets\"\nDataset: multi-origin-c3s-atlas [CDS]\nAspect: consistency_q03 | Category: Derived_Datasets\nSection: Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets > ℹ️ If you want to know more > Key resources\n---\n](https://github.com/ecmwf-projects/c3s-atlas)\n* [xclim](https://xclim.readthedocs.io/en/stable/) climate indicator tools\n\nMore about the Copernicus Interactive Climate Atlas and its IPCC predecessor:\n* [Copernicus Interactive Climate Atlas application](https://atlas.climate.copernicus.eu/)\n* [Gridded data underpinning the Copernicus Interactive Climate Atlas: Description of the datasets and variables](https://confluence.ecmwf.int/display/CKB/Gridded+data+underpinning+the+Copernicus+Interactive+Climate+Atlas%3A+Description+of+the+datasets+and+variables)\n* [The Copernicus Interactive Climate Atlas: a tool to explore regional climate change](https://doi.org/10.21957/ah52ufc369)\n* [Copernicus Interactive Climate Atlas: a new tool to visualise climate variability and change](https://www.ecmwf.int/en/newsletter/179/news/copernicus-interactive-climate-atlas-new-tool-visualise-climate-variability)\n* [Implementation of FAIR principles in the IPCC: the WGI AR6 Atlas repository](https://doi.org/10.1038/s41597-022-01739-y)\n* [Climate Change 2021 – The Physical Science Basis: Atlas](https://doi.org/10.1017/9781009157896.021)"} {"chunk_id": "derived_multi-origin-c3s-atlas_consistency_q03__cc0c0f0a560b", "report_id": "derived_multi-origin-c3s-atlas_consistency_q03", "dataset_id": "multi-origin-c3s-atlas", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q03", "aspect_base": "consistency", "category": "Derived_Datasets", "match_confidence": "exact", "section": "Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets > ℹ️ If you want to know more > References", "title": "Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets", "chunk_index": 29, "token_count": 441, "text_raw": "[[Avila+15](https://doi.org/10.1016/j.wace.2015.06.003)] F. B. Avila et al., ‘Systematic investigation of gridding-related scaling effects on annual statistics of daily temperature and precipitation maxima: A case study for south-east Australia’, Weather and Climate Extremes, vol. 9, pp. 6–16, Aug. 2015, doi: 10.1016/j.wace.2015.06.003.\n\n[[C3S Atlas dataset](https://doi.org/10.24381/cds.h35hb680)] Copernicus Climate Change Service, ‘Gridded dataset underpinning the Copernicus Interactive Climate Atlas’. Copernicus Climate Change Service (C3S) Climate Data Store (CDS), Jun. 17, 2024. doi: 10.24381/cds.h35hb680.\n\n[[CMIP6 dataset](https://doi.org/10.24381/cds.c866074c)] Copernicus Climate Change Service, ‘CMIP6 climate projections’. Copernicus Climate Change Service (C3S) Climate Data Store (CDS), Mar. 23, 2021. doi: 10.24381/cds.c866074c.\n\n[[Gutiérrez+24](https://doi.org/10.21957/ah52ufc369)] J. M. Gutiérrez et al., ‘The Copernicus Interactive Climate Atlas: a tool to explore regional climate change’, ECMWF Newsletter, vol. 181, pp. 38–45, Oct. 2024, doi: 10.21957/ah52ufc369.", "text_with_prefix": "EQC Quality Assessment: \"Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets\"\nDataset: multi-origin-c3s-atlas [CDS]\nAspect: consistency_q03 | Category: Derived_Datasets\nSection: Consistency between the C3S Atlas dataset and its origins: Multiple origin datasets > ℹ️ If you want to know more > References\n---\n[[Avila+15](https://doi.org/10.1016/j.wace.2015.06.003)] F. B. Avila et al., ‘Systematic investigation of gridding-related scaling effects on annual statistics of daily temperature and precipitation maxima: A case study for south-east Australia’, Weather and Climate Extremes, vol. 9, pp. 6–16, Aug. 2015, doi: 10.1016/j.wace.2015.06.003.\n\n[[C3S Atlas dataset](https://doi.org/10.24381/cds.h35hb680)] Copernicus Climate Change Service, ‘Gridded dataset underpinning the Copernicus Interactive Climate Atlas’. Copernicus Climate Change Service (C3S) Climate Data Store (CDS), Jun. 17, 2024. doi: 10.24381/cds.h35hb680.\n\n[[CMIP6 dataset](https://doi.org/10.24381/cds.c866074c)] Copernicus Climate Change Service, ‘CMIP6 climate projections’. Copernicus Climate Change Service (C3S) Climate Data Store (CDS), Mar. 23, 2021. doi: 10.24381/cds.c866074c.\n\n[[Gutiérrez+24](https://doi.org/10.21957/ah52ufc369)] J. M. Gutiérrez et al., ‘The Copernicus Interactive Climate Atlas: a tool to explore regional climate change’, ECMWF Newsletter, vol. 181, pp. 38–45, Oct. 2024, doi: 10.21957/ah52ufc369."} {"chunk_id": "satellite_satellite-total-column-water-vapour-ocean_validation_q01__5b92a5b02d96", "report_id": "satellite_satellite-total-column-water-vapour-ocean_validation_q01", "dataset_id": "satellite-total-column-water-vapour-ocean", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q01", "aspect_base": "validation", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Water vapor amplification of Earth's Greenhouse Effect", "title": "Water vapor amplification of Earth's Greenhouse Effect", "chunk_index": 0, "token_count": 115, "text_raw": "Production date: 17-04-2025\n\nProduced by : UVSQ/LSCE-IPSL (Hélène Brogniez) ; CNRS/LMD-IPSL (Giulio Mandorli, Claudia Stubenrauch)", "text_with_prefix": "EQC Quality Assessment: \"Water vapor amplification of Earth's Greenhouse Effect\"\nDataset: satellite-total-column-water-vapour-ocean [CDS]\nAspect: validation_q01 | Category: Satellite_ECVs\nSection: Water vapor amplification of Earth's Greenhouse Effect\n---\nProduction date: 17-04-2025\n\nProduced by : UVSQ/LSCE-IPSL (Hélène Brogniez) ; CNRS/LMD-IPSL (Giulio Mandorli, Claudia Stubenrauch)"} {"chunk_id": "satellite_satellite-total-column-water-vapour-ocean_validation_q01__17c779970ef7", "report_id": "satellite_satellite-total-column-water-vapour-ocean_validation_q01", "dataset_id": "satellite-total-column-water-vapour-ocean", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q01", "aspect_base": "validation", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Water vapor amplification of Earth's Greenhouse Effect > Quality assessment question:", "title": "Water vapor amplification of Earth's Greenhouse Effect", "chunk_index": 1, "token_count": 510, "text_raw": "- __Can satellite measurements reproduce the known relationship between clear sky greenhouse effect and total column water vapour (TCWV)?__\n\nThe water vapour is the most significant greenhouse gases, contributing to about half of the planet's global greenhouse effect ([[1]](https://doi.org/10.1029/2010JD014287), [[2]](https://doi.org/10.1175/1520-0477%281997%29078<0197:EAGMEB>2.0.CO;2)). As a result, it plays a important role in shaping the Earth's radiation budget.\n\nThe warming induced by a forcing, such as an increase in CO$_2$ concentrations, leads to higher levels of water vapor through the Clausius-Clapeyron relationship [[3]](https://doi.org/10.1175/1520-0469%281967%29024<0241:TEOTAW>2.0.CO;2). The increasing of water vapor, in turn, intensifies the greenhouse effect because of its strong radiative properties in the thermal infrared : this is the well-known positive water vapor feedback that amplifies by almost a factor of two the initial warming [[4]](https://doi.org/10.1146/annurev.energy.25.1.441). This positive feedback mechanism underscores the critical importance of monitoring and comprehending the water vapour concentrations to accurately assess and predict its impacts on the Earth's climate system.\n\nIn this analysis, we study the contribution of water vapor to the greenhouse effect and aim to determine the relationship between the two variables. We quantify the amount of water vapor in the atmosphere using its vertially integrated value, known as Total Column Water Vapor (TCWV), which is obtained from satellite observations. The data used in this study is __Monthly and 6-hourly total column water vapour over ocean from 1988 to 2020 derived from satellite observations__ [[described here]](https://cds.climate.copernicus.eu/datasets/satellite-total-column-water-vapour-ocean), available on the Climate Data Store of the Copernicus Climate Change Service.", "text_with_prefix": "EQC Quality Assessment: \"Water vapor amplification of Earth's Greenhouse Effect\"\nDataset: satellite-total-column-water-vapour-ocean [CDS]\nAspect: validation_q01 | Category: Satellite_ECVs\nSection: Water vapor amplification of Earth's Greenhouse Effect > Quality assessment question:\n---\n- __Can satellite measurements reproduce the known relationship between clear sky greenhouse effect and total column water vapour (TCWV)?__\n\nThe water vapour is the most significant greenhouse gases, contributing to about half of the planet's global greenhouse effect ([[1]](https://doi.org/10.1029/2010JD014287), [[2]](https://doi.org/10.1175/1520-0477%281997%29078<0197:EAGMEB>2.0.CO;2)). As a result, it plays a important role in shaping the Earth's radiation budget.\n\nThe warming induced by a forcing, such as an increase in CO$_2$ concentrations, leads to higher levels of water vapor through the Clausius-Clapeyron relationship [[3]](https://doi.org/10.1175/1520-0469%281967%29024<0241:TEOTAW>2.0.CO;2). The increasing of water vapor, in turn, intensifies the greenhouse effect because of its strong radiative properties in the thermal infrared : this is the well-known positive water vapor feedback that amplifies by almost a factor of two the initial warming [[4]](https://doi.org/10.1146/annurev.energy.25.1.441). This positive feedback mechanism underscores the critical importance of monitoring and comprehending the water vapour concentrations to accurately assess and predict its impacts on the Earth's climate system.\n\nIn this analysis, we study the contribution of water vapor to the greenhouse effect and aim to determine the relationship between the two variables. We quantify the amount of water vapor in the atmosphere using its vertially integrated value, known as Total Column Water Vapor (TCWV), which is obtained from satellite observations. The data used in this study is __Monthly and 6-hourly total column water vapour over ocean from 1988 to 2020 derived from satellite observations__ [[described here]](https://cds.climate.copernicus.eu/datasets/satellite-total-column-water-vapour-ocean), available on the Climate Data Store of the Copernicus Climate Change Service."} {"chunk_id": "satellite_satellite-total-column-water-vapour-ocean_validation_q01__2e10f150b47e", "report_id": "satellite_satellite-total-column-water-vapour-ocean_validation_q01", "dataset_id": "satellite-total-column-water-vapour-ocean", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q01", "aspect_base": "validation", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Water vapor amplification of Earth's Greenhouse Effect > Quality assessment statement", "title": "Water vapor amplification of Earth's Greenhouse Effect", "chunk_index": 2, "token_count": 419, "text_raw": "These are the key outcomes of this assessment\n\n- The clear sky greenhouse effect is approximately a linear function of the total column water vapour [[5]](https://doi.org/10.1073/pnas.1809868115).\n\n- This analysis is performed with one month of data. Over this month, the relationship between the greenhouse effect, estimated for clear sky situations, and the total column water vapour is consistent with the literature, the value of the slope being in agreement with [[6]](https://doi.org/full/10.1029/2000JD000040), [[7]](https://doi.org/10.1002/2013JD020184) and [[8]](https://doi.org/10.1029/2019JD031017).\n\n- An estimate of the clear sky water vapour feedback is computed by assuming that the relative humidity remains constant with warming, as in [[4]](https://doi.org/10.1146/annurev.energy.25.1.441). The humidity feedback obtained is $2.5 \\pm 0.2$ $W/m^{2}/ K$ which is in agreement with the IPCC report on this topic (Chapter 7, [[9]](https://www.ipcc.ch/report/ar6/wg1/downloads/report/IPCC_AR6_WGI_Chapter07.pdf))\n\n- Despite assumptions and the use of only one month, the general agreement with the literature highlights the value of this water vapor data record for studies of the greenhouse effect of water vapor. The combination with other data records (cloud amount, sea surface temperature, and outgoing longwave radiation) of the CDS shows also the relevance of these data records for such climate studies.\n```", "text_with_prefix": "EQC Quality Assessment: \"Water vapor amplification of Earth's Greenhouse Effect\"\nDataset: satellite-total-column-water-vapour-ocean [CDS]\nAspect: validation_q01 | Category: Satellite_ECVs\nSection: Water vapor amplification of Earth's Greenhouse Effect > Quality assessment statement\n---\nThese are the key outcomes of this assessment\n\n- The clear sky greenhouse effect is approximately a linear function of the total column water vapour [[5]](https://doi.org/10.1073/pnas.1809868115).\n\n- This analysis is performed with one month of data. Over this month, the relationship between the greenhouse effect, estimated for clear sky situations, and the total column water vapour is consistent with the literature, the value of the slope being in agreement with [[6]](https://doi.org/full/10.1029/2000JD000040), [[7]](https://doi.org/10.1002/2013JD020184) and [[8]](https://doi.org/10.1029/2019JD031017).\n\n- An estimate of the clear sky water vapour feedback is computed by assuming that the relative humidity remains constant with warming, as in [[4]](https://doi.org/10.1146/annurev.energy.25.1.441). The humidity feedback obtained is $2.5 \\pm 0.2$ $W/m^{2}/ K$ which is in agreement with the IPCC report on this topic (Chapter 7, [[9]](https://www.ipcc.ch/report/ar6/wg1/downloads/report/IPCC_AR6_WGI_Chapter07.pdf))\n\n- Despite assumptions and the use of only one month, the general agreement with the literature highlights the value of this water vapor data record for studies of the greenhouse effect of water vapor. The combination with other data records (cloud amount, sea surface temperature, and outgoing longwave radiation) of the CDS shows also the relevance of these data records for such climate studies.\n```"} {"chunk_id": "satellite_satellite-total-column-water-vapour-ocean_validation_q01__1f41e4a4ec0a", "report_id": "satellite_satellite-total-column-water-vapour-ocean_validation_q01", "dataset_id": "satellite-total-column-water-vapour-ocean", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q01", "aspect_base": "validation", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Water vapor amplification of Earth's Greenhouse Effect > Methodology", "title": "Water vapor amplification of Earth's Greenhouse Effect", "chunk_index": 3, "token_count": 136, "text_raw": "The relationship between clear sky geenhouse effect and total column of water vapour is examined using the available datasets from the Climate Data Record (CDR) for sea surface temperature, clouds, and outgoing longwave radiation, detailed below.\n\nThese datasets are daily and the analysis focuses on the __tropical ocean__ over February 2007.\n\n(Method-section-1)=", "text_with_prefix": "EQC Quality Assessment: \"Water vapor amplification of Earth's Greenhouse Effect\"\nDataset: satellite-total-column-water-vapour-ocean [CDS]\nAspect: validation_q01 | Category: Satellite_ECVs\nSection: Water vapor amplification of Earth's Greenhouse Effect > Methodology\n---\nThe relationship between clear sky geenhouse effect and total column of water vapour is examined using the available datasets from the Climate Data Record (CDR) for sea surface temperature, clouds, and outgoing longwave radiation, detailed below.\n\nThese datasets are daily and the analysis focuses on the __tropical ocean__ over February 2007.\n\n(Method-section-1)="} {"chunk_id": "satellite_satellite-total-column-water-vapour-ocean_validation_q01__e48a91349c39", "report_id": "satellite_satellite-total-column-water-vapour-ocean_validation_q01", "dataset_id": "satellite-total-column-water-vapour-ocean", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q01", "aspect_base": "validation", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Water vapor amplification of Earth's Greenhouse Effect > Methodology > The Greenhouse effect", "title": "Water vapor amplification of Earth's Greenhouse Effect", "chunk_index": 4, "token_count": 765, "text_raw": "The greenhouse effect (G, in W/m$^2$) is defined by the difference, in terms of longwave thermal radiation, between the radiation emitted to space at the top of the atmosphere, which is the OLR (in W/m$^2$), and the radiation emitted upward by the surface noted $F_{surf}$:\n\n$$ \n\\text{G} = \\text{F}_{surf} - \\text{OLR} \n$$\n\nWhere $\\text{F}_{surf} = \\epsilon_{surf} \\ \\sigma \\ T_{surf} ^4 $\n\n$T_{surf}$ is the surface temperature (in K), $\\epsilon_{surf}$ is the surface emissivity (unitless) and $\\sigma$ is the Stefan-Boltzmann constant ($\\sigma = 5.67 \\ 10^{-8}$ W/m$^2$/K$^4$).\n\nWith this definition, G has always positive values, the atmosphere trapping a part of the upward longwave radiation emitted by the surface, this trapping being done by greenhouse gases, clouds and, for a smaller part, aerosols. In the absence of clouds and aerosols, the clear sky greenhouse effect $\\text{G}_{CS}$ is dominated by water vapour which accounts for 50 to 60% of the total greenhouse effect ([[1]](https://doi.org/10.1029/2010JD014287), [[2]](https://doi.org/10.1175/1520-0477%281997%29078<0197:EAGMEB>2.0.CO;2)).\n\nattachment:ae4b0f92-acfa-4a45-81db-501203233bd9.png\n---\nheight: 300px\n---\nContributions of the individual absorbers to the clear sky greenhouse effect. Numbers taken from [[5]](https://doi.org/10.1073/pnas.1809868115).\n```\n\nThe estimation of $\\text{G}_{CS}$ is performed by selecting only clear sky regions identified at the daily scale using the cloud fraction cover.\n\nThe computation of $\\text{F}_{surf}$ assumes that the sea water emits thermal radiation very similar to blackbody emission, therefore $\\epsilon_{surf}=1$. Considering a blackbody emission rather the true emissivity of sea water yield a small difference of less than 1\\% as discussed by [[5]](https://doi.org/10.1073/pnas.1809868115).\n\nThe obtained $\\text{G}_{CS}$ is then compared with the corresponding TCWV, and their relationship is visualized as a function of TCWV.\nLastly, this derived relationship is utilized to estimate the water vapour feedback, $f_{wv}$.\n\nSeveral methods are used to compute the greenhouse effect and the water vapor feedbacks [[6]](https://doi.org/full/10.1029/2000JD000040):\n- a partial radiative perturbation method, comparing a control climate and a perturbed climate \n- a radiative kernel technique, comparing point-by-point perturbed profiles of water vapor and their radiative effect\n- a linear regression between the clear sky longwave top-of-the-atmosphere flux to the global mean.\n\nHere we use the latter method to assess the relevance of CDS data for estimating the water vapor feedback\n\n(method-section-2)=", "text_with_prefix": "EQC Quality Assessment: \"Water vapor amplification of Earth's Greenhouse Effect\"\nDataset: satellite-total-column-water-vapour-ocean [CDS]\nAspect: validation_q01 | Category: Satellite_ECVs\nSection: Water vapor amplification of Earth's Greenhouse Effect > Methodology > The Greenhouse effect\n---\nThe greenhouse effect (G, in W/m$^2$) is defined by the difference, in terms of longwave thermal radiation, between the radiation emitted to space at the top of the atmosphere, which is the OLR (in W/m$^2$), and the radiation emitted upward by the surface noted $F_{surf}$:\n\n$$ \n\\text{G} = \\text{F}_{surf} - \\text{OLR} \n$$\n\nWhere $\\text{F}_{surf} = \\epsilon_{surf} \\ \\sigma \\ T_{surf} ^4 $\n\n$T_{surf}$ is the surface temperature (in K), $\\epsilon_{surf}$ is the surface emissivity (unitless) and $\\sigma$ is the Stefan-Boltzmann constant ($\\sigma = 5.67 \\ 10^{-8}$ W/m$^2$/K$^4$).\n\nWith this definition, G has always positive values, the atmosphere trapping a part of the upward longwave radiation emitted by the surface, this trapping being done by greenhouse gases, clouds and, for a smaller part, aerosols. In the absence of clouds and aerosols, the clear sky greenhouse effect $\\text{G}_{CS}$ is dominated by water vapour which accounts for 50 to 60% of the total greenhouse effect ([[1]](https://doi.org/10.1029/2010JD014287), [[2]](https://doi.org/10.1175/1520-0477%281997%29078<0197:EAGMEB>2.0.CO;2)).\n\nattachment:ae4b0f92-acfa-4a45-81db-501203233bd9.png\n---\nheight: 300px\n---\nContributions of the individual absorbers to the clear sky greenhouse effect. Numbers taken from [[5]](https://doi.org/10.1073/pnas.1809868115).\n```\n\nThe estimation of $\\text{G}_{CS}$ is performed by selecting only clear sky regions identified at the daily scale using the cloud fraction cover.\n\nThe computation of $\\text{F}_{surf}$ assumes that the sea water emits thermal radiation very similar to blackbody emission, therefore $\\epsilon_{surf}=1$. Considering a blackbody emission rather the true emissivity of sea water yield a small difference of less than 1\\% as discussed by [[5]](https://doi.org/10.1073/pnas.1809868115).\n\nThe obtained $\\text{G}_{CS}$ is then compared with the corresponding TCWV, and their relationship is visualized as a function of TCWV.\nLastly, this derived relationship is utilized to estimate the water vapour feedback, $f_{wv}$.\n\nSeveral methods are used to compute the greenhouse effect and the water vapor feedbacks [[6]](https://doi.org/full/10.1029/2000JD000040):\n- a partial radiative perturbation method, comparing a control climate and a perturbed climate \n- a radiative kernel technique, comparing point-by-point perturbed profiles of water vapor and their radiative effect\n- a linear regression between the clear sky longwave top-of-the-atmosphere flux to the global mean.\n\nHere we use the latter method to assess the relevance of CDS data for estimating the water vapor feedback\n\n(method-section-2)="} {"chunk_id": "satellite_satellite-total-column-water-vapour-ocean_validation_q01__cd508e9672d4", "report_id": "satellite_satellite-total-column-water-vapour-ocean_validation_q01", "dataset_id": "satellite-total-column-water-vapour-ocean", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q01", "aspect_base": "validation", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Water vapor amplification of Earth's Greenhouse Effect > Methodology > The water vapour feedback", "title": "Water vapor amplification of Earth's Greenhouse Effect", "chunk_index": 5, "token_count": 1048, "text_raw": "The fundamental Clausius-Clapeyron equation gives the saturation vapor pressure $e_{sat}(T)$ (in hPa) at a given temperature $T$:\n\n$$ \ne_{sat}(T) = 6.11 \\times \\exp{ \\left[ -\\frac{L_v}{R_v} \\times \\left(\\frac{1}{T} - \\frac{1}{273.15}\\right) \\right]} \n$$\n\nwith $L_v = 2.5 \\ 10^6$ (J/kg) is the latent heat for vaporisation and $R_v = 461.51$ (J/kg/K) is the specific gas constant for water vapor. $e_{sat}(T)$ defines the capacity of the atmosphere to hold water vapor before it condensates, for a given air temperature $T$.\n\nAccording to this equation, it can be shown that the sensitivity of $e_{sat}(T)$ to a temperature increase is given by:\n\n$$ \n\\frac{ d (de_{sat}/ e_{sat})}{dT} = \\ \\frac{L_v}{R_v \\ T^2} \\ = \\ \\alpha(T) \n$$\n\nwith $\\alpha(T)$ standing for the Clausius-Clapeyron scaling, which depends strongly on $T$.\n\nMost water vapor resides in the lower troposphere and in the tropics because there temperatures are warmesest. For air temperature typical of the tropical lower atmosphere, ranging between 270 and 310K, $\\alpha(T)$ has values between 6.5 to 7.5\\% for a 1-K increase of air temperature (see [[8]](https://doi.org/10.1029/2019JD031017)).\n\nAs stated in numerous studies addressing the perturbation of the hydrological cycle to surface warming (see [[3]](https://doi.org/10.1175/1520-0469%281967%29024<0241:TEOTAW>2.0.CO;2), [[7]](https://doi.org/10.1002/2013JD020184), [[9]](https://www.ipcc.ch/report/ar6/wg1/downloads/report/IPCC_AR6_WGI_Chapter07.pdf) and [[10]](https://doi.org/abs/10.1029/1998JD900007)) tropospheric relative humidity tends to be maintained fixed as climate warms. The small changes observed in the tropospheric relative humidity do not affect significantly the column integrated water vapor [[11]](https://doi.org/10.1175/JCLI3990.1).\n\nFormally, the global climate feedback $f$ (in W/m$^2$/K) represents the sensitivity of the global energy balance, which considers longwave and shorwtave fluxes, to a perturbation of the surface temperature.\n\nHere we consider the clear sky longwave feedback and the water vapor contribution to this feedback, and this feeback parameter $f$ can be broken down into several components, among them the water vapor.\n\nFor the present analysis, we look at the part of the clear sky greenhouse effect $\\text{G}_{CSwv}$ due to water vapor. Then, and using the Clausius-Clapeyron scaling, the water vapor feedback $f_{wv}$ can be written as:\n\n$$\n\\begin{align}\nf_{wv} & = \\frac{d \\ \\text{G}_{CS wv}}{d \\ T_{surf}} \\\\\n & = \\frac{d \\ \\text{G}_{CS wv}}{d \\ TCWV} \\frac{d \\ TCWV wv}{d \\ T_{surf}}\n\\end{align}\n$$\n\nWhere $d \\ TCWV / d \\ T_{surf}$ represents the water vapor response to the temperature variations, while $d \\ \\text{G}_{CS w} / d \\ TCWV$ is the net change in upwelling thermal radiation caused by the associated changes in the water vapor content.\n\nThe TCWV is the integral of the water vapor amount q (in kg/kg) over the atmospheric column. Therefore, $TCWV \\sim q$.\nMoreover, with the approximation of a fixed atmospheric relative humidity with surface warming, then $ d q / \\overline{q} = d e_{sat}/ \\overline{e_{sat}}$, the notation $\\overline{X}$ denoting a spatial average of the parameter $X$.\n\nCombining all the parameters together, assuming constant RH and a 1-K surface warming, it comes that the water vapor feeback can be computed using TCWV according to:", "text_with_prefix": "EQC Quality Assessment: \"Water vapor amplification of Earth's Greenhouse Effect\"\nDataset: satellite-total-column-water-vapour-ocean [CDS]\nAspect: validation_q01 | Category: Satellite_ECVs\nSection: Water vapor amplification of Earth's Greenhouse Effect > Methodology > The water vapour feedback\n---\nThe fundamental Clausius-Clapeyron equation gives the saturation vapor pressure $e_{sat}(T)$ (in hPa) at a given temperature $T$:\n\n$$ \ne_{sat}(T) = 6.11 \\times \\exp{ \\left[ -\\frac{L_v}{R_v} \\times \\left(\\frac{1}{T} - \\frac{1}{273.15}\\right) \\right]} \n$$\n\nwith $L_v = 2.5 \\ 10^6$ (J/kg) is the latent heat for vaporisation and $R_v = 461.51$ (J/kg/K) is the specific gas constant for water vapor. $e_{sat}(T)$ defines the capacity of the atmosphere to hold water vapor before it condensates, for a given air temperature $T$.\n\nAccording to this equation, it can be shown that the sensitivity of $e_{sat}(T)$ to a temperature increase is given by:\n\n$$ \n\\frac{ d (de_{sat}/ e_{sat})}{dT} = \\ \\frac{L_v}{R_v \\ T^2} \\ = \\ \\alpha(T) \n$$\n\nwith $\\alpha(T)$ standing for the Clausius-Clapeyron scaling, which depends strongly on $T$.\n\nMost water vapor resides in the lower troposphere and in the tropics because there temperatures are warmesest. For air temperature typical of the tropical lower atmosphere, ranging between 270 and 310K, $\\alpha(T)$ has values between 6.5 to 7.5\\% for a 1-K increase of air temperature (see [[8]](https://doi.org/10.1029/2019JD031017)).\n\nAs stated in numerous studies addressing the perturbation of the hydrological cycle to surface warming (see [[3]](https://doi.org/10.1175/1520-0469%281967%29024<0241:TEOTAW>2.0.CO;2), [[7]](https://doi.org/10.1002/2013JD020184), [[9]](https://www.ipcc.ch/report/ar6/wg1/downloads/report/IPCC_AR6_WGI_Chapter07.pdf) and [[10]](https://doi.org/abs/10.1029/1998JD900007)) tropospheric relative humidity tends to be maintained fixed as climate warms. The small changes observed in the tropospheric relative humidity do not affect significantly the column integrated water vapor [[11]](https://doi.org/10.1175/JCLI3990.1).\n\nFormally, the global climate feedback $f$ (in W/m$^2$/K) represents the sensitivity of the global energy balance, which considers longwave and shorwtave fluxes, to a perturbation of the surface temperature.\n\nHere we consider the clear sky longwave feedback and the water vapor contribution to this feedback, and this feeback parameter $f$ can be broken down into several components, among them the water vapor.\n\nFor the present analysis, we look at the part of the clear sky greenhouse effect $\\text{G}_{CSwv}$ due to water vapor. Then, and using the Clausius-Clapeyron scaling, the water vapor feedback $f_{wv}$ can be written as:\n\n$$\n\\begin{align}\nf_{wv} & = \\frac{d \\ \\text{G}_{CS wv}}{d \\ T_{surf}} \\\\\n & = \\frac{d \\ \\text{G}_{CS wv}}{d \\ TCWV} \\frac{d \\ TCWV wv}{d \\ T_{surf}}\n\\end{align}\n$$\n\nWhere $d \\ TCWV / d \\ T_{surf}$ represents the water vapor response to the temperature variations, while $d \\ \\text{G}_{CS w} / d \\ TCWV$ is the net change in upwelling thermal radiation caused by the associated changes in the water vapor content.\n\nThe TCWV is the integral of the water vapor amount q (in kg/kg) over the atmospheric column. Therefore, $TCWV \\sim q$.\nMoreover, with the approximation of a fixed atmospheric relative humidity with surface warming, then $ d q / \\overline{q} = d e_{sat}/ \\overline{e_{sat}}$, the notation $\\overline{X}$ denoting a spatial average of the parameter $X$.\n\nCombining all the parameters together, assuming constant RH and a 1-K surface warming, it comes that the water vapor feeback can be computed using TCWV according to:"} {"chunk_id": "satellite_satellite-total-column-water-vapour-ocean_validation_q01__c70004be8134", "report_id": "satellite_satellite-total-column-water-vapour-ocean_validation_q01", "dataset_id": "satellite-total-column-water-vapour-ocean", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q01", "aspect_base": "validation", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Water vapor amplification of Earth's Greenhouse Effect > Methodology > The water vapour feedback", "title": "Water vapor amplification of Earth's Greenhouse Effect", "chunk_index": 6, "token_count": 164, "text_raw": "denoting a spatial average of the parameter $X$.\n\nCombining all the parameters together, assuming constant RH and a 1-K surface warming, it comes that the water vapor feeback can be computed using TCWV according to:\n\n$$ \nf_{wv} \\ = \\ \\frac{d \\ \\text{G}_{CS wv}}{d \\ TCWV} \\ \\overline{TCWV} \\ \\alpha(T)\n$$", "text_with_prefix": "EQC Quality Assessment: \"Water vapor amplification of Earth's Greenhouse Effect\"\nDataset: satellite-total-column-water-vapour-ocean [CDS]\nAspect: validation_q01 | Category: Satellite_ECVs\nSection: Water vapor amplification of Earth's Greenhouse Effect > Methodology > The water vapour feedback\n---\ndenoting a spatial average of the parameter $X$.\n\nCombining all the parameters together, assuming constant RH and a 1-K surface warming, it comes that the water vapor feeback can be computed using TCWV according to:\n\n$$ \nf_{wv} \\ = \\ \\frac{d \\ \\text{G}_{CS wv}}{d \\ TCWV} \\ \\overline{TCWV} \\ \\alpha(T)\n$$"} {"chunk_id": "satellite_satellite-total-column-water-vapour-ocean_validation_q01__320452b8709f", "report_id": "satellite_satellite-total-column-water-vapour-ocean_validation_q01", "dataset_id": "satellite-total-column-water-vapour-ocean", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q01", "aspect_base": "validation", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Water vapor amplification of Earth's Greenhouse Effect > Methodology > Method", "title": "Water vapor amplification of Earth's Greenhouse Effect", "chunk_index": 7, "token_count": 534, "text_raw": "The analysis comprises the following steps:\n\n__1. [](data-section-1)__\n - Import the relevant packages.\n - Define the parameters of the analysis and set the dataset requests\n\n__2. [](data-section-2)__\n - Download the variables of interest: Columnar water vapor is obtained from __Monthly and 6-hourly total column water vapour over ocean from 1988 to 2020 derived from satellite observations__ [[TCWV]](https://cds.climate.copernicus.eu/datasets/satellite-total-column-water-vapour-ocean), Clear sky is obtained from __Cloud properties global gridded monthly and daily data from 1979 to present derived from satellite observations__ [[clouds]](https://cds.climate.copernicus.eu/datasets/satellite-cloud-properties), Surface Temperature is obtained from __Sea surface temperature daily data from 1981 to present derived from satellite observations__ [[SST]](https://cds.climate.copernicus.eu/datasets/satellite-sea-surface-temperature), Outgoing Longwave Radiation is obtained from __Earth's radiation budget from 1979 to present derived from satellite observations__ [[OLR]](https://cds.climate.copernicus.eu/datasets/satellite-earth-radiation-budget).\n - The datasets are colocated in space and time over the Tropics.\n - Gridboxes with cloudy coverage greater than 10% are rejected. Moreover, only the well retrieved TCWV and SST are considered. The amount of radiation emitted by the surface is computed assuming blackbody behavior of the ocean ($\\epsilon_{ocean}=1$).\n - The clear sky greenhouse effect G$_\\text{CS}$ is obtained by subtracting outgoing lonwave radiation and surface emission longwave emission.\n\n__3. [](Analysis-section-1)__\n - Maps of TCWV, CFC, SST and OLR are described as well as of G$_\\text{CS}$.\n\n__4. [](Plot-and-results)__\n - The clear sky greenhouse effect is plotted as a function of the TCWV. The slope of the linear regression is computed and used to estimate the water vapor feedback $f_{wv}$. Final results are compared with proper references.", "text_with_prefix": "EQC Quality Assessment: \"Water vapor amplification of Earth's Greenhouse Effect\"\nDataset: satellite-total-column-water-vapour-ocean [CDS]\nAspect: validation_q01 | Category: Satellite_ECVs\nSection: Water vapor amplification of Earth's Greenhouse Effect > Methodology > Method\n---\nThe analysis comprises the following steps:\n\n__1. [](data-section-1)__\n - Import the relevant packages.\n - Define the parameters of the analysis and set the dataset requests\n\n__2. [](data-section-2)__\n - Download the variables of interest: Columnar water vapor is obtained from __Monthly and 6-hourly total column water vapour over ocean from 1988 to 2020 derived from satellite observations__ [[TCWV]](https://cds.climate.copernicus.eu/datasets/satellite-total-column-water-vapour-ocean), Clear sky is obtained from __Cloud properties global gridded monthly and daily data from 1979 to present derived from satellite observations__ [[clouds]](https://cds.climate.copernicus.eu/datasets/satellite-cloud-properties), Surface Temperature is obtained from __Sea surface temperature daily data from 1981 to present derived from satellite observations__ [[SST]](https://cds.climate.copernicus.eu/datasets/satellite-sea-surface-temperature), Outgoing Longwave Radiation is obtained from __Earth's radiation budget from 1979 to present derived from satellite observations__ [[OLR]](https://cds.climate.copernicus.eu/datasets/satellite-earth-radiation-budget).\n - The datasets are colocated in space and time over the Tropics.\n - Gridboxes with cloudy coverage greater than 10% are rejected. Moreover, only the well retrieved TCWV and SST are considered. The amount of radiation emitted by the surface is computed assuming blackbody behavior of the ocean ($\\epsilon_{ocean}=1$).\n - The clear sky greenhouse effect G$_\\text{CS}$ is obtained by subtracting outgoing lonwave radiation and surface emission longwave emission.\n\n__3. [](Analysis-section-1)__\n - Maps of TCWV, CFC, SST and OLR are described as well as of G$_\\text{CS}$.\n\n__4. [](Plot-and-results)__\n - The clear sky greenhouse effect is plotted as a function of the TCWV. The slope of the linear regression is computed and used to estimate the water vapor feedback $f_{wv}$. Final results are compared with proper references."} {"chunk_id": "satellite_satellite-total-column-water-vapour-ocean_validation_q01__3184d8654473", "report_id": "satellite_satellite-total-column-water-vapour-ocean_validation_q01", "dataset_id": "satellite-total-column-water-vapour-ocean", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q01", "aspect_base": "validation", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Water vapor amplification of Earth's Greenhouse Effect > Analysis and results > Choose the data to use and setup code > Define parameters", "title": "Water vapor amplification of Earth's Greenhouse Effect", "chunk_index": 8, "token_count": 116, "text_raw": "This use case is developed for one month of data (July 2007) and for the tropical belt, restricted to 30$^\\circ$S-30$^\\circ$N.", "text_with_prefix": "EQC Quality Assessment: \"Water vapor amplification of Earth's Greenhouse Effect\"\nDataset: satellite-total-column-water-vapour-ocean [CDS]\nAspect: validation_q01 | Category: Satellite_ECVs\nSection: Water vapor amplification of Earth's Greenhouse Effect > Analysis and results > Choose the data to use and setup code > Define parameters\n---\nThis use case is developed for one month of data (July 2007) and for the tropical belt, restricted to 30$^\\circ$S-30$^\\circ$N."} {"chunk_id": "satellite_satellite-total-column-water-vapour-ocean_validation_q01__c273144e5957", "report_id": "satellite_satellite-total-column-water-vapour-ocean_validation_q01", "dataset_id": "satellite-total-column-water-vapour-ocean", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q01", "aspect_base": "validation", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Water vapor amplification of Earth's Greenhouse Effect > Analysis and results > Choose the data to use and setup code > Set the data request", "title": "Water vapor amplification of Earth's Greenhouse Effect", "chunk_index": 9, "token_count": 125, "text_raw": "Four datasets are requested :\n- Total Column Water Vapor (TCWV)\n- Cloud Fraction Cover (CFC)\n- Sea Surface Temperature (SST)\n- Outgoing Longwave Radiation (OLR)\n\n(data-section-2)=", "text_with_prefix": "EQC Quality Assessment: \"Water vapor amplification of Earth's Greenhouse Effect\"\nDataset: satellite-total-column-water-vapour-ocean [CDS]\nAspect: validation_q01 | Category: Satellite_ECVs\nSection: Water vapor amplification of Earth's Greenhouse Effect > Analysis and results > Choose the data to use and setup code > Set the data request\n---\nFour datasets are requested :\n- Total Column Water Vapor (TCWV)\n- Cloud Fraction Cover (CFC)\n- Sea Surface Temperature (SST)\n- Outgoing Longwave Radiation (OLR)\n\n(data-section-2)="} {"chunk_id": "satellite_satellite-total-column-water-vapour-ocean_validation_q01__c153308c7a94", "report_id": "satellite_satellite-total-column-water-vapour-ocean_validation_q01", "dataset_id": "satellite-total-column-water-vapour-ocean", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q01", "aspect_base": "validation", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Water vapor amplification of Earth's Greenhouse Effect > Analysis and results > Download the datasets and perform space-time aggregation", "title": "Water vapor amplification of Earth's Greenhouse Effect", "chunk_index": 10, "token_count": 525, "text_raw": "The datasets are now downloaded.\n\n```text\n100%|██████████| 1/1 [00:00<00:00, 3.61it/s]\n/data/common/miniforge3/envs/wp5/lib/python3.11/site-packages/earthkit/data/readers/netcdf/fieldlist.py:318: FutureWarning: In a future version, xarray will not decode timedelta values based on the presence of a timedelta-like units attribute by default. Instead it will rely on the presence of a timedelta64 dtype attribute, which is now xarray's default way of encoding timedelta64 values. To continue decoding timedeltas based on the presence of a timedelta-like units attribute, users will need to explicitly opt-in by passing True or CFTimedeltaCoder(decode_via_units=True) to decode_timedelta. To silence this warning, set decode_timedelta to True, False, or a 'CFTimedeltaCoder' instance.\n return xr.open_dataset(self.path_or_url)\n/data/common/miniforge3/envs/wp5/lib/python3.11/site-packages/earthkit/data/readers/netcdf/fieldlist.py:202: FutureWarning: In a future version, xarray will not decode timedelta values based on the presence of a timedelta-like units attribute by default. Instead it will rely on the presence of a timedelta64 dtype attribute, which is now xarray's default way of encoding timedelta64 values. To continue decoding timedeltas based on the presence of a timedelta-like units attribute, users will need to explicitly opt-in by passing True or CFTimedeltaCoder(decode_via_units=True) to decode_timedelta. To silence this warning, set decode_timedelta to True, False, or a 'CFTimedeltaCoder' instance.\n return xr.open_mfdataset(\n100%|██████████| 1/1 [00:00<00:00, 3.92it/s]\n100%|██████████| 1/1 [00:00<00:00, 14.78it/s]\n100%|██████████| 1/1 [00:00<00:00, 1.62it/s]\n```", "text_with_prefix": "EQC Quality Assessment: \"Water vapor amplification of Earth's Greenhouse Effect\"\nDataset: satellite-total-column-water-vapour-ocean [CDS]\nAspect: validation_q01 | Category: Satellite_ECVs\nSection: Water vapor amplification of Earth's Greenhouse Effect > Analysis and results > Download the datasets and perform space-time aggregation\n---\nThe datasets are now downloaded.\n\n```text\n100%|██████████| 1/1 [00:00<00:00, 3.61it/s]\n/data/common/miniforge3/envs/wp5/lib/python3.11/site-packages/earthkit/data/readers/netcdf/fieldlist.py:318: FutureWarning: In a future version, xarray will not decode timedelta values based on the presence of a timedelta-like units attribute by default. Instead it will rely on the presence of a timedelta64 dtype attribute, which is now xarray's default way of encoding timedelta64 values. To continue decoding timedeltas based on the presence of a timedelta-like units attribute, users will need to explicitly opt-in by passing True or CFTimedeltaCoder(decode_via_units=True) to decode_timedelta. To silence this warning, set decode_timedelta to True, False, or a 'CFTimedeltaCoder' instance.\n return xr.open_dataset(self.path_or_url)\n/data/common/miniforge3/envs/wp5/lib/python3.11/site-packages/earthkit/data/readers/netcdf/fieldlist.py:202: FutureWarning: In a future version, xarray will not decode timedelta values based on the presence of a timedelta-like units attribute by default. Instead it will rely on the presence of a timedelta64 dtype attribute, which is now xarray's default way of encoding timedelta64 values. To continue decoding timedeltas based on the presence of a timedelta-like units attribute, users will need to explicitly opt-in by passing True or CFTimedeltaCoder(decode_via_units=True) to decode_timedelta. To silence this warning, set decode_timedelta to True, False, or a 'CFTimedeltaCoder' instance.\n return xr.open_mfdataset(\n100%|██████████| 1/1 [00:00<00:00, 3.92it/s]\n100%|██████████| 1/1 [00:00<00:00, 14.78it/s]\n100%|██████████| 1/1 [00:00<00:00, 1.62it/s]\n```"} {"chunk_id": "satellite_satellite-total-column-water-vapour-ocean_validation_q01__16546f40c4df", "report_id": "satellite_satellite-total-column-water-vapour-ocean_validation_q01", "dataset_id": "satellite-total-column-water-vapour-ocean", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q01", "aspect_base": "validation", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Water vapor amplification of Earth's Greenhouse Effect > Analysis and results > Download the datasets and perform space-time aggregation > Temporal and spatial aggregation", "title": "Water vapor amplification of Earth's Greenhouse Effect", "chunk_index": 11, "token_count": 276, "text_raw": "All data records are provided at different spatial and temporal resolutions : \n - Cloud Fraction Cover : 0.25$^\\circ$ / 1-day\n - Sea Surface Temperature : 0.05$^\\circ$ / 1-day\n - Outgoing Longwave Radiation : 0.25$^\\circ$ / 1-day\n - Total Column Water Vapor : 0.5$^\\circ$ / 6-hourly\n\nThis it is a cross-variable analysis, the study requires to process all the datasets at the same horizontal and temporal resolutions.\n\nNote that for the TCWV data record, a reversal of the latitudes is necessary as the data is stored North-to-South, while for the others data records the storage is South-to-North.\n\nThe common spatial aggregation is 0.5$^\\circ$ and the data records are processed at the daily and monthly scales.\n\ncoarsen datasets to 0.5° grid / 1-day", "text_with_prefix": "EQC Quality Assessment: \"Water vapor amplification of Earth's Greenhouse Effect\"\nDataset: satellite-total-column-water-vapour-ocean [CDS]\nAspect: validation_q01 | Category: Satellite_ECVs\nSection: Water vapor amplification of Earth's Greenhouse Effect > Analysis and results > Download the datasets and perform space-time aggregation > Temporal and spatial aggregation\n---\nAll data records are provided at different spatial and temporal resolutions : \n - Cloud Fraction Cover : 0.25$^\\circ$ / 1-day\n - Sea Surface Temperature : 0.05$^\\circ$ / 1-day\n - Outgoing Longwave Radiation : 0.25$^\\circ$ / 1-day\n - Total Column Water Vapor : 0.5$^\\circ$ / 6-hourly\n\nThis it is a cross-variable analysis, the study requires to process all the datasets at the same horizontal and temporal resolutions.\n\nNote that for the TCWV data record, a reversal of the latitudes is necessary as the data is stored North-to-South, while for the others data records the storage is South-to-North.\n\nThe common spatial aggregation is 0.5$^\\circ$ and the data records are processed at the daily and monthly scales.\n\ncoarsen datasets to 0.5° grid / 1-day"} {"chunk_id": "satellite_satellite-total-column-water-vapour-ocean_validation_q01__ac898c9ab56d", "report_id": "satellite_satellite-total-column-water-vapour-ocean_validation_q01", "dataset_id": "satellite-total-column-water-vapour-ocean", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q01", "aspect_base": "validation", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Water vapor amplification of Earth's Greenhouse Effect > Analysis and results > Download the datasets and perform space-time aggregation > Computation of the greenhouse effect", "title": "Water vapor amplification of Earth's Greenhouse Effect", "chunk_index": 12, "token_count": 452, "text_raw": "The variables (daylies and monthlies) are stored into data frames for a better handling.\n\nThen the clear sky scenes are selected using the CFC. A threshold of 10\\% of cloud cover within each grid cell is a good compromise between the definition of the clear sky greenhouse effect (mostly clear skies, regardless of cloud altitude) and the number of scenes that are retained for the representativity of the analysis. The grid cells are also filtered from erroneous estimates.\n\nThe clear sky greenhouse effect (G$_{CS}$) and the water vapor feedback paramter ($f_{wv}$) are then computed using the methods described above.\n\nMerge datasets\nCompute clear sky greenhouse effect and add to the merged dataset\nSelection of clear sky regions (at most 10\\% of cloud fraction within each grid cell) and quality of retrieved SST and TCVW (SST > 0°K and TCWV > 0 mm)\nCompute montly mean from daily & masked data\n\n```text\n/data/common/miniforge3/envs/wp5/lib/python3.11/site-packages/xarray/coding/times.py:650: RuntimeWarning: invalid value encountered in cast\n flat_num = flat_num.astype(np.int64)\n/data/common/miniforge3/envs/wp5/lib/python3.11/site-packages/xarray/coding/times.py:650: RuntimeWarning: invalid value encountered in cast\n flat_num = flat_num.astype(np.int64)\n/data/common/miniforge3/envs/wp5/lib/python3.11/site-packages/xarray/coding/times.py:650: RuntimeWarning: invalid value encountered in cast\n flat_num = flat_num.astype(np.int64)\n```\n\nConvert the 3D dataset into a dataframe for the analysis\ndaily dataset\nmonthly dataset\n\n(Analysis-section-1)=", "text_with_prefix": "EQC Quality Assessment: \"Water vapor amplification of Earth's Greenhouse Effect\"\nDataset: satellite-total-column-water-vapour-ocean [CDS]\nAspect: validation_q01 | Category: Satellite_ECVs\nSection: Water vapor amplification of Earth's Greenhouse Effect > Analysis and results > Download the datasets and perform space-time aggregation > Computation of the greenhouse effect\n---\nThe variables (daylies and monthlies) are stored into data frames for a better handling.\n\nThen the clear sky scenes are selected using the CFC. A threshold of 10\\% of cloud cover within each grid cell is a good compromise between the definition of the clear sky greenhouse effect (mostly clear skies, regardless of cloud altitude) and the number of scenes that are retained for the representativity of the analysis. The grid cells are also filtered from erroneous estimates.\n\nThe clear sky greenhouse effect (G$_{CS}$) and the water vapor feedback paramter ($f_{wv}$) are then computed using the methods described above.\n\nMerge datasets\nCompute clear sky greenhouse effect and add to the merged dataset\nSelection of clear sky regions (at most 10\\% of cloud fraction within each grid cell) and quality of retrieved SST and TCVW (SST > 0°K and TCWV > 0 mm)\nCompute montly mean from daily & masked data\n\n```text\n/data/common/miniforge3/envs/wp5/lib/python3.11/site-packages/xarray/coding/times.py:650: RuntimeWarning: invalid value encountered in cast\n flat_num = flat_num.astype(np.int64)\n/data/common/miniforge3/envs/wp5/lib/python3.11/site-packages/xarray/coding/times.py:650: RuntimeWarning: invalid value encountered in cast\n flat_num = flat_num.astype(np.int64)\n/data/common/miniforge3/envs/wp5/lib/python3.11/site-packages/xarray/coding/times.py:650: RuntimeWarning: invalid value encountered in cast\n flat_num = flat_num.astype(np.int64)\n```\n\nConvert the 3D dataset into a dataframe for the analysis\ndaily dataset\nmonthly dataset\n\n(Analysis-section-1)="} {"chunk_id": "satellite_satellite-total-column-water-vapour-ocean_validation_q01__9e72cff0038d", "report_id": "satellite_satellite-total-column-water-vapour-ocean_validation_q01", "dataset_id": "satellite-total-column-water-vapour-ocean", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q01", "aspect_base": "validation", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Water vapor amplification of Earth's Greenhouse Effect > Analysis and results > Overview of the geophysical variables of interest > Total Column of Water Vapour and Cloud Fraction Cover", "title": "Water vapor amplification of Earth's Greenhouse Effect", "chunk_index": 13, "token_count": 426, "text_raw": "The Total Column of Water Vapour (TCWV) is obtained from the HOAPS product, provided by EUMETSAT, at 0.5° resolution. The TCWV is retrieved over ice-free ocean surfaces using the microwave spectrum as measured by the instruments SSM/I and SSMIS using the Hamburg Ocean Atmosphere Parameters and Fluxes from Satellite Data (HOAPS) algorithm [[12]](https://essd.copernicus.org/articles/2/215/2010/essd-2-215-2010.html).\n\nThe Cloud Fraction Cover (CFC) is obtained from the CLARA-A3 product, provided by EUMETSAT at 0.25° resolution. The CFC is obtained over all surfaces using AVHRR measurements [[13]](https://doi.org/10.5676/EUM_SAF_CM/CLARA_AVHRR/V003).\n\nThe TCWV dataset in this assessment is considered at 0.5° over tropical oceans at a 6-hourly resolution.\n\nThe maps below show the TCWV (averaged over the 6-hourly time steps of the day) and the CFC for July 1st 2007.\n\nFrom these two maps, we can see that the highest values of TCWV are associated to cloudy situations and more specifically to storms. Some missing values are visible within areas of high TCWV, and those are where precipitation occurs in the cloudy areas for which no TCWV is retrieved [[12]](https://essd.copernicus.org/articles/2/215/2010/essd-2-215-2010.html).", "text_with_prefix": "EQC Quality Assessment: \"Water vapor amplification of Earth's Greenhouse Effect\"\nDataset: satellite-total-column-water-vapour-ocean [CDS]\nAspect: validation_q01 | Category: Satellite_ECVs\nSection: Water vapor amplification of Earth's Greenhouse Effect > Analysis and results > Overview of the geophysical variables of interest > Total Column of Water Vapour and Cloud Fraction Cover\n---\nThe Total Column of Water Vapour (TCWV) is obtained from the HOAPS product, provided by EUMETSAT, at 0.5° resolution. The TCWV is retrieved over ice-free ocean surfaces using the microwave spectrum as measured by the instruments SSM/I and SSMIS using the Hamburg Ocean Atmosphere Parameters and Fluxes from Satellite Data (HOAPS) algorithm [[12]](https://essd.copernicus.org/articles/2/215/2010/essd-2-215-2010.html).\n\nThe Cloud Fraction Cover (CFC) is obtained from the CLARA-A3 product, provided by EUMETSAT at 0.25° resolution. The CFC is obtained over all surfaces using AVHRR measurements [[13]](https://doi.org/10.5676/EUM_SAF_CM/CLARA_AVHRR/V003).\n\nThe TCWV dataset in this assessment is considered at 0.5° over tropical oceans at a 6-hourly resolution.\n\nThe maps below show the TCWV (averaged over the 6-hourly time steps of the day) and the CFC for July 1st 2007.\n\nFrom these two maps, we can see that the highest values of TCWV are associated to cloudy situations and more specifically to storms. Some missing values are visible within areas of high TCWV, and those are where precipitation occurs in the cloudy areas for which no TCWV is retrieved [[12]](https://essd.copernicus.org/articles/2/215/2010/essd-2-215-2010.html)."} {"chunk_id": "satellite_satellite-total-column-water-vapour-ocean_validation_q01__9d9fe9468519", "report_id": "satellite_satellite-total-column-water-vapour-ocean_validation_q01", "dataset_id": "satellite-total-column-water-vapour-ocean", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q01", "aspect_base": "validation", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Water vapor amplification of Earth's Greenhouse Effect > Analysis and results > Overview of the geophysical variables of interest > Sea Surface Temperature and the Outgoing Longwave Radiation", "title": "Water vapor amplification of Earth's Greenhouse Effect", "chunk_index": 14, "token_count": 469, "text_raw": "The daily estimates of Sea Surface Temperature (SST) are based on observations from multiple satellite sensors, and it is produced by the European Space Agency (ESA) SST Climate Change Initiative (CCI) project [[14]](https://doi.org/10.1038/s41597-019-0236-x).\nThis Climate Data Record cover the entire globe with a resolution of 0.05°.\n\nThe Outgoing Longwave Radiation (OLR) is obtained from the CLARA-A3 product, provided by EUMETSAT at 0.25° resolution. The OLR is obtained over all surfaces using AVHRR measurements [[13]](https://doi.org/10.5676/EUM_SAF_CM/CLARA_AVHRR/V003).\n\nThe maps below show the SST and the OLR for July 1st 2007.\n\nThe patterns of SST are similar to the patterns of TCWV, and this is mostly due to the Clausius-Clapeyron relationship recalled in [](method-section-2) : as surface temperature is higher, there is more evaporation and thus more water vapor in the atmosphere.\n\nThe OLR is a combination of several parameters that emit longwave radiation to space : the surface, the atmospheric gases and the clouds. \n- Minima of OLR correspond to high-altitude and optically thick clouds that can be also found from the CFC map above (although no cloud altitude is provided). The minima are thus linked to the major convective areas of the tropics: Central America, the Sahelian Africa band, India and the Indonesian Warm Pool.\n- Maxima of OLR are located over cloud-free regions, and is a combination of the surface (SST map) and the cloud-free atmosphere above (TCWV map). The maxima are found on both sides (North and South) of the convective areas.", "text_with_prefix": "EQC Quality Assessment: \"Water vapor amplification of Earth's Greenhouse Effect\"\nDataset: satellite-total-column-water-vapour-ocean [CDS]\nAspect: validation_q01 | Category: Satellite_ECVs\nSection: Water vapor amplification of Earth's Greenhouse Effect > Analysis and results > Overview of the geophysical variables of interest > Sea Surface Temperature and the Outgoing Longwave Radiation\n---\nThe daily estimates of Sea Surface Temperature (SST) are based on observations from multiple satellite sensors, and it is produced by the European Space Agency (ESA) SST Climate Change Initiative (CCI) project [[14]](https://doi.org/10.1038/s41597-019-0236-x).\nThis Climate Data Record cover the entire globe with a resolution of 0.05°.\n\nThe Outgoing Longwave Radiation (OLR) is obtained from the CLARA-A3 product, provided by EUMETSAT at 0.25° resolution. The OLR is obtained over all surfaces using AVHRR measurements [[13]](https://doi.org/10.5676/EUM_SAF_CM/CLARA_AVHRR/V003).\n\nThe maps below show the SST and the OLR for July 1st 2007.\n\nThe patterns of SST are similar to the patterns of TCWV, and this is mostly due to the Clausius-Clapeyron relationship recalled in [](method-section-2) : as surface temperature is higher, there is more evaporation and thus more water vapor in the atmosphere.\n\nThe OLR is a combination of several parameters that emit longwave radiation to space : the surface, the atmospheric gases and the clouds. \n- Minima of OLR correspond to high-altitude and optically thick clouds that can be also found from the CFC map above (although no cloud altitude is provided). The minima are thus linked to the major convective areas of the tropics: Central America, the Sahelian Africa band, India and the Indonesian Warm Pool.\n- Maxima of OLR are located over cloud-free regions, and is a combination of the surface (SST map) and the cloud-free atmosphere above (TCWV map). The maxima are found on both sides (North and South) of the convective areas."} {"chunk_id": "satellite_satellite-total-column-water-vapour-ocean_validation_q01__1311808fc0f9", "report_id": "satellite_satellite-total-column-water-vapour-ocean_validation_q01", "dataset_id": "satellite-total-column-water-vapour-ocean", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q01", "aspect_base": "validation", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Water vapor amplification of Earth's Greenhouse Effect > Analysis and results > Overview of the geophysical variables of interest > The monthly Clear Sky Greenhouse Effect", "title": "Water vapor amplification of Earth's Greenhouse Effect", "chunk_index": 15, "token_count": 170, "text_raw": "The maps below show the clear sky greenhouse effect G$_\\text{CS}$ for July 2007 and the corresponding map of column water vapor TCWV.\n\nAs expected, the largest values of G$_\\text{CS}$ are obtained where the TCWV is highest: as the amount of water vapour in the atmospheric column is high, there is a high trapping of upwelling thermal radiation.\n\n(Plot-and-results)=", "text_with_prefix": "EQC Quality Assessment: \"Water vapor amplification of Earth's Greenhouse Effect\"\nDataset: satellite-total-column-water-vapour-ocean [CDS]\nAspect: validation_q01 | Category: Satellite_ECVs\nSection: Water vapor amplification of Earth's Greenhouse Effect > Analysis and results > Overview of the geophysical variables of interest > The monthly Clear Sky Greenhouse Effect\n---\nThe maps below show the clear sky greenhouse effect G$_\\text{CS}$ for July 2007 and the corresponding map of column water vapor TCWV.\n\nAs expected, the largest values of G$_\\text{CS}$ are obtained where the TCWV is highest: as the amount of water vapour in the atmospheric column is high, there is a high trapping of upwelling thermal radiation.\n\n(Plot-and-results)="} {"chunk_id": "satellite_satellite-total-column-water-vapour-ocean_validation_q01__7dbdb4aea742", "report_id": "satellite_satellite-total-column-water-vapour-ocean_validation_q01", "dataset_id": "satellite-total-column-water-vapour-ocean", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q01", "aspect_base": "validation", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Water vapor amplification of Earth's Greenhouse Effect > Analysis and results > Plot and description of the results > Scatter plots: comparison of daily and monthly scales", "title": "Water vapor amplification of Earth's Greenhouse Effect", "chunk_index": 16, "token_count": 1045, "text_raw": "Before evaluating the contribution of water vapor to the clear sky greenhouse effect, a sanity check is done by comparing the rate of change of G$_\\text{CS}$ for a given change in sea surface temperature. This sanity check allows to compare with reference studies using satellite data.\n\nPlot linear fit\nPlot linear fit\n\n*Figure 1: Distributions of the clear sky greenhouse effect G$_{CS}$ according to SST for (a) daily data and (b) monthly data. The colorscales represent the density of points. The red dashed line are the linear regressions, with the slope and its uncertainty reported on the top left of each figures.*\n\nThe slopes of the regression lines slightly differ at the two scales: 5.94 W/m$^2$/K for daily data (Fig. 1a) and 5.86 W/m$^2$/K for monthly data (Fig. 1b).\n\nDespite these slight differences, the value of the slopes $d \\ G_{CS}/d T_{surf}$ are consistent with the findings of [[15]](https://doi.org/10.1175/JCLI3611.1), [[16]](https://doi.org/10.1038/351027a0) and [[17]](https://doi.org/epdf/10.1029/92JD00729), with reported values for given months ranging from 5.6 W/m$^2$/K to 6.3 W/m$^2$/K (monthly averages).\n\nThe contribution by water vapor to this sensitivity is estimated by looking at the linear regression $d \\ G_{CS}/d \\ TCWV$, and the slope of this regression is estimated using daily data and monthly data to have a sensitivity of this value to the temporal scale.\n\nPlot linear fit\nPlot linear fit\n\n*Figures 2 (top) and 4 (bottom) : Density diagrams of the clear sky greenhouse effect G$_{CS}$ according to bins of TCWV for daily data (top) and monthly data (bottom). The colorscales represent the density of points. The red dashed line are the linear regressions, with the slope and its uncertainty reported on the bottom left of each figures.*\n\nThe results show a linear relationship between the clear sky greenhouse effect and the Total Columns of Water Vapour, in agreement with previous research (see [[5]](https://doi.org/10.1073/pnas.1809868115), [[6]](https://doi.org/full/10.1029/2000JD000040) and [[7]](https://doi.org/10.1002/2013JD020184)).\n\nAccording to the section '[](method-section-2)' in the __Methodology__ Chapter, those results can be used to estimate $f_{wv}$:\n\n$$\n\\begin{align}\n\\frac{d \\ \\text{G}_{CS wv}}{d \\ TCWV} \\ & \\ \\text{is estimated from the slope of the linear regression G$_{CS}$=f(TCWV) in W/m$^2$/mm } \\\\ \n\\overline{TCWV} & \\ \\text{is the areal average of TCWV over the tropical belt in mm} \\\\\nCC & = 0.07 \\ \\text{K$^{-1}$} \\ \\ \\text{is the Clausius-Clapeyron rate of 7\\%/K}\n\\end{align}\n$$\n\nAs stated in the [](Method-section-1), the contribution of water vapor to the clear sky greenhouse effect is estimated to be $\\sim$ 60\\%. Therefore, we can reasonably consider that 60\\% of the slope G$_{CS}$=f(TCWV) corresponds to the water vapor contribution $\\text{G}_{CS wv}$\n\nVariables to estimate the water vapor feedback, from the equation in the Methodology section\n\n```text\nslope: 1.75 W/m^2/mm\ntropical mean TCWV: 33.81 mm\ntropical mean SST: 299.69 K\nfeedback: 2.48 W/m^2/K\n```\n\nThis result falls within the range of uncertainty of the water vapor feedback reported in the IPCC report Chapter 7 ([[9]](https://www.ipcc.ch/report/ar6/wg1/downloads/report/IPCC_AR6_WGI_Chapter07.pdf)) and summarized in Figure 7.20 (inset on the 'Global warming contributions from individual radiative feedbacks'). The reported values range between 2.2 and 3.5 W/m$^2$/K.", "text_with_prefix": "EQC Quality Assessment: \"Water vapor amplification of Earth's Greenhouse Effect\"\nDataset: satellite-total-column-water-vapour-ocean [CDS]\nAspect: validation_q01 | Category: Satellite_ECVs\nSection: Water vapor amplification of Earth's Greenhouse Effect > Analysis and results > Plot and description of the results > Scatter plots: comparison of daily and monthly scales\n---\nBefore evaluating the contribution of water vapor to the clear sky greenhouse effect, a sanity check is done by comparing the rate of change of G$_\\text{CS}$ for a given change in sea surface temperature. This sanity check allows to compare with reference studies using satellite data.\n\nPlot linear fit\nPlot linear fit\n\n*Figure 1: Distributions of the clear sky greenhouse effect G$_{CS}$ according to SST for (a) daily data and (b) monthly data. The colorscales represent the density of points. The red dashed line are the linear regressions, with the slope and its uncertainty reported on the top left of each figures.*\n\nThe slopes of the regression lines slightly differ at the two scales: 5.94 W/m$^2$/K for daily data (Fig. 1a) and 5.86 W/m$^2$/K for monthly data (Fig. 1b).\n\nDespite these slight differences, the value of the slopes $d \\ G_{CS}/d T_{surf}$ are consistent with the findings of [[15]](https://doi.org/10.1175/JCLI3611.1), [[16]](https://doi.org/10.1038/351027a0) and [[17]](https://doi.org/epdf/10.1029/92JD00729), with reported values for given months ranging from 5.6 W/m$^2$/K to 6.3 W/m$^2$/K (monthly averages).\n\nThe contribution by water vapor to this sensitivity is estimated by looking at the linear regression $d \\ G_{CS}/d \\ TCWV$, and the slope of this regression is estimated using daily data and monthly data to have a sensitivity of this value to the temporal scale.\n\nPlot linear fit\nPlot linear fit\n\n*Figures 2 (top) and 4 (bottom) : Density diagrams of the clear sky greenhouse effect G$_{CS}$ according to bins of TCWV for daily data (top) and monthly data (bottom). The colorscales represent the density of points. The red dashed line are the linear regressions, with the slope and its uncertainty reported on the bottom left of each figures.*\n\nThe results show a linear relationship between the clear sky greenhouse effect and the Total Columns of Water Vapour, in agreement with previous research (see [[5]](https://doi.org/10.1073/pnas.1809868115), [[6]](https://doi.org/full/10.1029/2000JD000040) and [[7]](https://doi.org/10.1002/2013JD020184)).\n\nAccording to the section '[](method-section-2)' in the __Methodology__ Chapter, those results can be used to estimate $f_{wv}$:\n\n$$\n\\begin{align}\n\\frac{d \\ \\text{G}_{CS wv}}{d \\ TCWV} \\ & \\ \\text{is estimated from the slope of the linear regression G$_{CS}$=f(TCWV) in W/m$^2$/mm } \\\\ \n\\overline{TCWV} & \\ \\text{is the areal average of TCWV over the tropical belt in mm} \\\\\nCC & = 0.07 \\ \\text{K$^{-1}$} \\ \\ \\text{is the Clausius-Clapeyron rate of 7\\%/K}\n\\end{align}\n$$\n\nAs stated in the [](Method-section-1), the contribution of water vapor to the clear sky greenhouse effect is estimated to be $\\sim$ 60\\%. Therefore, we can reasonably consider that 60\\% of the slope G$_{CS}$=f(TCWV) corresponds to the water vapor contribution $\\text{G}_{CS wv}$\n\nVariables to estimate the water vapor feedback, from the equation in the Methodology section\n\n```text\nslope: 1.75 W/m^2/mm\ntropical mean TCWV: 33.81 mm\ntropical mean SST: 299.69 K\nfeedback: 2.48 W/m^2/K\n```\n\nThis result falls within the range of uncertainty of the water vapor feedback reported in the IPCC report Chapter 7 ([[9]](https://www.ipcc.ch/report/ar6/wg1/downloads/report/IPCC_AR6_WGI_Chapter07.pdf)) and summarized in Figure 7.20 (inset on the 'Global warming contributions from individual radiative feedbacks'). The reported values range between 2.2 and 3.5 W/m$^2$/K."} {"chunk_id": "satellite_satellite-total-column-water-vapour-ocean_validation_q01__7d0e0bc6bc65", "report_id": "satellite_satellite-total-column-water-vapour-ocean_validation_q01", "dataset_id": "satellite-total-column-water-vapour-ocean", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q01", "aspect_base": "validation", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Water vapor amplification of Earth's Greenhouse Effect > Analysis and results > Plot and description of the results > Scatter plots: comparison of daily and monthly scales", "title": "Water vapor amplification of Earth's Greenhouse Effect", "chunk_index": 17, "token_count": 198, "text_raw": ".pdf)) and summarized in Figure 7.20 (inset on the 'Global warming contributions from individual radiative feedbacks'). The reported values range between 2.2 and 3.5 W/m$^2$/K.\n\nHere we have looked at the relationship between clear sky greenhouse effect and TCWV is studied on the time scale of one month. However, although the findings are consistent with the literature, climate feedbacks are generally estimated at longer time scale in order to take into account the role of natural climate variability (including El Nino events for instance).", "text_with_prefix": "EQC Quality Assessment: \"Water vapor amplification of Earth's Greenhouse Effect\"\nDataset: satellite-total-column-water-vapour-ocean [CDS]\nAspect: validation_q01 | Category: Satellite_ECVs\nSection: Water vapor amplification of Earth's Greenhouse Effect > Analysis and results > Plot and description of the results > Scatter plots: comparison of daily and monthly scales\n---\n.pdf)) and summarized in Figure 7.20 (inset on the 'Global warming contributions from individual radiative feedbacks'). The reported values range between 2.2 and 3.5 W/m$^2$/K.\n\nHere we have looked at the relationship between clear sky greenhouse effect and TCWV is studied on the time scale of one month. However, although the findings are consistent with the literature, climate feedbacks are generally estimated at longer time scale in order to take into account the role of natural climate variability (including El Nino events for instance)."} {"chunk_id": "satellite_satellite-total-column-water-vapour-ocean_validation_q01__d2263b2ec918", "report_id": "satellite_satellite-total-column-water-vapour-ocean_validation_q01", "dataset_id": "satellite-total-column-water-vapour-ocean", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q01", "aspect_base": "validation", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Water vapor amplification of Earth's Greenhouse Effect > ℹ️ If you want to know more > Key resources", "title": "Water vapor amplification of Earth's Greenhouse Effect", "chunk_index": 18, "token_count": 402, "text_raw": "Some key resources and further reading were linked throughout this assessment.\n\nThe CDS catalogue entries for the data used in this assessment are:\n- Monthly and 6-hourly total column water vapour over ocean from 1988 to 2020 derived from satellite observations:\nhttps://cds.climate.copernicus.eu/datasets/satellite-total-column-water-vapour-ocean\n\n- Cloud properties global gridded monthly and daily data from 1979 to present derived from satellite observations:\nhttps://cds.climate.copernicus.eu/datasets/satellite-cloud-properties\n\n- Sea surface temperature daily data from 1981 to present derived from satellite observations:\nhttps://cds.climate.copernicus.eu/datasets/satellite-sea-surface-temperature\n\n- Earth's radiation budget from 1979 to present derived from satellite observations:\nhttps://cds.climate.copernicus.eu/datasets/satellite-earth-radiation-budget\n\nCode libraries used:\n- [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)\n\nAnd some resources on the greenhouse effect of water vapor and its feedback :\n- https://science.nasa.gov/earth/climate-change/steamy-relationships-how-atmospheric-water-vapor-amplifies-earths-greenhouse-effect/\n\n- https://wmo.int/media/magazine-article/observing-water-vapour", "text_with_prefix": "EQC Quality Assessment: \"Water vapor amplification of Earth's Greenhouse Effect\"\nDataset: satellite-total-column-water-vapour-ocean [CDS]\nAspect: validation_q01 | Category: Satellite_ECVs\nSection: Water vapor amplification of Earth's Greenhouse Effect > ℹ️ If you want to know more > Key resources\n---\nSome key resources and further reading were linked throughout this assessment.\n\nThe CDS catalogue entries for the data used in this assessment are:\n- Monthly and 6-hourly total column water vapour over ocean from 1988 to 2020 derived from satellite observations:\nhttps://cds.climate.copernicus.eu/datasets/satellite-total-column-water-vapour-ocean\n\n- Cloud properties global gridded monthly and daily data from 1979 to present derived from satellite observations:\nhttps://cds.climate.copernicus.eu/datasets/satellite-cloud-properties\n\n- Sea surface temperature daily data from 1981 to present derived from satellite observations:\nhttps://cds.climate.copernicus.eu/datasets/satellite-sea-surface-temperature\n\n- Earth's radiation budget from 1979 to present derived from satellite observations:\nhttps://cds.climate.copernicus.eu/datasets/satellite-earth-radiation-budget\n\nCode libraries used:\n- [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)\n\nAnd some resources on the greenhouse effect of water vapor and its feedback :\n- https://science.nasa.gov/earth/climate-change/steamy-relationships-how-atmospheric-water-vapor-amplifies-earths-greenhouse-effect/\n\n- https://wmo.int/media/magazine-article/observing-water-vapour"} {"chunk_id": "satellite_satellite-total-column-water-vapour-ocean_validation_q01__3e2b038360a5", "report_id": "satellite_satellite-total-column-water-vapour-ocean_validation_q01", "dataset_id": "satellite-total-column-water-vapour-ocean", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q01", "aspect_base": "validation", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Water vapor amplification of Earth's Greenhouse Effect > ℹ️ If you want to know more > References", "title": "Water vapor amplification of Earth's Greenhouse Effect", "chunk_index": 19, "token_count": 1007, "text_raw": "[[1]](https://doi.org/10.1029/2010JD014287) Schmidt G. A., Ruedy R. A., Miller R. L., and A. A. Lacis (2010), Attribution of the present‐day total greenhouse effect,J. Geophys. Res.,115, D20106, doi:10.1029/2010JD014287\n\n[[2]](https://doi.org/10.1175/1520-0477%281997%29078<0197:EAGMEB>2.0.CO;2) Kiehl J. T. and K. Trenberth (1997), Earth's Annual Global Mean Energy Budget, Bull. Am. Meteorol. Soc., 78, 197-208, https://doi.org/10.1175/1520-0477%281997%29078<0197:EAGMEB>2.0.CO;2\n\n[[3]](https://doi.org/10.1175/1520-0469%281967%29024<0241:TEOTAW>2.0.CO;2) Manabe S., and R. T. Wetherald (1967), Thermal Equilibrium of the Atmosphere with a Given Distribution of Relative Humidity, J. Atmos. Sci., 24, 241–259, https://doi.org/10.1175/1520-0469%281967%29024<0241:TEOTAW>2.0.CO;2\n\n[[4]](https://doi.org/10.1146/annurev.energy.25.1.441) Held I. and B. Soden (2000), Water vapor feedback and global warming, Annu. Rev. Energy Environn., 25:441-75, doi:10.1146/annurev.energy.25.1.441\n\n[[5]](https://doi.org/10.1073/pnas.1809868115), Koll D. and T. Cronin (2018) Earth’s outgoing longwave radiation linear due to H2O greenhouse effect, Proc. Nat. Ac. Sci., 115 (41) 10293-10298, doi:10.1073/pnas.1809868115\n\n[[6]](https://doi.org/full/10.1029/2000JD000040) Roca R., Viollier M., Picon L., and M. Desbois (2002), A multisatellite analysis of deep convection and its moist environment over the Indian Ocean during the winter monsoon, J. Geophys. Res., 107(D19), doi:10.1029/2000JD000040.\n\n[[7]](https://doi.org/10.1002/2013JD020184) Gordon N., A. Jonko, P. Forster and K. Shell (2013), An observationally based constraint on the water-vapor feedback, J. Geophys. Res., 118, 12,435-12,443, doi:10.1002/2013JD020184.\n\n[[8]](https://doi.org/10.1029/2019JD031017) Raghuraman S. P., D. Paynter and V. Ramaswamy (2019), Quantifying the drivers of the clear sky greenhouse effect, 2000-2016, J. Geophys. Res, 124, 11,354-11,371, https://doi.org/10.1029/2019JD031017\n\n[[9]](https://www.ipcc.ch/report/ar6/wg1/downloads/report/IPCC_AR6_WGI_Chapter07.pdf) Forster P., Storelvmo T., Armour K., Collins W., Dufresne J-L, Frame D., Lunt D.J., Mauritsen T., Palmer M.D., Watanabe M., Wild M., and H. Zhang (2021), Chapter 7 The Earth’s Energy Budget, Climate Feedbacks, and Climate Sensitivity. In Climate Change 2021: The Physical Science Basis. Contribution of Working Group I to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, pp. 923–1054, doi: 10.1017/9781009157896.009.", "text_with_prefix": "EQC Quality Assessment: \"Water vapor amplification of Earth's Greenhouse Effect\"\nDataset: satellite-total-column-water-vapour-ocean [CDS]\nAspect: validation_q01 | Category: Satellite_ECVs\nSection: Water vapor amplification of Earth's Greenhouse Effect > ℹ️ If you want to know more > References\n---\n[[1]](https://doi.org/10.1029/2010JD014287) Schmidt G. A., Ruedy R. A., Miller R. L., and A. A. Lacis (2010), Attribution of the present‐day total greenhouse effect,J. Geophys. Res.,115, D20106, doi:10.1029/2010JD014287\n\n[[2]](https://doi.org/10.1175/1520-0477%281997%29078<0197:EAGMEB>2.0.CO;2) Kiehl J. T. and K. Trenberth (1997), Earth's Annual Global Mean Energy Budget, Bull. Am. Meteorol. Soc., 78, 197-208, https://doi.org/10.1175/1520-0477%281997%29078<0197:EAGMEB>2.0.CO;2\n\n[[3]](https://doi.org/10.1175/1520-0469%281967%29024<0241:TEOTAW>2.0.CO;2) Manabe S., and R. T. Wetherald (1967), Thermal Equilibrium of the Atmosphere with a Given Distribution of Relative Humidity, J. Atmos. Sci., 24, 241–259, https://doi.org/10.1175/1520-0469%281967%29024<0241:TEOTAW>2.0.CO;2\n\n[[4]](https://doi.org/10.1146/annurev.energy.25.1.441) Held I. and B. Soden (2000), Water vapor feedback and global warming, Annu. Rev. Energy Environn., 25:441-75, doi:10.1146/annurev.energy.25.1.441\n\n[[5]](https://doi.org/10.1073/pnas.1809868115), Koll D. and T. Cronin (2018) Earth’s outgoing longwave radiation linear due to H2O greenhouse effect, Proc. Nat. Ac. Sci., 115 (41) 10293-10298, doi:10.1073/pnas.1809868115\n\n[[6]](https://doi.org/full/10.1029/2000JD000040) Roca R., Viollier M., Picon L., and M. Desbois (2002), A multisatellite analysis of deep convection and its moist environment over the Indian Ocean during the winter monsoon, J. Geophys. Res., 107(D19), doi:10.1029/2000JD000040.\n\n[[7]](https://doi.org/10.1002/2013JD020184) Gordon N., A. Jonko, P. Forster and K. Shell (2013), An observationally based constraint on the water-vapor feedback, J. Geophys. Res., 118, 12,435-12,443, doi:10.1002/2013JD020184.\n\n[[8]](https://doi.org/10.1029/2019JD031017) Raghuraman S. P., D. Paynter and V. Ramaswamy (2019), Quantifying the drivers of the clear sky greenhouse effect, 2000-2016, J. Geophys. Res, 124, 11,354-11,371, https://doi.org/10.1029/2019JD031017\n\n[[9]](https://www.ipcc.ch/report/ar6/wg1/downloads/report/IPCC_AR6_WGI_Chapter07.pdf) Forster P., Storelvmo T., Armour K., Collins W., Dufresne J-L, Frame D., Lunt D.J., Mauritsen T., Palmer M.D., Watanabe M., Wild M., and H. Zhang (2021), Chapter 7 The Earth’s Energy Budget, Climate Feedbacks, and Climate Sensitivity. In Climate Change 2021: The Physical Science Basis. Contribution of Working Group I to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, pp. 923–1054, doi: 10.1017/9781009157896.009."} {"chunk_id": "satellite_satellite-total-column-water-vapour-ocean_validation_q01__19ebcc033797", "report_id": "satellite_satellite-total-column-water-vapour-ocean_validation_q01", "dataset_id": "satellite-total-column-water-vapour-ocean", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q01", "aspect_base": "validation", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Water vapor amplification of Earth's Greenhouse Effect > ℹ️ If you want to know more > References", "title": "Water vapor amplification of Earth's Greenhouse Effect", "chunk_index": 20, "token_count": 964, "text_raw": "al Panel on Climate Change. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, pp. 923–1054, doi: 10.1017/9781009157896.009.\n\n[[10]](https://doi.org/abs/10.1029/1998JD900007) Inamdar A. and V. Ramanathan (1998), Tropical and global scale interactions among water vapor, greenhouse effect and surface temperature, J. Geophys. Res, 103, D24, 32,177-32,194, doi: https://doi.org/10.1029/1998JD900007\n\n[[11]](https://doi.org/10.1175/JCLI3990.1) Held I. and B. Soden (2006), Robust responses of the hydrological cycle to global warming, J. Clim., 19, 5686–5699, https://doi.org/10.1175/JCLI3990.1\n\n[[12]](https://essd.copernicus.org/articles/2/215/2010/essd-2-215-2010.html) Andersson A., Fennig K., Klepp C., Bakan S., Graßl H., and J. Schulz (2010), The Hamburg Ocean Atmosphere Parameters and Fluxes from Satellite Data – HOAPS-3, Earth Syst. Sci. Data, 2, 215-234, doi:10.5194/essd-2-215-2010\n\n[[13]](https://doi.org/10.5676/EUM_SAF_CM/CLARA_AVHRR/V003) Karlsson K., Riihelä A., Trentmann J., Stengel M., Solodovnik I., Meirink J., Devasthale A., Jääskeläinen E., Kallio-Myers V., Eliasson S., Benas N., Johansson E., Stein D., Finkensieper S., Håkansson N., Akkermans T., Clerbaux N., Selbach N., Schröder M. and R. Hollmann (2023), CLARA-A3: CM SAF cLoud, Albedo and surface RAdiation dataset from AVHRR data - Edition 3, Satellite Application Facility on Climate Monitoring\nDOI:10.5676/EUM_SAF_CM/CLARA_AVHRR/V003, https://doi.org/10.5676/EUM_SAF_CM/CLARA_AVHRR/V003\n\n[[14]](https://doi.org/10.1038/s41597-019-0236-x) Merchant C. J., Embury O., Bulgin C. E., Block T., Corlett G. K., Fiedler E., Good S. A., Mittaz J., Rayner N. A., Berry D., Eastwood S., Taylor M., Tsushima Y., Waterfall A., Wilson R. and C. Donlon (2019), Satellite-based time-series of sea-surface temperature since 1981 for climate applications. Scientific Data, 6. 223. doi:10.1038/s41597-019-0236-x\n\n[[15]](https://doi.org/10.1175/JCLI3611.1) Forster P. and J. Gregory (2006), the climate sensitivity and its components diagnosed from earth radiation budget data, J. Clim., 19, 39-52, doi:10.1175/JCLI3611.1.\n\n[[16]](https://doi.org/10.1038/351027a0) Ramanathan V. and W. Collins (1991), Thermodynamic regulation of ocean warming by cirrus clouds deduced from observations of the 1987 El Nino, Nature, 351, 27-32.\n\n[[17]](https://doi.org/epdf/10.1029/92JD00729) Kiehl J. and B. Briegleb (1992), Comparison of the observed and calculated sky sky greenhouse effect: implications for climate studies, J. Geophys. Res, 97, 10,037-10,049.", "text_with_prefix": "EQC Quality Assessment: \"Water vapor amplification of Earth's Greenhouse Effect\"\nDataset: satellite-total-column-water-vapour-ocean [CDS]\nAspect: validation_q01 | Category: Satellite_ECVs\nSection: Water vapor amplification of Earth's Greenhouse Effect > ℹ️ If you want to know more > References\n---\nal Panel on Climate Change. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, pp. 923–1054, doi: 10.1017/9781009157896.009.\n\n[[10]](https://doi.org/abs/10.1029/1998JD900007) Inamdar A. and V. Ramanathan (1998), Tropical and global scale interactions among water vapor, greenhouse effect and surface temperature, J. Geophys. Res, 103, D24, 32,177-32,194, doi: https://doi.org/10.1029/1998JD900007\n\n[[11]](https://doi.org/10.1175/JCLI3990.1) Held I. and B. Soden (2006), Robust responses of the hydrological cycle to global warming, J. Clim., 19, 5686–5699, https://doi.org/10.1175/JCLI3990.1\n\n[[12]](https://essd.copernicus.org/articles/2/215/2010/essd-2-215-2010.html) Andersson A., Fennig K., Klepp C., Bakan S., Graßl H., and J. Schulz (2010), The Hamburg Ocean Atmosphere Parameters and Fluxes from Satellite Data – HOAPS-3, Earth Syst. Sci. Data, 2, 215-234, doi:10.5194/essd-2-215-2010\n\n[[13]](https://doi.org/10.5676/EUM_SAF_CM/CLARA_AVHRR/V003) Karlsson K., Riihelä A., Trentmann J., Stengel M., Solodovnik I., Meirink J., Devasthale A., Jääskeläinen E., Kallio-Myers V., Eliasson S., Benas N., Johansson E., Stein D., Finkensieper S., Håkansson N., Akkermans T., Clerbaux N., Selbach N., Schröder M. and R. Hollmann (2023), CLARA-A3: CM SAF cLoud, Albedo and surface RAdiation dataset from AVHRR data - Edition 3, Satellite Application Facility on Climate Monitoring\nDOI:10.5676/EUM_SAF_CM/CLARA_AVHRR/V003, https://doi.org/10.5676/EUM_SAF_CM/CLARA_AVHRR/V003\n\n[[14]](https://doi.org/10.1038/s41597-019-0236-x) Merchant C. J., Embury O., Bulgin C. E., Block T., Corlett G. K., Fiedler E., Good S. A., Mittaz J., Rayner N. A., Berry D., Eastwood S., Taylor M., Tsushima Y., Waterfall A., Wilson R. and C. Donlon (2019), Satellite-based time-series of sea-surface temperature since 1981 for climate applications. Scientific Data, 6. 223. doi:10.1038/s41597-019-0236-x\n\n[[15]](https://doi.org/10.1175/JCLI3611.1) Forster P. and J. Gregory (2006), the climate sensitivity and its components diagnosed from earth radiation budget data, J. Clim., 19, 39-52, doi:10.1175/JCLI3611.1.\n\n[[16]](https://doi.org/10.1038/351027a0) Ramanathan V. and W. Collins (1991), Thermodynamic regulation of ocean warming by cirrus clouds deduced from observations of the 1987 El Nino, Nature, 351, 27-32.\n\n[[17]](https://doi.org/epdf/10.1029/92JD00729) Kiehl J. and B. Briegleb (1992), Comparison of the observed and calculated sky sky greenhouse effect: implications for climate studies, J. Geophys. Res, 97, 10,037-10,049."} {"chunk_id": "satellite_satellite-upper-troposphere-humidity_validation_q01__62b9a7b8de2a", "report_id": "satellite_satellite-upper-troposphere-humidity_validation_q01", "dataset_id": "satellite-upper-troposphere-humidity", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q01", "aspect_base": "validation", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Studying the influence of atmospheric circulation on upper tropospheric humidity", "title": "Studying the influence of atmospheric circulation on upper tropospheric humidity", "chunk_index": 0, "token_count": 120, "text_raw": "Production date: 04-04-2025\n\nProduced by : UVSQ/LSCE-IPSL (Hélène Brogniez) ; CNRS/LMD-IPSL (Giulio Mandorli, Claudia Stubenrauch)", "text_with_prefix": "EQC Quality Assessment: \"Studying the influence of atmospheric circulation on upper tropospheric humidity\"\nDataset: satellite-upper-troposphere-humidity [CDS]\nAspect: validation_q01 | Category: Satellite_ECVs\nSection: Studying the influence of atmospheric circulation on upper tropospheric humidity\n---\nProduction date: 04-04-2025\n\nProduced by : UVSQ/LSCE-IPSL (Hélène Brogniez) ; CNRS/LMD-IPSL (Giulio Mandorli, Claudia Stubenrauch)"} {"chunk_id": "satellite_satellite-upper-troposphere-humidity_validation_q01__c4f696a5b31f", "report_id": "satellite_satellite-upper-troposphere-humidity_validation_q01", "dataset_id": "satellite-upper-troposphere-humidity", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q01", "aspect_base": "validation", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Studying the influence of atmospheric circulation on upper tropospheric humidity > Quality assessment question:", "title": "Studying the influence of atmospheric circulation on upper tropospheric humidity", "chunk_index": 1, "token_count": 462, "text_raw": "- __How well can satellite measurements reproduce the known relationship between large-scale circulation and upper tropospheric humidity (UTH)?__\n\nWater vapor in the mid- to upper troposphere, while representing a small fraction of the total column, has\na significant impact on the Earth's radiative budget thanks to its radiative and thermodynamical properties. In the tropics, the distribution of upper tropospheric humidity (UTH) is primarily controlled by large-scale transport, with influences by storm systems in the deep tropics [[1]](https://doi.org/10.1029/97GL03563).\n\nThe atmospheric circulation is the large-scale movement of air through the Earth's atmosphere, driven by the uneven heating of the planet's surface by the sun. This circulation is responsible for distributing heat and moisture around the globe, shaping weather patterns and climate.\nThe movement of the air masses influences the distribution of humidity in the upper troposphere. In regions where air rises, such as the Inter Tropical Convergence Zone, moisture from the lower atmosphere is transported upward, increasing humidity in the upper troposphere.\nConversely, in regions where air sinks, the upper troposphere becomes drier.\n\nThis notebook aims to examine the dependency of upper tropospheric humidity on the global atmospheric circulation. Specifically, it seeks to verify if the UTH dataset can reproduce this dependency in agreement with [[2]](https://doi.org/10.1029/2006GL029118). The analysis will compare the dataset's findings with established results to ensure consistency and reliability in representing the atmospheric dynamics.\n\nThe analysis is performed using the dataset __Upper tropospheric humidity gridded data from 1999 to present derived from satellite observations__ [[described here]](https://cds.climate.copernicus.eu/datasets/satellite-upper-troposphere-humidity) available on the Climate Data Store of the Copernicus Climate Change Service.", "text_with_prefix": "EQC Quality Assessment: \"Studying the influence of atmospheric circulation on upper tropospheric humidity\"\nDataset: satellite-upper-troposphere-humidity [CDS]\nAspect: validation_q01 | Category: Satellite_ECVs\nSection: Studying the influence of atmospheric circulation on upper tropospheric humidity > Quality assessment question:\n---\n- __How well can satellite measurements reproduce the known relationship between large-scale circulation and upper tropospheric humidity (UTH)?__\n\nWater vapor in the mid- to upper troposphere, while representing a small fraction of the total column, has\na significant impact on the Earth's radiative budget thanks to its radiative and thermodynamical properties. In the tropics, the distribution of upper tropospheric humidity (UTH) is primarily controlled by large-scale transport, with influences by storm systems in the deep tropics [[1]](https://doi.org/10.1029/97GL03563).\n\nThe atmospheric circulation is the large-scale movement of air through the Earth's atmosphere, driven by the uneven heating of the planet's surface by the sun. This circulation is responsible for distributing heat and moisture around the globe, shaping weather patterns and climate.\nThe movement of the air masses influences the distribution of humidity in the upper troposphere. In regions where air rises, such as the Inter Tropical Convergence Zone, moisture from the lower atmosphere is transported upward, increasing humidity in the upper troposphere.\nConversely, in regions where air sinks, the upper troposphere becomes drier.\n\nThis notebook aims to examine the dependency of upper tropospheric humidity on the global atmospheric circulation. Specifically, it seeks to verify if the UTH dataset can reproduce this dependency in agreement with [[2]](https://doi.org/10.1029/2006GL029118). The analysis will compare the dataset's findings with established results to ensure consistency and reliability in representing the atmospheric dynamics.\n\nThe analysis is performed using the dataset __Upper tropospheric humidity gridded data from 1999 to present derived from satellite observations__ [[described here]](https://cds.climate.copernicus.eu/datasets/satellite-upper-troposphere-humidity) available on the Climate Data Store of the Copernicus Climate Change Service."} {"chunk_id": "satellite_satellite-upper-troposphere-humidity_validation_q01__1b40ecd5ec0a", "report_id": "satellite_satellite-upper-troposphere-humidity_validation_q01", "dataset_id": "satellite-upper-troposphere-humidity", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q01", "aspect_base": "validation", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Studying the influence of atmospheric circulation on upper tropospheric humidity > Quality assessment statement", "title": "Studying the influence of atmospheric circulation on upper tropospheric humidity", "chunk_index": 2, "token_count": 428, "text_raw": "These are the key outcomes of this assessment\n\n- The analysis is done with year 2001 using monthly data over the tropical regions. The UTH dataset reproduces the known relationship between the atmospheric moisture and the tropical large-scale atmospheric circulation: ascending motions moisten the upper troposphere, whereas subsiding motions dry it. The amplitude and frequency of moist and dry situations are consistent with the literature [[2]](https://doi.org/10.1029/2006GL029118).\n\n- The tested subset of the UTH dataset shows little month-to-month variability over the year, in agreement with the known link between atmospheric circulation regimes and UTH which is independent on the season.\n\n- Some slight discrepancies with the literature arise and can be explained :\n\n- differences in the amplitude and frequency of the vertical velocity at 500hPa : the quality assessment uses ERA5 reanalyses while the literature relied on former versions of reanalysis (ERA40 and NCEP/NCAR reanalyses, see [[2]](https://doi.org/10.1029/2006GL029118) and [[3]](https://doi.org/10.1007/s00382-003-0369-6)). Differences in the results may arise from differences in their horizontal resolutions, in their vertical description of the atmosphere, in the version of the assimilation system and in the atmospheric model.\n\n- differences in the amplitude of UTH : the retrieval method of the UTH of the CDS data record [[see the ATBD]](https://confluence.ecmwf.int/pages/viewpage.action?pageId=288339045) is not strictly identical to the reference paper [[2]](https://doi.org/10.1029/2006GL029118).\n\n```", "text_with_prefix": "EQC Quality Assessment: \"Studying the influence of atmospheric circulation on upper tropospheric humidity\"\nDataset: satellite-upper-troposphere-humidity [CDS]\nAspect: validation_q01 | Category: Satellite_ECVs\nSection: Studying the influence of atmospheric circulation on upper tropospheric humidity > Quality assessment statement\n---\nThese are the key outcomes of this assessment\n\n- The analysis is done with year 2001 using monthly data over the tropical regions. The UTH dataset reproduces the known relationship between the atmospheric moisture and the tropical large-scale atmospheric circulation: ascending motions moisten the upper troposphere, whereas subsiding motions dry it. The amplitude and frequency of moist and dry situations are consistent with the literature [[2]](https://doi.org/10.1029/2006GL029118).\n\n- The tested subset of the UTH dataset shows little month-to-month variability over the year, in agreement with the known link between atmospheric circulation regimes and UTH which is independent on the season.\n\n- Some slight discrepancies with the literature arise and can be explained :\n\n- differences in the amplitude and frequency of the vertical velocity at 500hPa : the quality assessment uses ERA5 reanalyses while the literature relied on former versions of reanalysis (ERA40 and NCEP/NCAR reanalyses, see [[2]](https://doi.org/10.1029/2006GL029118) and [[3]](https://doi.org/10.1007/s00382-003-0369-6)). Differences in the results may arise from differences in their horizontal resolutions, in their vertical description of the atmosphere, in the version of the assimilation system and in the atmospheric model.\n\n- differences in the amplitude of UTH : the retrieval method of the UTH of the CDS data record [[see the ATBD]](https://confluence.ecmwf.int/pages/viewpage.action?pageId=288339045) is not strictly identical to the reference paper [[2]](https://doi.org/10.1029/2006GL029118).\n\n```"} {"chunk_id": "satellite_satellite-upper-troposphere-humidity_validation_q01__e93f515f1f09", "report_id": "satellite_satellite-upper-troposphere-humidity_validation_q01", "dataset_id": "satellite-upper-troposphere-humidity", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q01", "aspect_base": "validation", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Studying the influence of atmospheric circulation on upper tropospheric humidity > Methodology", "title": "Studying the influence of atmospheric circulation on upper tropospheric humidity", "chunk_index": 3, "token_count": 143, "text_raw": "The link between upper tropospheric humidity (UTH) and large-scale circulation is examined using only the vertical atmospheric motion in the mid-troposphere at 500hPa (noted $\\omega_{500}$) available from the Climate Data Store and detailed below.\n\nThe UTH data over the __tropics__ is downloaded for the full year 2001.", "text_with_prefix": "EQC Quality Assessment: \"Studying the influence of atmospheric circulation on upper tropospheric humidity\"\nDataset: satellite-upper-troposphere-humidity [CDS]\nAspect: validation_q01 | Category: Satellite_ECVs\nSection: Studying the influence of atmospheric circulation on upper tropospheric humidity > Methodology\n---\nThe link between upper tropospheric humidity (UTH) and large-scale circulation is examined using only the vertical atmospheric motion in the mid-troposphere at 500hPa (noted $\\omega_{500}$) available from the Climate Data Store and detailed below.\n\nThe UTH data over the __tropics__ is downloaded for the full year 2001."} {"chunk_id": "satellite_satellite-upper-troposphere-humidity_validation_q01__cb4a52a53047", "report_id": "satellite_satellite-upper-troposphere-humidity_validation_q01", "dataset_id": "satellite-upper-troposphere-humidity", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q01", "aspect_base": "validation", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Studying the influence of atmospheric circulation on upper tropospheric humidity > Methodology > Upper Tropospheric Humidity", "title": "Studying the influence of atmospheric circulation on upper tropospheric humidity", "chunk_index": 4, "token_count": 576, "text_raw": "The Upper Tropospheric Humidity (UTH) is a direct interpretation of the satellite measurements in strong water vapor absorption lines, such as those in the infra-red band at 6.2µm and in the micro-wave in the 183.31GHz line. Indeed, at these wavelengths, the emission to space is explained by water vapor concentration and temperature (see early studies by [[4]](https://doi.org/10.1175/1520-0450%281988%29027%3C0889:EOTUTR%3E2.0.CO;2) and [[5]](https://doi.org/10.1029/93JD01283)).\nUpon some simple assumptions on the local vertical lapse rate of the troposphere and on the behaviour of the absorption spectral lines, the UTH is obtained via a log-linear equation from the satellite measurement (brightness temperature, $TB$ in Kelvins):\n\n$$\n \\text{ln}( \\ UTH \\ ) = a + b \\cdot TB\n $$\n\nPhysically, the UTH is a vertical integration of the relative humidity, weighted by the vertical sensitivity of the measurement to water vapor distribution. It is therefore expressed in the unit of %, the S.I. unit of relative humidity, and, as such, is interpreted in the same way.\n\nThe computation of the regression parameters $a$ and $b$ is done thanks to a statistically representative dataset (see [[5]](https://doi.org/10.1029/93JD01283), [[6]](https://doi.org/10.1029/2007JD009314), [[7]](https://doi.org/10.1175/2009JCLI2963.1), [[8]](https://doi.org/10.5194/acp-14-11129-2014)), and the values depend on the definition of UTH itself. It also depends on the spectral domain (infra-red of microwave).\n\nUTH may be prefered over vertically resolved relative humidity profiles, provided for example by reanalyses such as ERA5, when the vertical resolution is not a key parameter and when independence from a modeling framework is sought ([[9]](https://doi.org/10.5194/acp-7-2489-2007), [[10]](https://doi.org/10.1175/JCLI-D-19-0046.1)).", "text_with_prefix": "EQC Quality Assessment: \"Studying the influence of atmospheric circulation on upper tropospheric humidity\"\nDataset: satellite-upper-troposphere-humidity [CDS]\nAspect: validation_q01 | Category: Satellite_ECVs\nSection: Studying the influence of atmospheric circulation on upper tropospheric humidity > Methodology > Upper Tropospheric Humidity\n---\nThe Upper Tropospheric Humidity (UTH) is a direct interpretation of the satellite measurements in strong water vapor absorption lines, such as those in the infra-red band at 6.2µm and in the micro-wave in the 183.31GHz line. Indeed, at these wavelengths, the emission to space is explained by water vapor concentration and temperature (see early studies by [[4]](https://doi.org/10.1175/1520-0450%281988%29027%3C0889:EOTUTR%3E2.0.CO;2) and [[5]](https://doi.org/10.1029/93JD01283)).\nUpon some simple assumptions on the local vertical lapse rate of the troposphere and on the behaviour of the absorption spectral lines, the UTH is obtained via a log-linear equation from the satellite measurement (brightness temperature, $TB$ in Kelvins):\n\n$$\n \\text{ln}( \\ UTH \\ ) = a + b \\cdot TB\n $$\n\nPhysically, the UTH is a vertical integration of the relative humidity, weighted by the vertical sensitivity of the measurement to water vapor distribution. It is therefore expressed in the unit of %, the S.I. unit of relative humidity, and, as such, is interpreted in the same way.\n\nThe computation of the regression parameters $a$ and $b$ is done thanks to a statistically representative dataset (see [[5]](https://doi.org/10.1029/93JD01283), [[6]](https://doi.org/10.1029/2007JD009314), [[7]](https://doi.org/10.1175/2009JCLI2963.1), [[8]](https://doi.org/10.5194/acp-14-11129-2014)), and the values depend on the definition of UTH itself. It also depends on the spectral domain (infra-red of microwave).\n\nUTH may be prefered over vertically resolved relative humidity profiles, provided for example by reanalyses such as ERA5, when the vertical resolution is not a key parameter and when independence from a modeling framework is sought ([[9]](https://doi.org/10.5194/acp-7-2489-2007), [[10]](https://doi.org/10.1175/JCLI-D-19-0046.1))."} {"chunk_id": "satellite_satellite-upper-troposphere-humidity_validation_q01__b7af72cfc0e1", "report_id": "satellite_satellite-upper-troposphere-humidity_validation_q01", "dataset_id": "satellite-upper-troposphere-humidity", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q01", "aspect_base": "validation", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Studying the influence of atmospheric circulation on upper tropospheric humidity > Methodology > Method", "title": "Studying the influence of atmospheric circulation on upper tropospheric humidity", "chunk_index": 5, "token_count": 514, "text_raw": "Both variables (UTH and $\\omega_{500}$) are averaged monthly to smooth out short-term fluctuations caused by convection. The parameter $\\omega_{500}$ is extracted from ERA5 reanalyses (see below). ERA5 data resolution is reduced from 0.25 degree to 1 degree to match the resolution of the UTH data. After collocating the monthly values at this resolution, the UTH values are categorized into bins based on the vertical velocities. The distribution of UTH within each bin is presented as a boxplot profile, providing a clear visualization of how UTH varies with changes in vertical velocity.\n\nThe analysis comprises the following steps:\n\n__1. [](data-section-1)__\n - Import the relevant packages.\n - Define the parameters of the analysis and set the dataset requests\n\n__2. [](data-section-2)__\n - Download the variables of interest: Upper tropospheric Humidity is obtained from __Upper tropospheric humidity gridded data from 1999 to present derived from satellite observations__ [[UTH]](https://cds.climate.copernicus.eu/datasets/satellite-upper-troposphere-humidity), Vertical component of wind $\\omega_{500}$ is obtained from __ERA5 hourly data on pressure levels from 1940 to present__ [[ERA5]](https://cds.climate.copernicus.eu/datasets/reanalysis-era5-pressure-levels?tab=overview).\n - ERA5 $\\omega_{500}$ is regridded to align with the resolution of UTH and its unit is converted from hPa/s to hPa/day to be consistent with published references.\n - The datasets are colocated in space and time over the Tropics.\n - Monthly averages are used.\n\n__3. [](Analysis-section-1)__\n - Maps and histograms of UTH and $\\omega_{500}$ are described at the monthly scale for the whole year 2001.\n\n__4. [](Plot-and-results)__\n - The composite of UTH is shown as a function of $\\omega_{500}$. The final results are compared with the appropriate references and the sources of differences are discussed.", "text_with_prefix": "EQC Quality Assessment: \"Studying the influence of atmospheric circulation on upper tropospheric humidity\"\nDataset: satellite-upper-troposphere-humidity [CDS]\nAspect: validation_q01 | Category: Satellite_ECVs\nSection: Studying the influence of atmospheric circulation on upper tropospheric humidity > Methodology > Method\n---\nBoth variables (UTH and $\\omega_{500}$) are averaged monthly to smooth out short-term fluctuations caused by convection. The parameter $\\omega_{500}$ is extracted from ERA5 reanalyses (see below). ERA5 data resolution is reduced from 0.25 degree to 1 degree to match the resolution of the UTH data. After collocating the monthly values at this resolution, the UTH values are categorized into bins based on the vertical velocities. The distribution of UTH within each bin is presented as a boxplot profile, providing a clear visualization of how UTH varies with changes in vertical velocity.\n\nThe analysis comprises the following steps:\n\n__1. [](data-section-1)__\n - Import the relevant packages.\n - Define the parameters of the analysis and set the dataset requests\n\n__2. [](data-section-2)__\n - Download the variables of interest: Upper tropospheric Humidity is obtained from __Upper tropospheric humidity gridded data from 1999 to present derived from satellite observations__ [[UTH]](https://cds.climate.copernicus.eu/datasets/satellite-upper-troposphere-humidity), Vertical component of wind $\\omega_{500}$ is obtained from __ERA5 hourly data on pressure levels from 1940 to present__ [[ERA5]](https://cds.climate.copernicus.eu/datasets/reanalysis-era5-pressure-levels?tab=overview).\n - ERA5 $\\omega_{500}$ is regridded to align with the resolution of UTH and its unit is converted from hPa/s to hPa/day to be consistent with published references.\n - The datasets are colocated in space and time over the Tropics.\n - Monthly averages are used.\n\n__3. [](Analysis-section-1)__\n - Maps and histograms of UTH and $\\omega_{500}$ are described at the monthly scale for the whole year 2001.\n\n__4. [](Plot-and-results)__\n - The composite of UTH is shown as a function of $\\omega_{500}$. The final results are compared with the appropriate references and the sources of differences are discussed."} {"chunk_id": "satellite_satellite-upper-troposphere-humidity_validation_q01__9d6ae4c955e9", "report_id": "satellite_satellite-upper-troposphere-humidity_validation_q01", "dataset_id": "satellite-upper-troposphere-humidity", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q01", "aspect_base": "validation", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Studying the influence of atmospheric circulation on upper tropospheric humidity > Analysis and results > Choose the parameters to use and setup code > Define parameters", "title": "Studying the influence of atmospheric circulation on upper tropospheric humidity", "chunk_index": 6, "token_count": 119, "text_raw": "This use case is developed for one year of data (2001) and for the tropical belt, restricted to 30$^\\circ$S-30$^\\circ$N.", "text_with_prefix": "EQC Quality Assessment: \"Studying the influence of atmospheric circulation on upper tropospheric humidity\"\nDataset: satellite-upper-troposphere-humidity [CDS]\nAspect: validation_q01 | Category: Satellite_ECVs\nSection: Studying the influence of atmospheric circulation on upper tropospheric humidity > Analysis and results > Choose the parameters to use and setup code > Define parameters\n---\nThis use case is developed for one year of data (2001) and for the tropical belt, restricted to 30$^\\circ$S-30$^\\circ$N."} {"chunk_id": "satellite_satellite-upper-troposphere-humidity_validation_q01__de10efc54693", "report_id": "satellite_satellite-upper-troposphere-humidity_validation_q01", "dataset_id": "satellite-upper-troposphere-humidity", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q01", "aspect_base": "validation", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Studying the influence of atmospheric circulation on upper tropospheric humidity > Analysis and results > Choose the parameters to use and setup code > Set the data request", "title": "Studying the influence of atmospheric circulation on upper tropospheric humidity", "chunk_index": 7, "token_count": 121, "text_raw": "Two datasets are requested :\n- Upper Tropospheric Humidity (UTH)\n- Vertical atmospheric velocity at 500hPa ($\\omega_{500}$)\n\n(data-section-2)=", "text_with_prefix": "EQC Quality Assessment: \"Studying the influence of atmospheric circulation on upper tropospheric humidity\"\nDataset: satellite-upper-troposphere-humidity [CDS]\nAspect: validation_q01 | Category: Satellite_ECVs\nSection: Studying the influence of atmospheric circulation on upper tropospheric humidity > Analysis and results > Choose the parameters to use and setup code > Set the data request\n---\nTwo datasets are requested :\n- Upper Tropospheric Humidity (UTH)\n- Vertical atmospheric velocity at 500hPa ($\\omega_{500}$)\n\n(data-section-2)="} {"chunk_id": "satellite_satellite-upper-troposphere-humidity_validation_q01__7d66eb956a26", "report_id": "satellite_satellite-upper-troposphere-humidity_validation_q01", "dataset_id": "satellite-upper-troposphere-humidity", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q01", "aspect_base": "validation", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Studying the influence of atmospheric circulation on upper tropospheric humidity > Analysis and results > Download the datasets and perform space-time aggregation", "title": "Studying the influence of atmospheric circulation on upper tropospheric humidity", "chunk_index": 8, "token_count": 143, "text_raw": "The datasets are now downloaded.\n\n```text\n100%|██████████| 12/12 [00:00<00:00, 15.31it/s]\n100%|██████████| 12/12 [00:00<00:00, 34.24it/s]\n```", "text_with_prefix": "EQC Quality Assessment: \"Studying the influence of atmospheric circulation on upper tropospheric humidity\"\nDataset: satellite-upper-troposphere-humidity [CDS]\nAspect: validation_q01 | Category: Satellite_ECVs\nSection: Studying the influence of atmospheric circulation on upper tropospheric humidity > Analysis and results > Download the datasets and perform space-time aggregation\n---\nThe datasets are now downloaded.\n\n```text\n100%|██████████| 12/12 [00:00<00:00, 15.31it/s]\n100%|██████████| 12/12 [00:00<00:00, 34.24it/s]\n```"} {"chunk_id": "satellite_satellite-upper-troposphere-humidity_validation_q01__b3e7c17c4213", "report_id": "satellite_satellite-upper-troposphere-humidity_validation_q01", "dataset_id": "satellite-upper-troposphere-humidity", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q01", "aspect_base": "validation", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Studying the influence of atmospheric circulation on upper tropospheric humidity > Analysis and results > Download the datasets and perform space-time aggregation > Temporal and spatial aggregation", "title": "Studying the influence of atmospheric circulation on upper tropospheric humidity", "chunk_index": 9, "token_count": 297, "text_raw": "All data records are provided at different spatial and temporal resolutions : \n- Upper Tropospheric Humidity : 1$^\\circ$ / 1-day\n- Vertical atmospheric velocity at 500hPa ($\\omega_{500}$) : 0.25$^\\circ$ / 1-hour\n\nSince this is a cross-variable analysis, the study requires to process all the datasets at the same horizontal and temporal resolutions.\n\nNote that for the UTH data record, two manipulations of the latitude $\\times$ longitude grid are done : first a reversal of the latitudes since the data is stored South-to-North (North-to-South for ERA5 $\\omega_{500}$), and second a change of the longitude values since they range from 0° to 360° (-180° to 180° for ERA5 $\\omega_{500}$) .\n\nThe common spatial aggregation is thus 1$^\\circ$ and the data records are processed at the monthly scale.\n\ncoarsen datasets to 1° grid / 1-month", "text_with_prefix": "EQC Quality Assessment: \"Studying the influence of atmospheric circulation on upper tropospheric humidity\"\nDataset: satellite-upper-troposphere-humidity [CDS]\nAspect: validation_q01 | Category: Satellite_ECVs\nSection: Studying the influence of atmospheric circulation on upper tropospheric humidity > Analysis and results > Download the datasets and perform space-time aggregation > Temporal and spatial aggregation\n---\nAll data records are provided at different spatial and temporal resolutions : \n- Upper Tropospheric Humidity : 1$^\\circ$ / 1-day\n- Vertical atmospheric velocity at 500hPa ($\\omega_{500}$) : 0.25$^\\circ$ / 1-hour\n\nSince this is a cross-variable analysis, the study requires to process all the datasets at the same horizontal and temporal resolutions.\n\nNote that for the UTH data record, two manipulations of the latitude $\\times$ longitude grid are done : first a reversal of the latitudes since the data is stored South-to-North (North-to-South for ERA5 $\\omega_{500}$), and second a change of the longitude values since they range from 0° to 360° (-180° to 180° for ERA5 $\\omega_{500}$) .\n\nThe common spatial aggregation is thus 1$^\\circ$ and the data records are processed at the monthly scale.\n\ncoarsen datasets to 1° grid / 1-month"} {"chunk_id": "satellite_satellite-upper-troposphere-humidity_validation_q01__8c94eb563d0f", "report_id": "satellite_satellite-upper-troposphere-humidity_validation_q01", "dataset_id": "satellite-upper-troposphere-humidity", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q01", "aspect_base": "validation", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Studying the influence of atmospheric circulation on upper tropospheric humidity > Analysis and results > Download the datasets and perform space-time aggregation > Change of unit for vertical velocity", "title": "Studying the influence of atmospheric circulation on upper tropospheric humidity", "chunk_index": 10, "token_count": 166, "text_raw": "Most scientific studies focusing on the large-scale atmospheric circulation use __hPa/day__ as the unit for the upward atmospheric wind (see [[3]](https://doi.org/10.1007/s00382-003-0369-6)), since it offers a more meaningful unit at the considered time scale.\n\nfrom Pa/s to hPa/day\nand the unit in the dataset is changed accordingly", "text_with_prefix": "EQC Quality Assessment: \"Studying the influence of atmospheric circulation on upper tropospheric humidity\"\nDataset: satellite-upper-troposphere-humidity [CDS]\nAspect: validation_q01 | Category: Satellite_ECVs\nSection: Studying the influence of atmospheric circulation on upper tropospheric humidity > Analysis and results > Download the datasets and perform space-time aggregation > Change of unit for vertical velocity\n---\nMost scientific studies focusing on the large-scale atmospheric circulation use __hPa/day__ as the unit for the upward atmospheric wind (see [[3]](https://doi.org/10.1007/s00382-003-0369-6)), since it offers a more meaningful unit at the considered time scale.\n\nfrom Pa/s to hPa/day\nand the unit in the dataset is changed accordingly"} {"chunk_id": "satellite_satellite-upper-troposphere-humidity_validation_q01__0f03e00c8398", "report_id": "satellite_satellite-upper-troposphere-humidity_validation_q01", "dataset_id": "satellite-upper-troposphere-humidity", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q01", "aspect_base": "validation", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Studying the influence of atmospheric circulation on upper tropospheric humidity > Analysis and results > Download the datasets and perform space-time aggregation > Organisation of the data", "title": "Studying the influence of atmospheric circulation on upper tropospheric humidity", "chunk_index": 11, "token_count": 138, "text_raw": "Creation of a dataframe that keeps also the information of each month to study the month-to-month variability\nsort and stack the data\nconvert into a serie\nextraction of the month\nCreation of the dataframe\nSupression of missing values\n\n(Analysis-section-1)=", "text_with_prefix": "EQC Quality Assessment: \"Studying the influence of atmospheric circulation on upper tropospheric humidity\"\nDataset: satellite-upper-troposphere-humidity [CDS]\nAspect: validation_q01 | Category: Satellite_ECVs\nSection: Studying the influence of atmospheric circulation on upper tropospheric humidity > Analysis and results > Download the datasets and perform space-time aggregation > Organisation of the data\n---\nCreation of a dataframe that keeps also the information of each month to study the month-to-month variability\nsort and stack the data\nconvert into a serie\nextraction of the month\nCreation of the dataframe\nSupression of missing values\n\n(Analysis-section-1)="} {"chunk_id": "satellite_satellite-upper-troposphere-humidity_validation_q01__c41802fa45cb", "report_id": "satellite_satellite-upper-troposphere-humidity_validation_q01", "dataset_id": "satellite-upper-troposphere-humidity", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q01", "aspect_base": "validation", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Studying the influence of atmospheric circulation on upper tropospheric humidity > Analysis and results > Overview of the geophysical variables of interest > Upper Tropospheric Humidity", "title": "Studying the influence of atmospheric circulation on upper tropospheric humidity", "chunk_index": 12, "token_count": 461, "text_raw": "The Upper Tropospheric Humidity (UTH, in %) is a single-layer geophysical variable provided by EUMETSAT on a 1°$\\times$1° regular grid.\n\nThe UTH data record of the CDS relies on measurements from the 183.31±1 GHz channels of the Advanced Microwave Sounding Unit-B (AMSU-B) and the Microwave Humidity Sounder (MHS) on board the NOAA- and MetOp- satellite series. The present use case analyses the AMSU-B data of the NOAA-16 satellite.\n\nFor the 183GHz microwave channel the UTH is retrieved under all situations, except where measurements are influenced by surface (high orography or extremely dry atmospheric conditions), or when there is a deep convective storm with large ice particules (see [[2]](https://doi.org/10.1029/2006GL029118), [[6]](https://doi.org/10.1029/2007JD009314)).\n\nThe map below shows the UTH for July 2001.\n\n*Figure 1: Monthly average of UTH (in % of relative humidity) for July 2001.*\n\nAs expected (see [[5]](https://doi.org/10.1029/93JD01283), [[8]](https://doi.org/10.5194/acp-14-11129-2014)) high UTH values are associated to monsoonal regions of strong storms that bring upward a high amount of moisture : in the Eastern Pacific, over West Africa and in the Indian Ocean and Indonesia (known as the Warm Pool).\nOn the other side of the UTH range, a drier atmosphere is visible in the continental desert regions, and more generally in both hemispheres, North and South of the Monsoonal regions.", "text_with_prefix": "EQC Quality Assessment: \"Studying the influence of atmospheric circulation on upper tropospheric humidity\"\nDataset: satellite-upper-troposphere-humidity [CDS]\nAspect: validation_q01 | Category: Satellite_ECVs\nSection: Studying the influence of atmospheric circulation on upper tropospheric humidity > Analysis and results > Overview of the geophysical variables of interest > Upper Tropospheric Humidity\n---\nThe Upper Tropospheric Humidity (UTH, in %) is a single-layer geophysical variable provided by EUMETSAT on a 1°$\\times$1° regular grid.\n\nThe UTH data record of the CDS relies on measurements from the 183.31±1 GHz channels of the Advanced Microwave Sounding Unit-B (AMSU-B) and the Microwave Humidity Sounder (MHS) on board the NOAA- and MetOp- satellite series. The present use case analyses the AMSU-B data of the NOAA-16 satellite.\n\nFor the 183GHz microwave channel the UTH is retrieved under all situations, except where measurements are influenced by surface (high orography or extremely dry atmospheric conditions), or when there is a deep convective storm with large ice particules (see [[2]](https://doi.org/10.1029/2006GL029118), [[6]](https://doi.org/10.1029/2007JD009314)).\n\nThe map below shows the UTH for July 2001.\n\n*Figure 1: Monthly average of UTH (in % of relative humidity) for July 2001.*\n\nAs expected (see [[5]](https://doi.org/10.1029/93JD01283), [[8]](https://doi.org/10.5194/acp-14-11129-2014)) high UTH values are associated to monsoonal regions of strong storms that bring upward a high amount of moisture : in the Eastern Pacific, over West Africa and in the Indian Ocean and Indonesia (known as the Warm Pool).\nOn the other side of the UTH range, a drier atmosphere is visible in the continental desert regions, and more generally in both hemispheres, North and South of the Monsoonal regions."} {"chunk_id": "satellite_satellite-upper-troposphere-humidity_validation_q01__8d641b4e5631", "report_id": "satellite_satellite-upper-troposphere-humidity_validation_q01", "dataset_id": "satellite-upper-troposphere-humidity", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q01", "aspect_base": "validation", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Studying the influence of atmospheric circulation on upper tropospheric humidity > Analysis and results > Overview of the geophysical variables of interest > Atmospheric vertical velocity at 500 hPa", "title": "Studying the influence of atmospheric circulation on upper tropospheric humidity", "chunk_index": 13, "token_count": 896, "text_raw": "The map below shows the $\\omega_{500}$ for the same month of July 2001, overlayed with contours of UTH of the same month.\n\n```text\nText(0.5, 1.0, 'July 2001 - monthly average')\n```\n\n*Figure 2: Monthly average of the vertical velocity at 500hPa (in hPa/day) for July 2001, from the ERA5 reanalyses. The overlaid contours are UTH from Figure 1.*\n\nNegative values of $\\omega_{500}$ indicate ascending motions, while positive values indicate descending motions.\n\nThe contours of UTH, overlayed on the map, help to interpret the structures.\n\nFrom the general atmospheric circulation, ascending motions ($\\omega_{500}$ < 0) are regions where storms form by convection. These regions form the moist Inter-Tropical Convergence Zone (ITCZ), a region of strong convergence of the surface winds of the Hadley cells. The convective storms moisten the troposphere by bringing upward the moisture from the boundary layer.\n\nDescending motions ($\\omega_{500}$ > 0) are dry regions that are free of storm clouds and are mainly located over the subtropics (near 20°S - 20°S). In these subsidence regions, the air sinks and warms adiabatically without changing its water vapour content, thus decreasing the rate of relative humidity.\n\nTherefore, $\\omega_{500}$ serves to define __large-scale circulation regimes__, as stated in [[3]](https://doi.org/10.1007/s00382-003-0369-6), extremely useful to decompose the dynamical structure of the tropical atmosphere and to analyze its properties (for the purpose of studying trends in precipitation [[11]](https://doi.org/10.1256/qj.04.176) or for the purpose of studying cloud radiative forcing [[3]](https://doi.org/10.1007/s00382-003-0369-6), among other applications). Many studies use this framework to assess global climate models ([[2]](https://doi.org/10.1029/2006GL029118), [[3]](https://doi.org/10.1007/s00382-003-0369-6), [[11]](https://doi.org/10.1256/qj.04.176)).\n\nThe following density histogram illustrates the probability density function (PDF) of $\\omega_{500}$ for the full year 2001.\n\n```text\n<>:18: SyntaxWarning: invalid escape sequence '\\o'\n<>:18: SyntaxWarning: invalid escape sequence '\\o'\n/data/wp5/.tmp/ipykernel_2407812/2054132186.py:18: SyntaxWarning: invalid escape sequence '\\o'\n plt.xlabel('circulation regimes - $\\omega_{500}$ [hPa/day]')\n```\n\n*Figure 3: Probability density functions (PDF) of the monthly $\\omega_{500}$ for the year 2001 in the tropics (30°N - 30°S). Each light-blue PDF is for one given month. The red marks and vertical bars show the average PDF for year 2001 and the associated standard deviation, considered as the uncertainty, within each 10-hPa bin.*\n\nOverall the distribution of $\\omega_{500}$ is negatively skewed, with a peak around 20 hPa/day. This peak, compliant with [[3]](https://doi.org/10.1007/s00382-003-0369-6), shows that the tropics are dominated by a weak subsiding motion found in the subtropics and in the western parts of the ocean basins.\nThe negative tail reflects the magnitude of vertical motions within clouds.\nThe small vertical bars highlight the small temporal variability in the occurrence of each bins of $\\omega_{500}$ throughout the year.\n\n(Plot-and-results)=", "text_with_prefix": "EQC Quality Assessment: \"Studying the influence of atmospheric circulation on upper tropospheric humidity\"\nDataset: satellite-upper-troposphere-humidity [CDS]\nAspect: validation_q01 | Category: Satellite_ECVs\nSection: Studying the influence of atmospheric circulation on upper tropospheric humidity > Analysis and results > Overview of the geophysical variables of interest > Atmospheric vertical velocity at 500 hPa\n---\nThe map below shows the $\\omega_{500}$ for the same month of July 2001, overlayed with contours of UTH of the same month.\n\n```text\nText(0.5, 1.0, 'July 2001 - monthly average')\n```\n\n*Figure 2: Monthly average of the vertical velocity at 500hPa (in hPa/day) for July 2001, from the ERA5 reanalyses. The overlaid contours are UTH from Figure 1.*\n\nNegative values of $\\omega_{500}$ indicate ascending motions, while positive values indicate descending motions.\n\nThe contours of UTH, overlayed on the map, help to interpret the structures.\n\nFrom the general atmospheric circulation, ascending motions ($\\omega_{500}$ < 0) are regions where storms form by convection. These regions form the moist Inter-Tropical Convergence Zone (ITCZ), a region of strong convergence of the surface winds of the Hadley cells. The convective storms moisten the troposphere by bringing upward the moisture from the boundary layer.\n\nDescending motions ($\\omega_{500}$ > 0) are dry regions that are free of storm clouds and are mainly located over the subtropics (near 20°S - 20°S). In these subsidence regions, the air sinks and warms adiabatically without changing its water vapour content, thus decreasing the rate of relative humidity.\n\nTherefore, $\\omega_{500}$ serves to define __large-scale circulation regimes__, as stated in [[3]](https://doi.org/10.1007/s00382-003-0369-6), extremely useful to decompose the dynamical structure of the tropical atmosphere and to analyze its properties (for the purpose of studying trends in precipitation [[11]](https://doi.org/10.1256/qj.04.176) or for the purpose of studying cloud radiative forcing [[3]](https://doi.org/10.1007/s00382-003-0369-6), among other applications). Many studies use this framework to assess global climate models ([[2]](https://doi.org/10.1029/2006GL029118), [[3]](https://doi.org/10.1007/s00382-003-0369-6), [[11]](https://doi.org/10.1256/qj.04.176)).\n\nThe following density histogram illustrates the probability density function (PDF) of $\\omega_{500}$ for the full year 2001.\n\n```text\n<>:18: SyntaxWarning: invalid escape sequence '\\o'\n<>:18: SyntaxWarning: invalid escape sequence '\\o'\n/data/wp5/.tmp/ipykernel_2407812/2054132186.py:18: SyntaxWarning: invalid escape sequence '\\o'\n plt.xlabel('circulation regimes - $\\omega_{500}$ [hPa/day]')\n```\n\n*Figure 3: Probability density functions (PDF) of the monthly $\\omega_{500}$ for the year 2001 in the tropics (30°N - 30°S). Each light-blue PDF is for one given month. The red marks and vertical bars show the average PDF for year 2001 and the associated standard deviation, considered as the uncertainty, within each 10-hPa bin.*\n\nOverall the distribution of $\\omega_{500}$ is negatively skewed, with a peak around 20 hPa/day. This peak, compliant with [[3]](https://doi.org/10.1007/s00382-003-0369-6), shows that the tropics are dominated by a weak subsiding motion found in the subtropics and in the western parts of the ocean basins.\nThe negative tail reflects the magnitude of vertical motions within clouds.\nThe small vertical bars highlight the small temporal variability in the occurrence of each bins of $\\omega_{500}$ throughout the year.\n\n(Plot-and-results)="} {"chunk_id": "satellite_satellite-upper-troposphere-humidity_validation_q01__18ff0ee00cff", "report_id": "satellite_satellite-upper-troposphere-humidity_validation_q01", "dataset_id": "satellite-upper-troposphere-humidity", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q01", "aspect_base": "validation", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Studying the influence of atmospheric circulation on upper tropospheric humidity > Analysis and results > Plot and description of the results > Variation of UTH with the vertical velocity at 500hPa", "title": "Studying the influence of atmospheric circulation on upper tropospheric humidity", "chunk_index": 14, "token_count": 936, "text_raw": "For a given latitude$\\times$longitude$\\times$month the UTH is associated a value of $\\omega_{500}$, defining a circulation regime, and each bin of $\\omega_{500}$ groups all UTH for this particular regime.\n\nThe entire year 2001 is considered, and month-to-month variability is presented.\n\nmean within each bin\n\n```text\n<>:25: SyntaxWarning: invalid escape sequence '\\o'\n<>:25: SyntaxWarning: invalid escape sequence '\\o'\n/data/wp5/.tmp/ipykernel_2407812/1983021514.py:25: SyntaxWarning: invalid escape sequence '\\o'\n xlabel=\"circulation regimes - $\\omega_{500}$ [hPa/day]\",\n```\n\n*Figure 4: Evolution of UTH upon different circulation regimes of the tropics, defined by 10-hPa bins of $\\omega_{500}$, for each month of year 2001. For each bin of $\\omega_{500}$, a box-and-whiskers diagram is used to show the month-to-month variability: the gray boxes are the interquartiles and the median are in grey. The green symbols represent the mean of the UTH of each bin of the given months. The red circles are the mean of the means (ie the annual mean) within each bin.*\n\nOverall the results are in agreement with Figure 1 of [[2]](https://doi.org/10.1029/2006GL029118), which associates the retrieved monthly UTH from NOAA-16 to the vertical velocity at 500hPa, but for only 2 months of data which are November and December 2001. In agreement with this study, the UTH shows a sharp transition from convergence regions to subsidence regions. The UTH value remains stable at around 20% over regions where $\\omega_{500}$ exceeds 30 hPa/day.\n\nFigure 4 shows that there is little month-to-month variability within a given circulation regime, the largest variability being observed in the extreme ascending regimes, for $\\omega_{500}$ > 55 hPa/day. This is essentially explained by the small occurrence of such regimes, as visible on Figure 1.\n\nCaution must be taken though when comparing the occurence of circulation regimes between the present use case and the literature : the use case is based on ERA5 vertical velocity while the literature ([[2]](https://doi.org/10.1029/2006GL029118) and [[3]](https://doi.org/10.1007/s00382-003-0369-6)) rely on former versions of ECMWF reanalysis ERA40 ([[11]](https://doi.org/10.1256/qj.04.176)) and NCEP/NCAR reanalyses ([[12]](https://doi.org/10.1175/1520-0477%281996%29077<0437:TNYRP>2.0.CO;2)). Therefore, differences in the present results and the literature may arise from differences in their horizontal resolutions, in their vertical description of the atmosphere, in the version of the assimilation system and in the atmospheric model\n\nHere, the UTH reaches a plateau of maximum value of approximately 45-50% when upward motions reach -60 hPa/day. The reference reports slightly lower values, which still fall within the UTH distribution observed in this study. The discrepancies between these results and those presented in [[2]](https://doi.org/10.1029/2006GL029118) can arise from the following reasons:\n\n- The retrieval method for UTH is similar, based on a simple linear equation ([[2]](https://doi.org/10.1029/2006GL029118), [[4]](https://doi.org/10.1175/1520-0450%281988%29027%3C0889:EOTUTR%3E2.0.CO;2)) (see also the __\"Methodology\"__ section):\n\n$$\n \\text{ln}(UTH) = a + b \\cdot TB\n $$", "text_with_prefix": "EQC Quality Assessment: \"Studying the influence of atmospheric circulation on upper tropospheric humidity\"\nDataset: satellite-upper-troposphere-humidity [CDS]\nAspect: validation_q01 | Category: Satellite_ECVs\nSection: Studying the influence of atmospheric circulation on upper tropospheric humidity > Analysis and results > Plot and description of the results > Variation of UTH with the vertical velocity at 500hPa\n---\nFor a given latitude$\\times$longitude$\\times$month the UTH is associated a value of $\\omega_{500}$, defining a circulation regime, and each bin of $\\omega_{500}$ groups all UTH for this particular regime.\n\nThe entire year 2001 is considered, and month-to-month variability is presented.\n\nmean within each bin\n\n```text\n<>:25: SyntaxWarning: invalid escape sequence '\\o'\n<>:25: SyntaxWarning: invalid escape sequence '\\o'\n/data/wp5/.tmp/ipykernel_2407812/1983021514.py:25: SyntaxWarning: invalid escape sequence '\\o'\n xlabel=\"circulation regimes - $\\omega_{500}$ [hPa/day]\",\n```\n\n*Figure 4: Evolution of UTH upon different circulation regimes of the tropics, defined by 10-hPa bins of $\\omega_{500}$, for each month of year 2001. For each bin of $\\omega_{500}$, a box-and-whiskers diagram is used to show the month-to-month variability: the gray boxes are the interquartiles and the median are in grey. The green symbols represent the mean of the UTH of each bin of the given months. The red circles are the mean of the means (ie the annual mean) within each bin.*\n\nOverall the results are in agreement with Figure 1 of [[2]](https://doi.org/10.1029/2006GL029118), which associates the retrieved monthly UTH from NOAA-16 to the vertical velocity at 500hPa, but for only 2 months of data which are November and December 2001. In agreement with this study, the UTH shows a sharp transition from convergence regions to subsidence regions. The UTH value remains stable at around 20% over regions where $\\omega_{500}$ exceeds 30 hPa/day.\n\nFigure 4 shows that there is little month-to-month variability within a given circulation regime, the largest variability being observed in the extreme ascending regimes, for $\\omega_{500}$ > 55 hPa/day. This is essentially explained by the small occurrence of such regimes, as visible on Figure 1.\n\nCaution must be taken though when comparing the occurence of circulation regimes between the present use case and the literature : the use case is based on ERA5 vertical velocity while the literature ([[2]](https://doi.org/10.1029/2006GL029118) and [[3]](https://doi.org/10.1007/s00382-003-0369-6)) rely on former versions of ECMWF reanalysis ERA40 ([[11]](https://doi.org/10.1256/qj.04.176)) and NCEP/NCAR reanalyses ([[12]](https://doi.org/10.1175/1520-0477%281996%29077<0437:TNYRP>2.0.CO;2)). Therefore, differences in the present results and the literature may arise from differences in their horizontal resolutions, in their vertical description of the atmosphere, in the version of the assimilation system and in the atmospheric model\n\nHere, the UTH reaches a plateau of maximum value of approximately 45-50% when upward motions reach -60 hPa/day. The reference reports slightly lower values, which still fall within the UTH distribution observed in this study. The discrepancies between these results and those presented in [[2]](https://doi.org/10.1029/2006GL029118) can arise from the following reasons:\n\n- The retrieval method for UTH is similar, based on a simple linear equation ([[2]](https://doi.org/10.1029/2006GL029118), [[4]](https://doi.org/10.1175/1520-0450%281988%29027%3C0889:EOTUTR%3E2.0.CO;2)) (see also the __\"Methodology\"__ section):\n\n$$\n \\text{ln}(UTH) = a + b \\cdot TB\n $$"} {"chunk_id": "satellite_satellite-upper-troposphere-humidity_validation_q01__8b1319b141ce", "report_id": "satellite_satellite-upper-troposphere-humidity_validation_q01", "dataset_id": "satellite-upper-troposphere-humidity", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q01", "aspect_base": "validation", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Studying the influence of atmospheric circulation on upper tropospheric humidity > Analysis and results > Plot and description of the results > Variation of UTH with the vertical velocity at 500hPa", "title": "Studying the influence of atmospheric circulation on upper tropospheric humidity", "chunk_index": 15, "token_count": 538, "text_raw": "%3E2.0.CO;2)) (see also the __\"Methodology\"__ section):\n\n$$\n \\text{ln}(UTH) = a + b \\cdot TB\n $$\n\nhowever, the regression coefficients $a$ and $b$ used to obtain the UTH from the TB measured by the instrument are not identical between the UTH dataset of the CDS [[see the ATBD]](https://confluence.ecmwf.int/pages/viewpage.action?pageId=288339045) and the UTH of the reference paper [[2]](https://doi.org/10.1029/2006GL029118), which may explain some of the differences in the binned UTH per $\\omega_{500}$. Firstly, those parameters are computed using offline radiative transfer codes that have some discrepancies in the water vapor absorption channels (see for instance [[[14]](https://doi.org/10.1175/1520-0477%282000%29081%3C0797:AIORCF%3E2.3.CO;2), [[15]](https://www.wcrp-climate.org/resources/wcrp-publications/1095-pub-2017)). Secondly, the dataset used to compute the regression coefficients, and considered as a reference dataset, is built from different sources : a collection of 5000 representative profiles for the CDS dataset [[see the ATBD, pages 9 and 10]](https://confluence.ecmwf.int/pages/viewpage.action?pageId=288339045), and an extraction of profiles from ERA40 reanalyses (see explanations in [[2]](https://doi.org/10.1029/2006GL029118)).\n\n- The results in [[2]](https://doi.org/10.1029/2006GL029118) are based on only two months of data (November and December 2001) using former versions of reanalyses (ERA40) for $\\omega_{500}$, which can lead to discrepancies in the amplitudes and occurences when compared to our more comprehensive dataset, even if limited to only one year.", "text_with_prefix": "EQC Quality Assessment: \"Studying the influence of atmospheric circulation on upper tropospheric humidity\"\nDataset: satellite-upper-troposphere-humidity [CDS]\nAspect: validation_q01 | Category: Satellite_ECVs\nSection: Studying the influence of atmospheric circulation on upper tropospheric humidity > Analysis and results > Plot and description of the results > Variation of UTH with the vertical velocity at 500hPa\n---\n%3E2.0.CO;2)) (see also the __\"Methodology\"__ section):\n\n$$\n \\text{ln}(UTH) = a + b \\cdot TB\n $$\n\nhowever, the regression coefficients $a$ and $b$ used to obtain the UTH from the TB measured by the instrument are not identical between the UTH dataset of the CDS [[see the ATBD]](https://confluence.ecmwf.int/pages/viewpage.action?pageId=288339045) and the UTH of the reference paper [[2]](https://doi.org/10.1029/2006GL029118), which may explain some of the differences in the binned UTH per $\\omega_{500}$. Firstly, those parameters are computed using offline radiative transfer codes that have some discrepancies in the water vapor absorption channels (see for instance [[[14]](https://doi.org/10.1175/1520-0477%282000%29081%3C0797:AIORCF%3E2.3.CO;2), [[15]](https://www.wcrp-climate.org/resources/wcrp-publications/1095-pub-2017)). Secondly, the dataset used to compute the regression coefficients, and considered as a reference dataset, is built from different sources : a collection of 5000 representative profiles for the CDS dataset [[see the ATBD, pages 9 and 10]](https://confluence.ecmwf.int/pages/viewpage.action?pageId=288339045), and an extraction of profiles from ERA40 reanalyses (see explanations in [[2]](https://doi.org/10.1029/2006GL029118)).\n\n- The results in [[2]](https://doi.org/10.1029/2006GL029118) are based on only two months of data (November and December 2001) using former versions of reanalyses (ERA40) for $\\omega_{500}$, which can lead to discrepancies in the amplitudes and occurences when compared to our more comprehensive dataset, even if limited to only one year."} {"chunk_id": "satellite_satellite-upper-troposphere-humidity_validation_q01__f717b83750b9", "report_id": "satellite_satellite-upper-troposphere-humidity_validation_q01", "dataset_id": "satellite-upper-troposphere-humidity", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q01", "aspect_base": "validation", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Studying the influence of atmospheric circulation on upper tropospheric humidity > ℹ️ If you want to know more > Key resources", "title": "Studying the influence of atmospheric circulation on upper tropospheric humidity", "chunk_index": 16, "token_count": 547, "text_raw": "Some key resources and further reading were linked throughout this assessment.\n\nThe CDS catalogue entries for the data used is:\n- Upper tropospheric humidity gridded data from 1999 to present derived from satellite observations:\nhttps://cds.climate.copernicus.eu/datasets/satellite-upper-troposphere-humidity\n\n- ERA5 hourly data on pressure levels from 1940 to present:\nhttps://cds.climate.copernicus.eu/datasets/reanalysis-era5-pressure-levels\n\nCode libraries used:\n- [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)\n\nAnd more resources on tropical atmosphere, large-scale circulation and upper tropospheric humidity :\n\n- https://www.nature.com/scitable/knowledge/library/tropical-weather-84224797/\n\n- Additonal scientific articles on Upper Tropospheric Humidity :\n\n- IPCC Assessment Report 4, 2007 : 3.4.2.2 Upper Tropospheric Water Vapor, in WG1: The Physical Science Basis, https://archive.ipcc.ch/publications_and_data/ar4/wg1/en/ch3s3-4-2-2.html\n - Shi, L., Schreck III, C. J., John, V. O., Chung, E.-S., Lang, T., Buehler, S. A., and Soden, B. J., 2022: Assessing the consistency of satellite-derived upper tropospheric humidity measurements, Atmos. Meas. Tech., 15, 6949–6963, https://doi.org/10.5194/amt-15-6949-2022\n - Chung E., B. Soden, B.J. Sohn, and L. Shi, 2014: Upper-tropospheric moistening in response to anthropogenic warming, Proc. Natl. Acad. Sci. U.S.A. 111 (32) 11636-11641, https://doi.org/10.1073/pnas.1409659111", "text_with_prefix": "EQC Quality Assessment: \"Studying the influence of atmospheric circulation on upper tropospheric humidity\"\nDataset: satellite-upper-troposphere-humidity [CDS]\nAspect: validation_q01 | Category: Satellite_ECVs\nSection: Studying the influence of atmospheric circulation on upper tropospheric humidity > ℹ️ If you want to know more > Key resources\n---\nSome key resources and further reading were linked throughout this assessment.\n\nThe CDS catalogue entries for the data used is:\n- Upper tropospheric humidity gridded data from 1999 to present derived from satellite observations:\nhttps://cds.climate.copernicus.eu/datasets/satellite-upper-troposphere-humidity\n\n- ERA5 hourly data on pressure levels from 1940 to present:\nhttps://cds.climate.copernicus.eu/datasets/reanalysis-era5-pressure-levels\n\nCode libraries used:\n- [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)\n\nAnd more resources on tropical atmosphere, large-scale circulation and upper tropospheric humidity :\n\n- https://www.nature.com/scitable/knowledge/library/tropical-weather-84224797/\n\n- Additonal scientific articles on Upper Tropospheric Humidity :\n\n- IPCC Assessment Report 4, 2007 : 3.4.2.2 Upper Tropospheric Water Vapor, in WG1: The Physical Science Basis, https://archive.ipcc.ch/publications_and_data/ar4/wg1/en/ch3s3-4-2-2.html\n - Shi, L., Schreck III, C. J., John, V. O., Chung, E.-S., Lang, T., Buehler, S. A., and Soden, B. J., 2022: Assessing the consistency of satellite-derived upper tropospheric humidity measurements, Atmos. Meas. Tech., 15, 6949–6963, https://doi.org/10.5194/amt-15-6949-2022\n - Chung E., B. Soden, B.J. Sohn, and L. Shi, 2014: Upper-tropospheric moistening in response to anthropogenic warming, Proc. Natl. Acad. Sci. U.S.A. 111 (32) 11636-11641, https://doi.org/10.1073/pnas.1409659111"} {"chunk_id": "satellite_satellite-upper-troposphere-humidity_validation_q01__b03f12feeb48", "report_id": "satellite_satellite-upper-troposphere-humidity_validation_q01", "dataset_id": "satellite-upper-troposphere-humidity", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q01", "aspect_base": "validation", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Studying the influence of atmospheric circulation on upper tropospheric humidity > ℹ️ If you want to know more > References", "title": "Studying the influence of atmospheric circulation on upper tropospheric humidity", "chunk_index": 17, "token_count": 980, "text_raw": "[[1]](https://doi.org/10.1029/97GL03563) Pierrehumbert R. T (1998), Lateral mixing as a source of subtropical water vapor, Geophys. Res. Lett., 25, 151-154, doi:10.1029/97GL03563\n\n[[2]](https://doi.org/10.1029/2006GL029118) Brogniez, H., and R. T. Pierrehumbert (2007), Intercomparison of tropical tropospheric humidity in GCMs with AMSU-B water vapor data, Geophys. Res. Lett., 34, L17812, doi:10.1029/2006GL029118.\n\n[[3]](https://doi.org/10.1007/s00382-003-0369-6) Bony, S., Dufresne, JL., Le Treut, H. et al. (2004), On dynamic and thermodynamic components of cloud changes. Climate Dynamics 22, 71–86. doi:10.1007/s00382-003-0369-6\n\n[[4]](https://doi.org/10.1175/1520-0450%281988%29027%3C0889:EOTUTR%3E2.0.CO;2) Schmetz J. and O. Turpeinen (1988), Estimation of the Upper Tropospheric Relative Humidity Field from METEOSAT Water Vapor Image Data, J. Appl. Meteor. Clim., 27, 8, 889-899. doi:10.1175/1520-0450(1988)027%3C0889:EOTUTR%3E2.0.CO;2.\n\n[[5]](https://doi.org/10.1029/93JD01283) Soden B. and F. Bretherton (1993), Upper tropospheric relative humidity from the GOES 6.7 μm channel: Method and climatology for July 1987, J. Geophys. Res, 98, D9, 16,669-16,688. doi:/10.1029/93JD01283.\n\n[[6]](https://doi.org/10.1029/2007JD009314) Buehler, S. A., M. Kuvatov, V. O. John, M. Milz, B. J. Soden, D. L. Jackson, and J. Notholt (2008), An upper tropospheric humidity data set from operational satellite microwave data, J. Geophys. Res., 113, D14110, doi:10.1029/2007JD009314.\n\n[[7]](https://doi.org/10.1175/2009JCLI2963.1) Brogniez H., R. Roca and L. Picon (2009), A study of the free tropospheric humidity interannual variability using Meteosat data and an advection–condensation transport model, J. Clim., 22, 24, 6773–6787, doi:10.1175/2009JCLI2963.1.\n\n[[8]](https://doi.org/10.5194/acp-14-11129-2014) Schröder, M., R. Roca, L. Picon, A. Kniffka, and H. Brogniez (2014) Climatology of free tropospheric humidity: extension into the SEVIRI era, evaluation and exemplary analysis, Atmos. Chem. Phys., 14, 11129-11148, doi:10.5194/acp-14-11129-2014.\n\n[[9]](https://doi.org/10.5194/acp-7-2489-2007) Chung E. S., B. J. Sohn, J. Schmetz and M. Koening (2007) Diurnal variation of upper tropospheric humidity and its relations to convective activities over tropical Africa, Atmos. Chem. Phys., 7, 2489–2502, doi:10.5194/acp-7-2489-2007.", "text_with_prefix": "EQC Quality Assessment: \"Studying the influence of atmospheric circulation on upper tropospheric humidity\"\nDataset: satellite-upper-troposphere-humidity [CDS]\nAspect: validation_q01 | Category: Satellite_ECVs\nSection: Studying the influence of atmospheric circulation on upper tropospheric humidity > ℹ️ If you want to know more > References\n---\n[[1]](https://doi.org/10.1029/97GL03563) Pierrehumbert R. T (1998), Lateral mixing as a source of subtropical water vapor, Geophys. Res. Lett., 25, 151-154, doi:10.1029/97GL03563\n\n[[2]](https://doi.org/10.1029/2006GL029118) Brogniez, H., and R. T. Pierrehumbert (2007), Intercomparison of tropical tropospheric humidity in GCMs with AMSU-B water vapor data, Geophys. Res. Lett., 34, L17812, doi:10.1029/2006GL029118.\n\n[[3]](https://doi.org/10.1007/s00382-003-0369-6) Bony, S., Dufresne, JL., Le Treut, H. et al. (2004), On dynamic and thermodynamic components of cloud changes. Climate Dynamics 22, 71–86. doi:10.1007/s00382-003-0369-6\n\n[[4]](https://doi.org/10.1175/1520-0450%281988%29027%3C0889:EOTUTR%3E2.0.CO;2) Schmetz J. and O. Turpeinen (1988), Estimation of the Upper Tropospheric Relative Humidity Field from METEOSAT Water Vapor Image Data, J. Appl. Meteor. Clim., 27, 8, 889-899. doi:10.1175/1520-0450(1988)027%3C0889:EOTUTR%3E2.0.CO;2.\n\n[[5]](https://doi.org/10.1029/93JD01283) Soden B. and F. Bretherton (1993), Upper tropospheric relative humidity from the GOES 6.7 μm channel: Method and climatology for July 1987, J. Geophys. Res, 98, D9, 16,669-16,688. doi:/10.1029/93JD01283.\n\n[[6]](https://doi.org/10.1029/2007JD009314) Buehler, S. A., M. Kuvatov, V. O. John, M. Milz, B. J. Soden, D. L. Jackson, and J. Notholt (2008), An upper tropospheric humidity data set from operational satellite microwave data, J. Geophys. Res., 113, D14110, doi:10.1029/2007JD009314.\n\n[[7]](https://doi.org/10.1175/2009JCLI2963.1) Brogniez H., R. Roca and L. Picon (2009), A study of the free tropospheric humidity interannual variability using Meteosat data and an advection–condensation transport model, J. Clim., 22, 24, 6773–6787, doi:10.1175/2009JCLI2963.1.\n\n[[8]](https://doi.org/10.5194/acp-14-11129-2014) Schröder, M., R. Roca, L. Picon, A. Kniffka, and H. Brogniez (2014) Climatology of free tropospheric humidity: extension into the SEVIRI era, evaluation and exemplary analysis, Atmos. Chem. Phys., 14, 11129-11148, doi:10.5194/acp-14-11129-2014.\n\n[[9]](https://doi.org/10.5194/acp-7-2489-2007) Chung E. S., B. J. Sohn, J. Schmetz and M. Koening (2007) Diurnal variation of upper tropospheric humidity and its relations to convective activities over tropical Africa, Atmos. Chem. Phys., 7, 2489–2502, doi:10.5194/acp-7-2489-2007."} {"chunk_id": "satellite_satellite-upper-troposphere-humidity_validation_q01__fc4b57c8b454", "report_id": "satellite_satellite-upper-troposphere-humidity_validation_q01", "dataset_id": "satellite-upper-troposphere-humidity", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q01", "aspect_base": "validation", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Studying the influence of atmospheric circulation on upper tropospheric humidity > ℹ️ If you want to know more > References", "title": "Studying the influence of atmospheric circulation on upper tropospheric humidity", "chunk_index": 18, "token_count": 848, "text_raw": "humidity and its relations to convective activities over tropical Africa, Atmos. Chem. Phys., 7, 2489–2502, doi:10.5194/acp-7-2489-2007.\n\n[[10]](https://doi.org/10.1175/JCLI-D-19-0046.1) Tivig M., V. Grützun, V. O. John and S. Buehler (2020), Trends in upper-tropospheric humidity: expansion of the subtropical dry zones? J. Clim. 33, 6, 2149–2161, doi:10.1175/JCLI-D-19-0046.1.\n\n[[11]](https://doi.org/10.1256/qj.04.176) Uppala S. and co authors (2005), The ERA-40 re-analysis, Q. J. R. Meteorol. Soc., 31, 2961– 3012, doi: 10.1256/qj.04.176.\n\n[[12]](https://doi.org/10.1175/1520-0477%281996%29077<0437:TNYRP>2.0.CO;2) Kalnay E. and co authors (1996) The NCEP/NCAR 40-year reanalysis project. Bull. Am. Meteorol. Soc., 77, 437–471, doi:10.1175/1520-0477(1996)077<0437:TNYRP>2.0.CO;2.\n\n[[13]](https://doi.org/10.1088/1748-9326/5/2/025205) Allan R., B. Soden, V. John, W. Ingram and P. Good (2010) Current changes in tropical precipitation, Environ. Res. Lett., 5, 025205, doi:10.1088/1748-9326/5/2/025205.\n\n[[14]](https://doi.org/10.1175/1520-0477%282000%29081%3C0797:AIORCF%3E2.3.CO;2) Soden, B., and coauthors (2000), An Intercomparison of Radiation Codes for Retrieving Upper-Tropospheric Humidity in the 6.3-μm Band: A Report from the First GVaP Workshop. Bull. Amer. Meteor. Soc., 81, 797–808. doi:10.1175/1520-0477%282000%29081<0797:AIORCF>2.3.CO;2\n\n[[15]](https://www.wcrp-climate.org/resources/wcrp-publications/1095-pub-2017) Schroeder, M., M. Lockhoff, L. Shi, T. August, R. Bennartz, E. Borbas, H. Brogniez, X. Calbet, S. Crewell, S. Eikenberg, F. Fell, J. Forsythe, A. Gambacorta, K. Graw, S.-P. Ho, H. Hoeschen, J. Kinzel, E. R. Kursinski, A. Reale, J. Roman, N. Scott, S. Steinke, B. Sun, T. Trent, A. Walther, U. Willen, Q. Yang, 2017: GEWEX water vapor assessment (G-VAP). WCRP Report 16/2017, World Climate Research Programme, Geneva, Switzerland, 216 pp.", "text_with_prefix": "EQC Quality Assessment: \"Studying the influence of atmospheric circulation on upper tropospheric humidity\"\nDataset: satellite-upper-troposphere-humidity [CDS]\nAspect: validation_q01 | Category: Satellite_ECVs\nSection: Studying the influence of atmospheric circulation on upper tropospheric humidity > ℹ️ If you want to know more > References\n---\nhumidity and its relations to convective activities over tropical Africa, Atmos. Chem. Phys., 7, 2489–2502, doi:10.5194/acp-7-2489-2007.\n\n[[10]](https://doi.org/10.1175/JCLI-D-19-0046.1) Tivig M., V. Grützun, V. O. John and S. Buehler (2020), Trends in upper-tropospheric humidity: expansion of the subtropical dry zones? J. Clim. 33, 6, 2149–2161, doi:10.1175/JCLI-D-19-0046.1.\n\n[[11]](https://doi.org/10.1256/qj.04.176) Uppala S. and co authors (2005), The ERA-40 re-analysis, Q. J. R. Meteorol. Soc., 31, 2961– 3012, doi: 10.1256/qj.04.176.\n\n[[12]](https://doi.org/10.1175/1520-0477%281996%29077<0437:TNYRP>2.0.CO;2) Kalnay E. and co authors (1996) The NCEP/NCAR 40-year reanalysis project. Bull. Am. Meteorol. Soc., 77, 437–471, doi:10.1175/1520-0477(1996)077<0437:TNYRP>2.0.CO;2.\n\n[[13]](https://doi.org/10.1088/1748-9326/5/2/025205) Allan R., B. Soden, V. John, W. Ingram and P. Good (2010) Current changes in tropical precipitation, Environ. Res. Lett., 5, 025205, doi:10.1088/1748-9326/5/2/025205.\n\n[[14]](https://doi.org/10.1175/1520-0477%282000%29081%3C0797:AIORCF%3E2.3.CO;2) Soden, B., and coauthors (2000), An Intercomparison of Radiation Codes for Retrieving Upper-Tropospheric Humidity in the 6.3-μm Band: A Report from the First GVaP Workshop. Bull. Amer. Meteor. Soc., 81, 797–808. doi:10.1175/1520-0477%282000%29081<0797:AIORCF>2.3.CO;2\n\n[[15]](https://www.wcrp-climate.org/resources/wcrp-publications/1095-pub-2017) Schroeder, M., M. Lockhoff, L. Shi, T. August, R. Bennartz, E. Borbas, H. Brogniez, X. Calbet, S. Crewell, S. Eikenberg, F. Fell, J. Forsythe, A. Gambacorta, K. Graw, S.-P. Ho, H. Hoeschen, J. Kinzel, E. R. Kursinski, A. Reale, J. Roman, N. Scott, S. Steinke, B. Sun, T. Trent, A. Walther, U. Willen, Q. Yang, 2017: GEWEX water vapor assessment (G-VAP). WCRP Report 16/2017, World Climate Research Programme, Geneva, Switzerland, 216 pp."} {"chunk_id": "satellite_satellite-aerosol-properties_consistency_q01__a98f9b9eece6", "report_id": "satellite_satellite-aerosol-properties_consistency_q01", "dataset_id": "satellite-aerosol-properties", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q01", "aspect_base": "consistency", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Assessment of Seasonal Variability and Completeness of Aerosol Optical Depth from SLSTR Satellite Data (2017-2023)", "title": "Assessment of Seasonal Variability and Completeness of Aerosol Optical Depth from SLSTR Satellite Data (2017-2023)", "chunk_index": 0, "token_count": 138, "text_raw": " \n\nProduction date: 30-09-2024\n\nProduced by: Consiglio Nazionale delle Ricerche ([CNR](https://www.cnr.it/en)- Fabio Madonna and Faezeh Karimian.)", "text_with_prefix": "EQC Quality Assessment: \"Assessment of Seasonal Variability and Completeness of Aerosol Optical Depth from SLSTR Satellite Data (2017-2023)\"\nDataset: satellite-aerosol-properties [CDS]\nAspect: consistency_q01 | Category: Satellite_ECVs\nSection: Assessment of Seasonal Variability and Completeness of Aerosol Optical Depth from SLSTR Satellite Data (2017-2023)\n---\n \n\nProduction date: 30-09-2024\n\nProduced by: Consiglio Nazionale delle Ricerche ([CNR](https://www.cnr.it/en)- Fabio Madonna and Faezeh Karimian.)"} {"chunk_id": "satellite_satellite-aerosol-properties_consistency_q01__d11648162241", "report_id": "satellite_satellite-aerosol-properties_consistency_q01", "dataset_id": "satellite-aerosol-properties", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q01", "aspect_base": "consistency", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Assessment of Seasonal Variability and Completeness of Aerosol Optical Depth from SLSTR Satellite Data (2017-2023) > Quality assessment questions", "title": "Assessment of Seasonal Variability and Completeness of Aerosol Optical Depth from SLSTR Satellite Data (2017-2023)", "chunk_index": 1, "token_count": 517, "text_raw": "**• Can satellite data consistently and homogeneously capture the seasonal variability of AOD in different parts of the world?**\n\n**• Is the temporal coverage of aerosol satellite data sufficient to characterize the aerosol global distribution in the period 2017-2023?**\n\nIn the context of climate change and natural atmospheric variability, understanding the spatial and temporal distribution of aerosols is crucial for analyzing phenomena such as air quality, global warming, and public health. Seasonal variations in aerosols are influenced by numerous factors, including biological cycles, seasonal fires, desert dust, and industrial activities [1, 2, 3]. However, uncertainties in satellite data, stemming from limitations in temporal coverage and spatial resolution, could affect our ability to draw reliable conclusions [1].\n\nThis assessment uses the SLSTR aerosol ensemble product at Level-2 super-pixel resolution (10 × 10 km²), an uncertainty-weighted average of the SDVV, ORAC and Swansea retrievals, to evaluate its ability to consistently and homogeneously capture the seasonal variability of Aerosol Optical Depth (AOD) across different global regions. The central question we aim to answer is whether satellite data, despite their limitations and uncertainties, can accurately capture the atmospheric dynamics that influence the distribution of aerosols over the years [2, 4].\n\nThe dataset used in this study is the CDS (Climate Data Store) AOD product derived from SLSTR onboard Sentinel-3, providing monthly gridded AOD values at 550 nm at approximately 1° spatial resolution in NetCDF format, covering the period 2017–2023.\n\nIn summary, this study investigates whether SLSTR data can be reliably used to monitor the seasonal variability of AOD by comparing different geographic regions and assessing whether the spatial coverage is sufficient to characterize AOD distribution globally.\n\nThe results show that the satellite-derived AOD estimates are consistent with other findings reported in the literature [1, 4, 5] and are supported by a thorough scientific analysis. This consistency suggests that SLSTR data can be a robust tool for monitoring global aerosol distribution.", "text_with_prefix": "EQC Quality Assessment: \"Assessment of Seasonal Variability and Completeness of Aerosol Optical Depth from SLSTR Satellite Data (2017-2023)\"\nDataset: satellite-aerosol-properties [CDS]\nAspect: consistency_q01 | Category: Satellite_ECVs\nSection: Assessment of Seasonal Variability and Completeness of Aerosol Optical Depth from SLSTR Satellite Data (2017-2023) > Quality assessment questions\n---\n**• Can satellite data consistently and homogeneously capture the seasonal variability of AOD in different parts of the world?**\n\n**• Is the temporal coverage of aerosol satellite data sufficient to characterize the aerosol global distribution in the period 2017-2023?**\n\nIn the context of climate change and natural atmospheric variability, understanding the spatial and temporal distribution of aerosols is crucial for analyzing phenomena such as air quality, global warming, and public health. Seasonal variations in aerosols are influenced by numerous factors, including biological cycles, seasonal fires, desert dust, and industrial activities [1, 2, 3]. However, uncertainties in satellite data, stemming from limitations in temporal coverage and spatial resolution, could affect our ability to draw reliable conclusions [1].\n\nThis assessment uses the SLSTR aerosol ensemble product at Level-2 super-pixel resolution (10 × 10 km²), an uncertainty-weighted average of the SDVV, ORAC and Swansea retrievals, to evaluate its ability to consistently and homogeneously capture the seasonal variability of Aerosol Optical Depth (AOD) across different global regions. The central question we aim to answer is whether satellite data, despite their limitations and uncertainties, can accurately capture the atmospheric dynamics that influence the distribution of aerosols over the years [2, 4].\n\nThe dataset used in this study is the CDS (Climate Data Store) AOD product derived from SLSTR onboard Sentinel-3, providing monthly gridded AOD values at 550 nm at approximately 1° spatial resolution in NetCDF format, covering the period 2017–2023.\n\nIn summary, this study investigates whether SLSTR data can be reliably used to monitor the seasonal variability of AOD by comparing different geographic regions and assessing whether the spatial coverage is sufficient to characterize AOD distribution globally.\n\nThe results show that the satellite-derived AOD estimates are consistent with other findings reported in the literature [1, 4, 5] and are supported by a thorough scientific analysis. This consistency suggests that SLSTR data can be a robust tool for monitoring global aerosol distribution."} {"chunk_id": "satellite_satellite-aerosol-properties_consistency_q01__63f7b0fe3963", "report_id": "satellite_satellite-aerosol-properties_consistency_q01", "dataset_id": "satellite-aerosol-properties", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q01", "aspect_base": "consistency", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Assessment of Seasonal Variability and Completeness of Aerosol Optical Depth from SLSTR Satellite Data (2017-2023) > Quality assessment statement", "title": "Assessment of Seasonal Variability and Completeness of Aerosol Optical Depth from SLSTR Satellite Data (2017-2023)", "chunk_index": 2, "token_count": 672, "text_raw": "These are the key outcomes of this assessment\n\n• The SLSTR aerosol data exhibits consistency with other satellite-derived datasets, particularly in regions like Southern Africa and Southeast Asia [3, 6].\n\n• Coarse-mode AOD trends over North Africa and the Mediterranean Basin align with findings from previous studies, such as Cuevas-Agulló et al. (2024), highlighting an increase in dust intrusions during 2020–2022 [2].\n\n• The absence of data for several months in East Asia and Siberia restricts a full understanding of regional aerosol dynamics.\n\n• A weaker correlation between Fine Mode AOD and Total AOD in regions like North Africa and India suggests the presence of varying aerosol types and the influence of mineral dust transport.\n\n## 📋 Methodology\n\nThis assessment aims to assess the consistency and completeness of SLSTR-derived satellite data in capturing the seasonal variability of Aerosol Optical Depth (AOD) over different global regions from 2017 to 2023.\n\nThe methodology adopted for the analysis is split into the following steps:\n\n[](section-1) \n* Import all required libraries \n* Spatial and temporal definitions\n\n[](section-2) \n* Download AOD data \n* Computation of time series \n* Define required functions\n\n[](section-3) \n* Prepare the dataframe \n* Global map plots: mean AOD and FM-AOD\n* Mapping of subregions\n* Regional AOD and FM-AOD time series plots \n* Observation count plot for AOD550 averages \n* Scatter plot of fine-mode AOD vs. total AOD\n\n[](section-4)\n\n## 📈 Analysis and results\n(section-1)=\n### Choose the data to use and set up the code\n\n#### Import all required libraries\nIn this section, we import all the relevant packages needed for running the notebook.\n\n#### Spatial and temporal definitions\n\nThe analysis performed in this assessment focuses on data from July 2017 to June 2023, using monthly averages from the SLSTR sensor on Sentinel-3A. The regions are grouped into meaningful areas like North America, Europe, Africa, and Asia, each defined by specific latitude and longitude ranges. These subregions are color-coded for easy comparison in plots.\n\nAssign time dimension and compute means\nOptional color palette for consistency\nDefine subregions\n\n(section-2)=\n### Data Retrieval and Preparation\n\nIn this step, the selected satellite data is downloaded using predefined time and variable settings. The data is then formatted to include a clear time coordinate, and average maps of key variables like AOD550 and FM_AOD550 are calculated over the full time period to prepare for analysis and visualization.\n\n#### Download AOD data\n\n```text\n100%|██████████| 72/72 [00:07<00:00, 9.06it/s]\n```", "text_with_prefix": "EQC Quality Assessment: \"Assessment of Seasonal Variability and Completeness of Aerosol Optical Depth from SLSTR Satellite Data (2017-2023)\"\nDataset: satellite-aerosol-properties [CDS]\nAspect: consistency_q01 | Category: Satellite_ECVs\nSection: Assessment of Seasonal Variability and Completeness of Aerosol Optical Depth from SLSTR Satellite Data (2017-2023) > Quality assessment statement\n---\nThese are the key outcomes of this assessment\n\n• The SLSTR aerosol data exhibits consistency with other satellite-derived datasets, particularly in regions like Southern Africa and Southeast Asia [3, 6].\n\n• Coarse-mode AOD trends over North Africa and the Mediterranean Basin align with findings from previous studies, such as Cuevas-Agulló et al. (2024), highlighting an increase in dust intrusions during 2020–2022 [2].\n\n• The absence of data for several months in East Asia and Siberia restricts a full understanding of regional aerosol dynamics.\n\n• A weaker correlation between Fine Mode AOD and Total AOD in regions like North Africa and India suggests the presence of varying aerosol types and the influence of mineral dust transport.\n\n## 📋 Methodology\n\nThis assessment aims to assess the consistency and completeness of SLSTR-derived satellite data in capturing the seasonal variability of Aerosol Optical Depth (AOD) over different global regions from 2017 to 2023.\n\nThe methodology adopted for the analysis is split into the following steps:\n\n[](section-1) \n* Import all required libraries \n* Spatial and temporal definitions\n\n[](section-2) \n* Download AOD data \n* Computation of time series \n* Define required functions\n\n[](section-3) \n* Prepare the dataframe \n* Global map plots: mean AOD and FM-AOD\n* Mapping of subregions\n* Regional AOD and FM-AOD time series plots \n* Observation count plot for AOD550 averages \n* Scatter plot of fine-mode AOD vs. total AOD\n\n[](section-4)\n\n## 📈 Analysis and results\n(section-1)=\n### Choose the data to use and set up the code\n\n#### Import all required libraries\nIn this section, we import all the relevant packages needed for running the notebook.\n\n#### Spatial and temporal definitions\n\nThe analysis performed in this assessment focuses on data from July 2017 to June 2023, using monthly averages from the SLSTR sensor on Sentinel-3A. The regions are grouped into meaningful areas like North America, Europe, Africa, and Asia, each defined by specific latitude and longitude ranges. These subregions are color-coded for easy comparison in plots.\n\nAssign time dimension and compute means\nOptional color palette for consistency\nDefine subregions\n\n(section-2)=\n### Data Retrieval and Preparation\n\nIn this step, the selected satellite data is downloaded using predefined time and variable settings. The data is then formatted to include a clear time coordinate, and average maps of key variables like AOD550 and FM_AOD550 are calculated over the full time period to prepare for analysis and visualization.\n\n#### Download AOD data\n\n```text\n100%|██████████| 72/72 [00:07<00:00, 9.06it/s]\n```"} {"chunk_id": "satellite_satellite-aerosol-properties_consistency_q01__df758a11f61c", "report_id": "satellite_satellite-aerosol-properties_consistency_q01", "dataset_id": "satellite-aerosol-properties", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q01", "aspect_base": "consistency", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Assessment of Seasonal Variability and Completeness of Aerosol Optical Depth from SLSTR Satellite Data (2017-2023) > Quality assessment statement", "title": "Assessment of Seasonal Variability and Completeness of Aerosol Optical Depth from SLSTR Satellite Data (2017-2023)", "chunk_index": 3, "token_count": 1111, "text_raw": "%|██████████| 72/72 [00:07<00:00, 9.06it/s]\n```\n\n```text\n Size: 75MB\nDimensions: (source: 72, latitude: 180, longitude: 360)\nCoordinates:\n * source (source) object 576B '148c29e9a49fd1261d149a...\n * latitude (latitude) float32 720B -89.5 -88.5 ... 89.5\n * longitude (longitude) float32 1kB -179.5 -178.5 ... 179.5\nData variables:\n AOD550 (source, latitude, longitude) float32 19MB dask.array\n FM_AOD550 (source, latitude, longitude) float32 19MB dask.array\n AOD550_UNCERTAINTY_ENSEMBLE (source, latitude, longitude) float32 19MB dask.array\n NMEAS (source, latitude, longitude) float32 19MB dask.array\nAttributes: (12/18)\n Conventions: CF-1.6\n creator_email: thomas.popp@dlr.de\n creator_name: German Aerospace Center, DFD\n geospatial_lat_max: 90.0\n geospatial_lat_min: -90.0\n geospatial_lon_max: 180.0\n ... ...\n sensor: SLSTR\n standard_name_vocabulary: NetCDF Climate and Forecast (CF) Metadata Conv...\n summary: Level 3 aerosol properties retreived using sat...\n title: Ensemble aerosol product level 3\n tracking_id: d79c7d5c-a788-11ee-9e43-0050569370b0\n version: v2.3\n```\n\n#### Computation of time series\n\nTo analyze changes over time, a monthly time coordinate was created and assigned to the dataset. The data’s internal “source” dimension was replaced with this new time axis, allowing for clear temporal analysis. Then, the mean values of AOD550 and FM_AOD550 were calculated across the full time period, producing overall average maps for these key aerosol indicators.\n\nAssign time dimension and compute means\nAdd time as a coordinate based on the 'source' dimension\n\n#### Define required functions\n\nThe following functions were defined to support the analysis of aerosol data:\n\nThe function \"plot_global_map\" creates a global map of AOD values, adding coastlines, borders, and gridlines for geographic context.\n\n\"plot_multi_series_aod\" generates seasonal time series for each subregion, showing trends in AOD or fine-mode AOD, with uncertainty as a shading.\n\n\"plot_observation_counts_with_stats_on_plot\" shows how many data points were used in each region per month, helping assess data coverage.\n\n\"scatter_fine_vs_total\" creates scatter plots comparing total AOD with fine-mode AOD, including regression lines and R² values to reveal correlations across regions.\n\nCollect all monthly means to find global y-limits\nConcatenate all monthly mean series and find min/max ignoring NaNs\nOptionally, add a small margin for visual comfort\nCalculate area in km² and normalize\nOptionally set upper limit to 100 if values exceed it\nPlot 1:1 line and set axis limits\nRemove any empty axes\n\n(section-3)=\n### Plot and describe the results\n\n#### Global map plots: mean AOD and FM-AOD\n\nThe first map shows the mean Aerosol Optical Depth (AOD550) over the globe. Higher AOD values, shown in darker orange, are observed over regions with frequent dust storms (e.g. the Sahara), biomass burning (e.g. central Africa), and industrial pollution (e.g. parts of India and China).\n\nThe second map displays the mean Fine-mode AOD550, which focuses on smaller particles like smoke, urban pollution, and secondary aerosols, dominating aerosol loading over central Africa, Southeast Asia, and parts of South America. These areas associated with biomass burning and anthropogenic emissions.\n\nTogether, these plots help distinguish between regions dominated by coarse-mode aerosols (e.g. dust) and those with higher concentrations of fine-mode particles.\n\n*Figure 1.* Global distribution of mean Aerosol Optical Depth (AOD550) and Fine-mode AOD550 from the SLSTR sensor onboard Sentinel-3A, averaged over the period July 2017 to June 2023.", "text_with_prefix": "EQC Quality Assessment: \"Assessment of Seasonal Variability and Completeness of Aerosol Optical Depth from SLSTR Satellite Data (2017-2023)\"\nDataset: satellite-aerosol-properties [CDS]\nAspect: consistency_q01 | Category: Satellite_ECVs\nSection: Assessment of Seasonal Variability and Completeness of Aerosol Optical Depth from SLSTR Satellite Data (2017-2023) > Quality assessment statement\n---\n%|██████████| 72/72 [00:07<00:00, 9.06it/s]\n```\n\n```text\n Size: 75MB\nDimensions: (source: 72, latitude: 180, longitude: 360)\nCoordinates:\n * source (source) object 576B '148c29e9a49fd1261d149a...\n * latitude (latitude) float32 720B -89.5 -88.5 ... 89.5\n * longitude (longitude) float32 1kB -179.5 -178.5 ... 179.5\nData variables:\n AOD550 (source, latitude, longitude) float32 19MB dask.array\n FM_AOD550 (source, latitude, longitude) float32 19MB dask.array\n AOD550_UNCERTAINTY_ENSEMBLE (source, latitude, longitude) float32 19MB dask.array\n NMEAS (source, latitude, longitude) float32 19MB dask.array\nAttributes: (12/18)\n Conventions: CF-1.6\n creator_email: thomas.popp@dlr.de\n creator_name: German Aerospace Center, DFD\n geospatial_lat_max: 90.0\n geospatial_lat_min: -90.0\n geospatial_lon_max: 180.0\n ... ...\n sensor: SLSTR\n standard_name_vocabulary: NetCDF Climate and Forecast (CF) Metadata Conv...\n summary: Level 3 aerosol properties retreived using sat...\n title: Ensemble aerosol product level 3\n tracking_id: d79c7d5c-a788-11ee-9e43-0050569370b0\n version: v2.3\n```\n\n#### Computation of time series\n\nTo analyze changes over time, a monthly time coordinate was created and assigned to the dataset. The data’s internal “source” dimension was replaced with this new time axis, allowing for clear temporal analysis. Then, the mean values of AOD550 and FM_AOD550 were calculated across the full time period, producing overall average maps for these key aerosol indicators.\n\nAssign time dimension and compute means\nAdd time as a coordinate based on the 'source' dimension\n\n#### Define required functions\n\nThe following functions were defined to support the analysis of aerosol data:\n\nThe function \"plot_global_map\" creates a global map of AOD values, adding coastlines, borders, and gridlines for geographic context.\n\n\"plot_multi_series_aod\" generates seasonal time series for each subregion, showing trends in AOD or fine-mode AOD, with uncertainty as a shading.\n\n\"plot_observation_counts_with_stats_on_plot\" shows how many data points were used in each region per month, helping assess data coverage.\n\n\"scatter_fine_vs_total\" creates scatter plots comparing total AOD with fine-mode AOD, including regression lines and R² values to reveal correlations across regions.\n\nCollect all monthly means to find global y-limits\nConcatenate all monthly mean series and find min/max ignoring NaNs\nOptionally, add a small margin for visual comfort\nCalculate area in km² and normalize\nOptionally set upper limit to 100 if values exceed it\nPlot 1:1 line and set axis limits\nRemove any empty axes\n\n(section-3)=\n### Plot and describe the results\n\n#### Global map plots: mean AOD and FM-AOD\n\nThe first map shows the mean Aerosol Optical Depth (AOD550) over the globe. Higher AOD values, shown in darker orange, are observed over regions with frequent dust storms (e.g. the Sahara), biomass burning (e.g. central Africa), and industrial pollution (e.g. parts of India and China).\n\nThe second map displays the mean Fine-mode AOD550, which focuses on smaller particles like smoke, urban pollution, and secondary aerosols, dominating aerosol loading over central Africa, Southeast Asia, and parts of South America. These areas associated with biomass burning and anthropogenic emissions.\n\nTogether, these plots help distinguish between regions dominated by coarse-mode aerosols (e.g. dust) and those with higher concentrations of fine-mode particles.\n\n*Figure 1.* Global distribution of mean Aerosol Optical Depth (AOD550) and Fine-mode AOD550 from the SLSTR sensor onboard Sentinel-3A, averaged over the period July 2017 to June 2023."} {"chunk_id": "satellite_satellite-aerosol-properties_consistency_q01__3102d96e70b1", "report_id": "satellite_satellite-aerosol-properties_consistency_q01", "dataset_id": "satellite-aerosol-properties", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q01", "aspect_base": "consistency", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Assessment of Seasonal Variability and Completeness of Aerosol Optical Depth from SLSTR Satellite Data (2017-2023) > Quality assessment statement", "title": "Assessment of Seasonal Variability and Completeness of Aerosol Optical Depth from SLSTR Satellite Data (2017-2023)", "chunk_index": 4, "token_count": 1113, "text_raw": "*Figure 1.* Global distribution of mean Aerosol Optical Depth (AOD550) and Fine-mode AOD550 from the SLSTR sensor onboard Sentinel-3A, averaged over the period July 2017 to June 2023.\n\n#### Mapping of subregions\nThis section defines the geographic boundaries of the subregions used for regional aerosol analysis. Each domain is represented as a simple rectangular box defined by latitude and longitude limits, enabling consistent spatial aggregation of Aerosol Optical Depth (AOD) data across different parts of the globe. The map visualizes these subregions to aid interpretation of subsequent regional time series and correlation analyses.\n\n*Figure 2.* The spatial domains of the subregions used in the aerosol analysis. Each red rectangle represents the geographic boundaries defined by latitude and longitude for a specific subregion.\n\n#### Regional AOD and FM-AOD time series plots\nThe time series of 550 nm AOD monthly means from SLSTR satellite data refer to subregions selected because of the similarity in the observed aerosol types: Northern Industrial Economies, Southern Biomass Burning, Dust-Dominated Regions, and Asia. Each subplot includes shaded uncertainty areas (±1σ) quantifying AOD variability.\nIn the AOD550 plot, Northern industrial regions display often have AOD peaks in summertime, while, since 2020 the Mediterranean Basin shows high AOD levels throughout the year, strongly influenced by the mineral dust intrusion consistently with a documented enhancement of dust emissions in Northern Africa [4].\nThe FM_AOD550 is also reported showing that biomass burning regions, like Southern Africa, have peaks during dry periods typically dominated by forest fires [2, 3].\n\nThe observed data gaps in regions such as Siberia primarily arise from the inherent limitations of satellite aerosol retrievals in polar and subpolar areas. According to the Product User Guide (PUG), the SLSTR aerosol product has reduced temporal coverage in these high-latitude regions due to the absence of sunlight during extended winter periods, which prevents reliable satellite observations (Table 6, footnote 1). This sunlight dependency means that aerosol data cannot be retrieved when there is no solar illumination, resulting in missing data during polar night [7].\n\n*Figure 3.* Monthly mean time series of total AOD at 550 nm from SLSTR for different sub-regions. Each panel shows regional time series with uncertainty as a shaded area (±1σ). Northern industrial regions show peaks in summer. The Mediterranean Basin maintains high AOD values year-round, especially in the period 2020-2022.\n\n*Figure 4.* Monthly mean time series of fine-mode AOD550 from SLSTR for selected global sub-regions. Fine mode aerosols include smoke, urban pollution, and secondary particles. Biomass burning regions like South America and Southern Africa show peaks during dry periods.\n\n#### Observation count plot for AOD550 averages\n\nFor assessing the spatial completeness of the data used in the time series analysis, this plot displays the normalized observation density, i.e., the number of valid AOD550 observations per 1,000,000 km², used each month to compute regional averages. Each line represents a distinct subregion, allowing visual comparison of data coverage over time, normalized by region size.\n\nMost regions, such as South America, exhibit consistently high normalized observation densities, typically ranging from 60 to 80 counts per 1,000,000 km². In contrast, areas like Siberia show larger variability and lower density, implying a higher uncertainty for the derived monthly statistics and trend assessments. \nThe lower number of aerosol observations in Siberia is mainly due to its high-latitude location and harsh environmental conditions. As highlighted in the Product User Guide (PUG), satellite retrievals rely on sunlight, which is limited or absent during the extended polar night in winter months, reducing the temporal coverage in this region [7].\n\n```text\n/data/wp5/.tmp/ipykernel_1257448/4247846240.py:144: MatplotlibDeprecationWarning: The get_cmap function was deprecated in Matplotlib 3.7 and will be removed in 3.11. Use ``matplotlib.colormaps[name]`` or ``matplotlib.colormaps.get_cmap()`` or ``pyplot.get_cmap()`` instead.\n colors = cm.get_cmap('tab20', len(all_regions))\n```\n\n```text\nY-axis limits before setting: (np.float64(-4.058112166220274), np.float64(85.22035549062575))\nY-axis limits after setting: (np.float64(-4.058112166220274), np.float64(85.22035549062575))\n```\n\n*Figure 5.* Normalized observation density for each sub-region, showing the number of valid observations per 1,000,000 km² over time. The maximum normalized observation density reached approximately 85 counts / 1,000,000 km².\n\n#### Scatter plot of Fine Mode AOD vs total AOD", "text_with_prefix": "EQC Quality Assessment: \"Assessment of Seasonal Variability and Completeness of Aerosol Optical Depth from SLSTR Satellite Data (2017-2023)\"\nDataset: satellite-aerosol-properties [CDS]\nAspect: consistency_q01 | Category: Satellite_ECVs\nSection: Assessment of Seasonal Variability and Completeness of Aerosol Optical Depth from SLSTR Satellite Data (2017-2023) > Quality assessment statement\n---\n*Figure 1.* Global distribution of mean Aerosol Optical Depth (AOD550) and Fine-mode AOD550 from the SLSTR sensor onboard Sentinel-3A, averaged over the period July 2017 to June 2023.\n\n#### Mapping of subregions\nThis section defines the geographic boundaries of the subregions used for regional aerosol analysis. Each domain is represented as a simple rectangular box defined by latitude and longitude limits, enabling consistent spatial aggregation of Aerosol Optical Depth (AOD) data across different parts of the globe. The map visualizes these subregions to aid interpretation of subsequent regional time series and correlation analyses.\n\n*Figure 2.* The spatial domains of the subregions used in the aerosol analysis. Each red rectangle represents the geographic boundaries defined by latitude and longitude for a specific subregion.\n\n#### Regional AOD and FM-AOD time series plots\nThe time series of 550 nm AOD monthly means from SLSTR satellite data refer to subregions selected because of the similarity in the observed aerosol types: Northern Industrial Economies, Southern Biomass Burning, Dust-Dominated Regions, and Asia. Each subplot includes shaded uncertainty areas (±1σ) quantifying AOD variability.\nIn the AOD550 plot, Northern industrial regions display often have AOD peaks in summertime, while, since 2020 the Mediterranean Basin shows high AOD levels throughout the year, strongly influenced by the mineral dust intrusion consistently with a documented enhancement of dust emissions in Northern Africa [4].\nThe FM_AOD550 is also reported showing that biomass burning regions, like Southern Africa, have peaks during dry periods typically dominated by forest fires [2, 3].\n\nThe observed data gaps in regions such as Siberia primarily arise from the inherent limitations of satellite aerosol retrievals in polar and subpolar areas. According to the Product User Guide (PUG), the SLSTR aerosol product has reduced temporal coverage in these high-latitude regions due to the absence of sunlight during extended winter periods, which prevents reliable satellite observations (Table 6, footnote 1). This sunlight dependency means that aerosol data cannot be retrieved when there is no solar illumination, resulting in missing data during polar night [7].\n\n*Figure 3.* Monthly mean time series of total AOD at 550 nm from SLSTR for different sub-regions. Each panel shows regional time series with uncertainty as a shaded area (±1σ). Northern industrial regions show peaks in summer. The Mediterranean Basin maintains high AOD values year-round, especially in the period 2020-2022.\n\n*Figure 4.* Monthly mean time series of fine-mode AOD550 from SLSTR for selected global sub-regions. Fine mode aerosols include smoke, urban pollution, and secondary particles. Biomass burning regions like South America and Southern Africa show peaks during dry periods.\n\n#### Observation count plot for AOD550 averages\n\nFor assessing the spatial completeness of the data used in the time series analysis, this plot displays the normalized observation density, i.e., the number of valid AOD550 observations per 1,000,000 km², used each month to compute regional averages. Each line represents a distinct subregion, allowing visual comparison of data coverage over time, normalized by region size.\n\nMost regions, such as South America, exhibit consistently high normalized observation densities, typically ranging from 60 to 80 counts per 1,000,000 km². In contrast, areas like Siberia show larger variability and lower density, implying a higher uncertainty for the derived monthly statistics and trend assessments. \nThe lower number of aerosol observations in Siberia is mainly due to its high-latitude location and harsh environmental conditions. As highlighted in the Product User Guide (PUG), satellite retrievals rely on sunlight, which is limited or absent during the extended polar night in winter months, reducing the temporal coverage in this region [7].\n\n```text\n/data/wp5/.tmp/ipykernel_1257448/4247846240.py:144: MatplotlibDeprecationWarning: The get_cmap function was deprecated in Matplotlib 3.7 and will be removed in 3.11. Use ``matplotlib.colormaps[name]`` or ``matplotlib.colormaps.get_cmap()`` or ``pyplot.get_cmap()`` instead.\n colors = cm.get_cmap('tab20', len(all_regions))\n```\n\n```text\nY-axis limits before setting: (np.float64(-4.058112166220274), np.float64(85.22035549062575))\nY-axis limits after setting: (np.float64(-4.058112166220274), np.float64(85.22035549062575))\n```\n\n*Figure 5.* Normalized observation density for each sub-region, showing the number of valid observations per 1,000,000 km² over time. The maximum normalized observation density reached approximately 85 counts / 1,000,000 km².\n\n#### Scatter plot of Fine Mode AOD vs total AOD"} {"chunk_id": "satellite_satellite-aerosol-properties_consistency_q01__75e3d16c6eb1", "report_id": "satellite_satellite-aerosol-properties_consistency_q01", "dataset_id": "satellite-aerosol-properties", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q01", "aspect_base": "consistency", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Assessment of Seasonal Variability and Completeness of Aerosol Optical Depth from SLSTR Satellite Data (2017-2023) > Quality assessment statement", "title": "Assessment of Seasonal Variability and Completeness of Aerosol Optical Depth from SLSTR Satellite Data (2017-2023)", "chunk_index": 5, "token_count": 1123, "text_raw": "number of valid observations per 1,000,000 km² over time. The maximum normalized observation density reached approximately 85 counts / 1,000,000 km².\n\n#### Scatter plot of Fine Mode AOD vs total AOD\n\nThis figure shows a series of scatter plots comparing Fine Mode Aerosol Optical Depth (FM_AOD550) with Total AOD (AOD550) for each subregion, based on monthly averages. Each subplot includes a red regression line with its corresponding equation and R² value, indicating the strength of the linear relationship.\nAll regions display a high correlation, suggesting that fine-mode particles—such as smoke, continental aerosols, and anthropogenic pollution—contribute significantly to total aerosol loading.\nRegions with smaller R² values, such as North Africa and India, indicate a greater contribution from coarse and giant particles to the total AOD.\nThese plots are useful for understanding regional differences in aerosol size and type, the latter known the local aerosol climatology.\n\n*Figure 6.* Scatter plots of monthly mean Fine Mode Aerosol Optical Depth (FM_AOD550) versus Total Aerosol Optical Depth (AOD550) for selected global subregions. Each panel includes a linear regression line (in red) with the equation and coefficient of determination (R²) indicated. The black dashed line represents the 1:1 line, highlighting perfect agreement between the two variables.\n\n(section-4)=\n### Take-Home Messages\n\nThe current aerosol climatology shows substantial deviations from earlier studies [1]. Over North Africa and the Mediterranean Basin, a significant increase in mineral dust intrusions was observed during 2020–2022, as reported by Cuevas-Agulló et al. (2024). This increase is associated with enhanced activity of high-pressure systems over the Euro-Atlantic sector, which favour the obstruction of the westerlies and the occurrence of cut-off lows at subtropical latitudes.\nDuring the same period, the coarse-mode AOD fraction increased over the Mediterranean, further confirming the influence of mineral dust transport due to the atmospheric circulation in the Mediterranean Basin.\nThe weaker correlation between fine-mode and total AOD reflects the greater contribution of coarse-mode particles, particularly dust, rather than inconsistencies in retrieval.\nIn the Northern Hemisphere, seasonal variability remains generally consistent across the selected regions, dominated by continental aerosols and anthropogenic pollution.\nStrong seasonal peaks are also observed in:\n- Southern Africa, where higher AOD values in the Southern hemisphere are consistent with the occurrence of forest fires during the dry austral season (Aug-Sep-Oct) [2, 3].\n- Southeast Asia, where AOD is influenced by local fires and aerosol transport from northern Australia [5].\nIn Siberia, several months of data are missing from the ensemble retrieval, making it difficult to assess seasonal variability.\n\n## ℹ️ If you want to know more\n\n### Key Resources\n\n• Aerosol properties gridded data from 1995 to present derived from satellite observations:\n\nhttps://cds.climate.copernicus.eu/datasets/satellite-aerosol-properties?tab=overview\n\nCode libraries used:\n\n• C3S EQC custom function, c3s_eqc_automatic_quality_control, prepared by B-Open\n\n### References\n\n[1] Remer, L. A., et al. (2008). Global aerosol climatology from the MODIS satellite sensors. Journal of Geophysical Research: Atmospheres, 113(D14S07). https://doi.org/10.1029/2007JD009661\n\n[2] Nyasulu, M., Haque, M.M., Musonda, B. et al. The long-term spatial and temporal distribution of aerosol optical depth and its associated atmospheric circulation over Southeast Africa. Environ Sci Pollut Res 29, 30073–30089 (2022). https://doi.org/10.1007/s11356-021-18446-7\n\n[3] Flamant, C., Chaboureau, J.-P., Gaetani, M., Schepanski, K., and Formenti, P.: The radiative impact of biomass burning aerosols on dust emissions over Namibia and the long-range transport of smoke observed during the Aerosols, Radiation and Clouds in southern Africa (AEROCLO-sA) campaign, Atmos. Chem. Phys., 24, 4265–4288, https://doi.org/10.5194/acp-24-4265-2024, 2024.\n\n[4] Cuevas-Agulló, E. et al. (2024). Sharp increase in Saharan dust intrusions over the western Euro-Mediterranean... https://doi.org/10.5194/acp-24-4083-2024\n\n[5] Jin, S. et al. (2023). A comprehensive reappraisal of long-term aerosol characteristics... https://doi.org/10.5194/acp-23-8187-2023", "text_with_prefix": "EQC Quality Assessment: \"Assessment of Seasonal Variability and Completeness of Aerosol Optical Depth from SLSTR Satellite Data (2017-2023)\"\nDataset: satellite-aerosol-properties [CDS]\nAspect: consistency_q01 | Category: Satellite_ECVs\nSection: Assessment of Seasonal Variability and Completeness of Aerosol Optical Depth from SLSTR Satellite Data (2017-2023) > Quality assessment statement\n---\nnumber of valid observations per 1,000,000 km² over time. The maximum normalized observation density reached approximately 85 counts / 1,000,000 km².\n\n#### Scatter plot of Fine Mode AOD vs total AOD\n\nThis figure shows a series of scatter plots comparing Fine Mode Aerosol Optical Depth (FM_AOD550) with Total AOD (AOD550) for each subregion, based on monthly averages. Each subplot includes a red regression line with its corresponding equation and R² value, indicating the strength of the linear relationship.\nAll regions display a high correlation, suggesting that fine-mode particles—such as smoke, continental aerosols, and anthropogenic pollution—contribute significantly to total aerosol loading.\nRegions with smaller R² values, such as North Africa and India, indicate a greater contribution from coarse and giant particles to the total AOD.\nThese plots are useful for understanding regional differences in aerosol size and type, the latter known the local aerosol climatology.\n\n*Figure 6.* Scatter plots of monthly mean Fine Mode Aerosol Optical Depth (FM_AOD550) versus Total Aerosol Optical Depth (AOD550) for selected global subregions. Each panel includes a linear regression line (in red) with the equation and coefficient of determination (R²) indicated. The black dashed line represents the 1:1 line, highlighting perfect agreement between the two variables.\n\n(section-4)=\n### Take-Home Messages\n\nThe current aerosol climatology shows substantial deviations from earlier studies [1]. Over North Africa and the Mediterranean Basin, a significant increase in mineral dust intrusions was observed during 2020–2022, as reported by Cuevas-Agulló et al. (2024). This increase is associated with enhanced activity of high-pressure systems over the Euro-Atlantic sector, which favour the obstruction of the westerlies and the occurrence of cut-off lows at subtropical latitudes.\nDuring the same period, the coarse-mode AOD fraction increased over the Mediterranean, further confirming the influence of mineral dust transport due to the atmospheric circulation in the Mediterranean Basin.\nThe weaker correlation between fine-mode and total AOD reflects the greater contribution of coarse-mode particles, particularly dust, rather than inconsistencies in retrieval.\nIn the Northern Hemisphere, seasonal variability remains generally consistent across the selected regions, dominated by continental aerosols and anthropogenic pollution.\nStrong seasonal peaks are also observed in:\n- Southern Africa, where higher AOD values in the Southern hemisphere are consistent with the occurrence of forest fires during the dry austral season (Aug-Sep-Oct) [2, 3].\n- Southeast Asia, where AOD is influenced by local fires and aerosol transport from northern Australia [5].\nIn Siberia, several months of data are missing from the ensemble retrieval, making it difficult to assess seasonal variability.\n\n## ℹ️ If you want to know more\n\n### Key Resources\n\n• Aerosol properties gridded data from 1995 to present derived from satellite observations:\n\nhttps://cds.climate.copernicus.eu/datasets/satellite-aerosol-properties?tab=overview\n\nCode libraries used:\n\n• C3S EQC custom function, c3s_eqc_automatic_quality_control, prepared by B-Open\n\n### References\n\n[1] Remer, L. A., et al. (2008). Global aerosol climatology from the MODIS satellite sensors. Journal of Geophysical Research: Atmospheres, 113(D14S07). https://doi.org/10.1029/2007JD009661\n\n[2] Nyasulu, M., Haque, M.M., Musonda, B. et al. The long-term spatial and temporal distribution of aerosol optical depth and its associated atmospheric circulation over Southeast Africa. Environ Sci Pollut Res 29, 30073–30089 (2022). https://doi.org/10.1007/s11356-021-18446-7\n\n[3] Flamant, C., Chaboureau, J.-P., Gaetani, M., Schepanski, K., and Formenti, P.: The radiative impact of biomass burning aerosols on dust emissions over Namibia and the long-range transport of smoke observed during the Aerosols, Radiation and Clouds in southern Africa (AEROCLO-sA) campaign, Atmos. Chem. Phys., 24, 4265–4288, https://doi.org/10.5194/acp-24-4265-2024, 2024.\n\n[4] Cuevas-Agulló, E. et al. (2024). Sharp increase in Saharan dust intrusions over the western Euro-Mediterranean... https://doi.org/10.5194/acp-24-4083-2024\n\n[5] Jin, S. et al. (2023). A comprehensive reappraisal of long-term aerosol characteristics... https://doi.org/10.5194/acp-23-8187-2023"} {"chunk_id": "satellite_satellite-aerosol-properties_consistency_q01__74f231f9e989", "report_id": "satellite_satellite-aerosol-properties_consistency_q01", "dataset_id": "satellite-aerosol-properties", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q01", "aspect_base": "consistency", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Assessment of Seasonal Variability and Completeness of Aerosol Optical Depth from SLSTR Satellite Data (2017-2023) > Quality assessment statement", "title": "Assessment of Seasonal Variability and Completeness of Aerosol Optical Depth from SLSTR Satellite Data (2017-2023)", "chunk_index": 6, "token_count": 378, "text_raw": "-2024\n\n[5] Jin, S. et al. (2023). A comprehensive reappraisal of long-term aerosol characteristics... https://doi.org/10.5194/acp-23-8187-2023\n\n[6] Garrigues, S., Remy, S., Chimot, J., Ades, M., Inness, A., Flemming, J., Kipling, Z., Laszlo, I., Benedetti, A., Ribas, R. and Jafariserajehlou, S., 2022. Monitoring multiple satellite aerosol optical depth (AOD) products within the Copernicus Atmosphere Monitoring Service (CAMS) data assimilation system. Atmospheric Chemistry and Physics Discussions, 2022, pp.1-80.\n\n[7] Copernicus Climate Change Service (2021). Product User Guide and Specifications for Satellite Aerosol Properties (C3S2_312a_Lot2_DLR_2021SC1). Copernicus Climate Change Service. https://dast.copernicus-climate.eu/documents/satellite-aerosol-properties/C3S2_312a_Lot2_FDDP-AER/C3S2_312a_Lot2_D-WP2-FDDP-AER_202311_PUGS_AER_v2.0_final2.pdf", "text_with_prefix": "EQC Quality Assessment: \"Assessment of Seasonal Variability and Completeness of Aerosol Optical Depth from SLSTR Satellite Data (2017-2023)\"\nDataset: satellite-aerosol-properties [CDS]\nAspect: consistency_q01 | Category: Satellite_ECVs\nSection: Assessment of Seasonal Variability and Completeness of Aerosol Optical Depth from SLSTR Satellite Data (2017-2023) > Quality assessment statement\n---\n-2024\n\n[5] Jin, S. et al. (2023). A comprehensive reappraisal of long-term aerosol characteristics... https://doi.org/10.5194/acp-23-8187-2023\n\n[6] Garrigues, S., Remy, S., Chimot, J., Ades, M., Inness, A., Flemming, J., Kipling, Z., Laszlo, I., Benedetti, A., Ribas, R. and Jafariserajehlou, S., 2022. Monitoring multiple satellite aerosol optical depth (AOD) products within the Copernicus Atmosphere Monitoring Service (CAMS) data assimilation system. Atmospheric Chemistry and Physics Discussions, 2022, pp.1-80.\n\n[7] Copernicus Climate Change Service (2021). Product User Guide and Specifications for Satellite Aerosol Properties (C3S2_312a_Lot2_DLR_2021SC1). Copernicus Climate Change Service. https://dast.copernicus-climate.eu/documents/satellite-aerosol-properties/C3S2_312a_Lot2_FDDP-AER/C3S2_312a_Lot2_D-WP2-FDDP-AER_202311_PUGS_AER_v2.0_final2.pdf"} {"chunk_id": "satellite_satellite-carbon-dioxide_consistency_q04__1bc3fc304929", "report_id": "satellite_satellite-carbon-dioxide_consistency_q04", "dataset_id": "satellite-carbon-dioxide", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q04", "aspect_base": "consistency", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Consistency of carbon dioxide satellite observations for the evaluation of Earth system models. > Quality assessment question", "title": "Consistency of carbon dioxide satellite observations for the evaluation of Earth system models.", "chunk_index": 0, "token_count": 683, "text_raw": "* **Is the temporal consistency of the Level-3 Obs4MIPs merged dataset suitable for calculating changes in the annual growth rates of XCO₂ and the amplitude of the seasonal cycles?**\n\nThe most important anthropogenic greenhouse gas is carbon dioxide (CO$_2$), with CO$_2$ emissions contributing 66% of the total radiative forcing by long-lived greenhouse gases [[1]](https://library.wmo.int/idurl/4/69057). It is therefore important to monitor the long-term changes in atmospheric CO$_2$ concentrations, to understand the sources and sinks of the atmospheric carbon content and to provide reliable projections of future CO$_2$ concentrations under various scenarios.\n\nEarth System Models (ESMs) are computer simulations that couple the physical climate (i.e. the atmosphere, the oceans and sea ice) with other components of the Earth, such as the land surface, vegetation, biogeochemistry (i.e. the carbon and nitrogen cycles), the cryosphere and, in some cases, human systems, in order to simulate how the entire Earth behaves now and in the future [[2]](https://doi.org/10.5194/gmd-9-1937-2016). They are widely used, among other aims, to provide climate projections under different greenhouse gas and land use scenarios, to determine the contribution of human activities to observed changes or specific extreme events, to investigate biogeochemical cycles, and to inform mitigation pathways and climate risk analysis for policymakers.\n\nIn this assessment, we show how the column-averaged mixing ratios of CO$_2$ (XCO$_2$) provided by the Obs4MIPs Level 3 data product of [[3]](https://cds.climate.copernicus.eu/datasets/satellite-carbon-dioxide?tab=overview) can be used to evaluate the annual growth rates and seasonal cycles of CO$_2$ provided by ESMs. Previous versions of this dataset have been already used to calculate global annual growth rates of CO$_2$ ([[4]](https://doi.org/10.5194/acp-18-17355-2018), [[5]](https://doi.org/10.5194/amt-13-789-2020)). These studies reported good agreement between growth rates derived from satellite data and those derived from surface observations by the National Oceanic and Atmospheric Administration (NOAA).\n\nThe Obs4MIPs Level 3 data product results from integrating XCO₂ data from different combinations of satellite sensors and algorithms over time and space. Based on the results of references [[6]](https://doi.org/10.5194/bg-17-6115-2020) and independent analyses, we investigate the impact of changes in the sampling characteristics of the Obs4MIPs dataset on the ability to use it to evaluate the capacity of ESMs to reproduce CO₂ seasonal cycle amplitudes and related trends/inter-annual variability, across its entire temporal coverage.", "text_with_prefix": "EQC Quality Assessment: \"Consistency of carbon dioxide satellite observations for the evaluation of Earth system models.\"\nDataset: satellite-carbon-dioxide [CDS]\nAspect: consistency_q04 | Category: Satellite_ECVs\nSection: Consistency of carbon dioxide satellite observations for the evaluation of Earth system models. > Quality assessment question\n---\n* **Is the temporal consistency of the Level-3 Obs4MIPs merged dataset suitable for calculating changes in the annual growth rates of XCO₂ and the amplitude of the seasonal cycles?**\n\nThe most important anthropogenic greenhouse gas is carbon dioxide (CO$_2$), with CO$_2$ emissions contributing 66% of the total radiative forcing by long-lived greenhouse gases [[1]](https://library.wmo.int/idurl/4/69057). It is therefore important to monitor the long-term changes in atmospheric CO$_2$ concentrations, to understand the sources and sinks of the atmospheric carbon content and to provide reliable projections of future CO$_2$ concentrations under various scenarios.\n\nEarth System Models (ESMs) are computer simulations that couple the physical climate (i.e. the atmosphere, the oceans and sea ice) with other components of the Earth, such as the land surface, vegetation, biogeochemistry (i.e. the carbon and nitrogen cycles), the cryosphere and, in some cases, human systems, in order to simulate how the entire Earth behaves now and in the future [[2]](https://doi.org/10.5194/gmd-9-1937-2016). They are widely used, among other aims, to provide climate projections under different greenhouse gas and land use scenarios, to determine the contribution of human activities to observed changes or specific extreme events, to investigate biogeochemical cycles, and to inform mitigation pathways and climate risk analysis for policymakers.\n\nIn this assessment, we show how the column-averaged mixing ratios of CO$_2$ (XCO$_2$) provided by the Obs4MIPs Level 3 data product of [[3]](https://cds.climate.copernicus.eu/datasets/satellite-carbon-dioxide?tab=overview) can be used to evaluate the annual growth rates and seasonal cycles of CO$_2$ provided by ESMs. Previous versions of this dataset have been already used to calculate global annual growth rates of CO$_2$ ([[4]](https://doi.org/10.5194/acp-18-17355-2018), [[5]](https://doi.org/10.5194/amt-13-789-2020)). These studies reported good agreement between growth rates derived from satellite data and those derived from surface observations by the National Oceanic and Atmospheric Administration (NOAA).\n\nThe Obs4MIPs Level 3 data product results from integrating XCO₂ data from different combinations of satellite sensors and algorithms over time and space. Based on the results of references [[6]](https://doi.org/10.5194/bg-17-6115-2020) and independent analyses, we investigate the impact of changes in the sampling characteristics of the Obs4MIPs dataset on the ability to use it to evaluate the capacity of ESMs to reproduce CO₂ seasonal cycle amplitudes and related trends/inter-annual variability, across its entire temporal coverage."} {"chunk_id": "satellite_satellite-carbon-dioxide_consistency_q04__0aebe6af1e86", "report_id": "satellite_satellite-carbon-dioxide_consistency_q04", "dataset_id": "satellite-carbon-dioxide", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q04", "aspect_base": "consistency", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Consistency of carbon dioxide satellite observations for the evaluation of Earth system models. > Quality assessment statement", "title": "Consistency of carbon dioxide satellite observations for the evaluation of Earth system models.", "chunk_index": 1, "token_count": 283, "text_raw": "These are the key outcomes of this assessment\n\n* The Obs4MIPs Lev3 merged dataset is suitable for assessing the capacity of Earth System Models to reproduce CO₂ growth rates and seasonal cycle variability, however the dataset features related to the spatial and temporal data coverage must be carefully considered (see points below).\n* For annual aggregation, high-latitude regions (i.e.. latitudes higher than 60°N and 60°S) should not be considered due to non-uniform data coverage across the different seasons.\n* The introduction of new satellite data (e.g., GOSAT in 2009 and OCO-2 in 2015) affects the data coverage at global scale with changing coverage of high latitude regions and oceans.\n* The months characterised by changes in the contributing satellite data can show inconsistent CO₂ growth rates; however, this feature is averaged out when computing the annual growth.\n* Changes in data coverage over the period covered by the Obs4MIPs Level 3 dataset affect the calculation of the seasonal cycle amplitude.\n```", "text_with_prefix": "EQC Quality Assessment: \"Consistency of carbon dioxide satellite observations for the evaluation of Earth system models.\"\nDataset: satellite-carbon-dioxide [CDS]\nAspect: consistency_q04 | Category: Satellite_ECVs\nSection: Consistency of carbon dioxide satellite observations for the evaluation of Earth system models. > Quality assessment statement\n---\nThese are the key outcomes of this assessment\n\n* The Obs4MIPs Lev3 merged dataset is suitable for assessing the capacity of Earth System Models to reproduce CO₂ growth rates and seasonal cycle variability, however the dataset features related to the spatial and temporal data coverage must be carefully considered (see points below).\n* For annual aggregation, high-latitude regions (i.e.. latitudes higher than 60°N and 60°S) should not be considered due to non-uniform data coverage across the different seasons.\n* The introduction of new satellite data (e.g., GOSAT in 2009 and OCO-2 in 2015) affects the data coverage at global scale with changing coverage of high latitude regions and oceans.\n* The months characterised by changes in the contributing satellite data can show inconsistent CO₂ growth rates; however, this feature is averaged out when computing the annual growth.\n* Changes in data coverage over the period covered by the Obs4MIPs Level 3 dataset affect the calculation of the seasonal cycle amplitude.\n```"} {"chunk_id": "satellite_satellite-carbon-dioxide_consistency_q04__99de4b6231dc", "report_id": "satellite_satellite-carbon-dioxide_consistency_q04", "dataset_id": "satellite-carbon-dioxide", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q04", "aspect_base": "consistency", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Consistency of carbon dioxide satellite observations for the evaluation of Earth system models. > Methodology", "title": "Consistency of carbon dioxide satellite observations for the evaluation of Earth system models.", "chunk_index": 2, "token_count": 817, "text_raw": "To assess if the temporal consistency of the Level-3 Obs4MIPs merged dataset from the Climate Data Store “Carbon dioxide data from 2002 to present derived from satellite observations” [[3]](https://cds.climate.copernicus.eu/datasets/satellite-carbon-dioxide?tab=overview) is suitable for calculating changes in the annual growth rates of XCO$_2$ and the amplitude of the seasonal cycles, we inspected a specific case study available from scientific literature [[6]](https://doi.org/10.5194/bg-17-6115-2020) and we performed independent analyses. In particular, we analysed the changes of data coverage as a function of months and years, we calculated the monthly resolved CO$_2$ annual growth rates and the annual seasonal cycles. To do that we used a methodology consistent with [[6]](https://doi.org/10.5194/bg-17-6115-2020). However, we perform our analyses over the period 2003 - 2023 by using the most recent version (4.6) of the Level-3 Obs4MIPs merged dataset [[7]](https://confluence.ecmwf.int/pages/viewpage.action?pageId=567160605). This dataset is obtained by gridding the merged level-2 product generated with the ensemble median algorithm (XCO2_EMMA). EMMA combines several different XCO2 level-2 satellite data products: SCIAMACHY/Envisat (2003–2012), TANSO-FTS/GOSAT (2009–2022), OCO-2 (2014-2022), TANSO-FTS-2/GOSAT2 (2019-2023).\n\n[[4]](https://doi.org/10.5194/acp-18-17355-2018) used a previous version (O4Mv3) of the Level-3 Obs4MIPs merged data product available by the Climate Data Store to calulate the global atmospheric CO$_2$ annual mean growth rates during 2003–2016 and to analyse the relative contributions of anthropogenic emissions and El Niño Southern Oscillation (ENSO) to the observed growth rates. This study reported a good agreement (mean difference ± standard deviation: 0.0 ± 0.3 ppm/year; R: 0.82)\n\nMore specifically, this notebook aims to:\n* Generate global maps of monthly XCO$_2$ data coverage, to show that the spatial coverage of data is not uniform across the different regions of the world for the different seasons. \n* Generate global maps of XCO$_2$ data coverage as a function of the different temporal periods covered by different satellite/sensors, i.e. SCIAMACHY/Envisat (2003–2012), TANSO-FTS/GOSAT (2009–2023), OCO-2 (2014-2023), TANSO-FTS-2/GOSAT2 (2019-2023).\n* To calculate the global annual growth rates of XCO$_2$ and the seasonal cycle amplitudes for the entire measurement period 2003-2022.\n\nThe analysis and results are organized in the following steps, which are detailed in the sections below:\n\n**[](template:section-1)**\n * Import all relevant packages.\n * Select the variable of interest\n * Cache needed functions.\n\n**[](template:section-2)**\n * Define the data request to CDS.\n\n**[](template:section-3)**", "text_with_prefix": "EQC Quality Assessment: \"Consistency of carbon dioxide satellite observations for the evaluation of Earth system models.\"\nDataset: satellite-carbon-dioxide [CDS]\nAspect: consistency_q04 | Category: Satellite_ECVs\nSection: Consistency of carbon dioxide satellite observations for the evaluation of Earth system models. > Methodology\n---\nTo assess if the temporal consistency of the Level-3 Obs4MIPs merged dataset from the Climate Data Store “Carbon dioxide data from 2002 to present derived from satellite observations” [[3]](https://cds.climate.copernicus.eu/datasets/satellite-carbon-dioxide?tab=overview) is suitable for calculating changes in the annual growth rates of XCO$_2$ and the amplitude of the seasonal cycles, we inspected a specific case study available from scientific literature [[6]](https://doi.org/10.5194/bg-17-6115-2020) and we performed independent analyses. In particular, we analysed the changes of data coverage as a function of months and years, we calculated the monthly resolved CO$_2$ annual growth rates and the annual seasonal cycles. To do that we used a methodology consistent with [[6]](https://doi.org/10.5194/bg-17-6115-2020). However, we perform our analyses over the period 2003 - 2023 by using the most recent version (4.6) of the Level-3 Obs4MIPs merged dataset [[7]](https://confluence.ecmwf.int/pages/viewpage.action?pageId=567160605). This dataset is obtained by gridding the merged level-2 product generated with the ensemble median algorithm (XCO2_EMMA). EMMA combines several different XCO2 level-2 satellite data products: SCIAMACHY/Envisat (2003–2012), TANSO-FTS/GOSAT (2009–2022), OCO-2 (2014-2022), TANSO-FTS-2/GOSAT2 (2019-2023).\n\n[[4]](https://doi.org/10.5194/acp-18-17355-2018) used a previous version (O4Mv3) of the Level-3 Obs4MIPs merged data product available by the Climate Data Store to calulate the global atmospheric CO$_2$ annual mean growth rates during 2003–2016 and to analyse the relative contributions of anthropogenic emissions and El Niño Southern Oscillation (ENSO) to the observed growth rates. This study reported a good agreement (mean difference ± standard deviation: 0.0 ± 0.3 ppm/year; R: 0.82)\n\nMore specifically, this notebook aims to:\n* Generate global maps of monthly XCO$_2$ data coverage, to show that the spatial coverage of data is not uniform across the different regions of the world for the different seasons. \n* Generate global maps of XCO$_2$ data coverage as a function of the different temporal periods covered by different satellite/sensors, i.e. SCIAMACHY/Envisat (2003–2012), TANSO-FTS/GOSAT (2009–2023), OCO-2 (2014-2023), TANSO-FTS-2/GOSAT2 (2019-2023).\n* To calculate the global annual growth rates of XCO$_2$ and the seasonal cycle amplitudes for the entire measurement period 2003-2022.\n\nThe analysis and results are organized in the following steps, which are detailed in the sections below:\n\n**[](template:section-1)**\n * Import all relevant packages.\n * Select the variable of interest\n * Cache needed functions.\n\n**[](template:section-2)**\n * Define the data request to CDS.\n\n**[](template:section-3)**"} {"chunk_id": "satellite_satellite-carbon-dioxide_consistency_q04__8ff9e1bdb0f7", "report_id": "satellite_satellite-carbon-dioxide_consistency_q04", "dataset_id": "satellite-carbon-dioxide", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q04", "aspect_base": "consistency", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Consistency of carbon dioxide satellite observations for the evaluation of Earth system models. > Methodology", "title": "Consistency of carbon dioxide satellite observations for the evaluation of Earth system models.", "chunk_index": 3, "token_count": 489, "text_raw": "Select the variable of interest\n * Cache needed functions.\n\n**[](template:section-2)**\n * Define the data request to CDS.\n\n**[](template:section-3)**\n\nThis section presents several results for the XCO$_2$ products, i.e.:\n * The monthtly mean data coverage of XCO$_2$ over the period 2002-2022.\n * The mean data coverage of XCO$_2$ over specific time periods characterised by the use of observations from different satellites.\n * The calculation of the XCO$_2$ growth rates and annual seasonal cycles by using an approach similar to that adopted by [[6]](https://doi.org/10.5194/bg-17-6115-2020). Specifically, we computed monthly resolved annual growth rates by subtracting the XCO$_2$ value 6 months in the future from the one 6 months in the past. Then these monthly resolved growth rates are averaged to a yearly GR for a calendar year. We calculated the annual seasonal cycled of global XCO$_2$ by detrending the monthly time series with the cumulative sum of monthly growth rates, using the annual mean growth rates as substitution for missing values where necessary. Finally, we normalised the time series of detrended monthly values by subtracting the average seasonal cycle over the whole period 2003 - 2023. Spatial averages are calculated by taking the arithmetic averages over all grid cells weighted by their area.\n * Similarly to [[6]](https://doi.org/10.5194/bg-17-6115-2020), we calculated of the annual seasonal cycle amplitudes (SCAs) as the peak-to-trough amplitude in a calendar year of the detrended time series. The SCA was calculated for a shorter time period (2004 – 2022) with respect to the full available dataset (2003 - 2023), to avoid artefacts related to the fact that monthly growth rates were only available for six months in 2003 and 2023.", "text_with_prefix": "EQC Quality Assessment: \"Consistency of carbon dioxide satellite observations for the evaluation of Earth system models.\"\nDataset: satellite-carbon-dioxide [CDS]\nAspect: consistency_q04 | Category: Satellite_ECVs\nSection: Consistency of carbon dioxide satellite observations for the evaluation of Earth system models. > Methodology\n---\nSelect the variable of interest\n * Cache needed functions.\n\n**[](template:section-2)**\n * Define the data request to CDS.\n\n**[](template:section-3)**\n\nThis section presents several results for the XCO$_2$ products, i.e.:\n * The monthtly mean data coverage of XCO$_2$ over the period 2002-2022.\n * The mean data coverage of XCO$_2$ over specific time periods characterised by the use of observations from different satellites.\n * The calculation of the XCO$_2$ growth rates and annual seasonal cycles by using an approach similar to that adopted by [[6]](https://doi.org/10.5194/bg-17-6115-2020). Specifically, we computed monthly resolved annual growth rates by subtracting the XCO$_2$ value 6 months in the future from the one 6 months in the past. Then these monthly resolved growth rates are averaged to a yearly GR for a calendar year. We calculated the annual seasonal cycled of global XCO$_2$ by detrending the monthly time series with the cumulative sum of monthly growth rates, using the annual mean growth rates as substitution for missing values where necessary. Finally, we normalised the time series of detrended monthly values by subtracting the average seasonal cycle over the whole period 2003 - 2023. Spatial averages are calculated by taking the arithmetic averages over all grid cells weighted by their area.\n * Similarly to [[6]](https://doi.org/10.5194/bg-17-6115-2020), we calculated of the annual seasonal cycle amplitudes (SCAs) as the peak-to-trough amplitude in a calendar year of the detrended time series. The SCA was calculated for a shorter time period (2004 – 2022) with respect to the full available dataset (2003 - 2023), to avoid artefacts related to the fact that monthly growth rates were only available for six months in 2003 and 2023."} {"chunk_id": "satellite_satellite-carbon-dioxide_consistency_q04__37c8156de66f", "report_id": "satellite_satellite-carbon-dioxide_consistency_q04", "dataset_id": "satellite-carbon-dioxide", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q04", "aspect_base": "consistency", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Consistency of carbon dioxide satellite observations for the evaluation of Earth system models. > Analysis and results > 1. Choose the data to use and set-up the code > Import all relevant packages", "title": "Consistency of carbon dioxide satellite observations for the evaluation of Earth system models.", "chunk_index": 4, "token_count": 136, "text_raw": "In this section, we import all the relevant packages needed for running the notebook and we define the style for plot appearance (\"seaborn-v0_8-notebook\") as well as the style of hatches used in the maps.", "text_with_prefix": "EQC Quality Assessment: \"Consistency of carbon dioxide satellite observations for the evaluation of Earth system models.\"\nDataset: satellite-carbon-dioxide [CDS]\nAspect: consistency_q04 | Category: Satellite_ECVs\nSection: Consistency of carbon dioxide satellite observations for the evaluation of Earth system models. > Analysis and results > 1. Choose the data to use and set-up the code > Import all relevant packages\n---\nIn this section, we import all the relevant packages needed for running the notebook and we define the style for plot appearance (\"seaborn-v0_8-notebook\") as well as the style of hatches used in the maps."} {"chunk_id": "satellite_satellite-carbon-dioxide_consistency_q04__e64d616c059a", "report_id": "satellite_satellite-carbon-dioxide_consistency_q04", "dataset_id": "satellite-carbon-dioxide", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q04", "aspect_base": "consistency", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Consistency of carbon dioxide satellite observations for the evaluation of Earth system models. > Analysis and results > 1. Choose the data to use and set-up the code > Define the variable of interest", "title": "Consistency of carbon dioxide satellite observations for the evaluation of Earth system models.", "chunk_index": 5, "token_count": 135, "text_raw": "In this section, we define the parameters to be ingested by the code (that can be customized by the user), i.e.: \n* the variable of interest.\n\nChoose variable (xch4 or xco2)", "text_with_prefix": "EQC Quality Assessment: \"Consistency of carbon dioxide satellite observations for the evaluation of Earth system models.\"\nDataset: satellite-carbon-dioxide [CDS]\nAspect: consistency_q04 | Category: Satellite_ECVs\nSection: Consistency of carbon dioxide satellite observations for the evaluation of Earth system models. > Analysis and results > 1. Choose the data to use and set-up the code > Define the variable of interest\n---\nIn this section, we define the parameters to be ingested by the code (that can be customized by the user), i.e.: \n* the variable of interest.\n\nChoose variable (xch4 or xco2)"} {"chunk_id": "satellite_satellite-carbon-dioxide_consistency_q04__17308101dfda", "report_id": "satellite_satellite-carbon-dioxide_consistency_q04", "dataset_id": "satellite-carbon-dioxide", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q04", "aspect_base": "consistency", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Consistency of carbon dioxide satellite observations for the evaluation of Earth system models. > Analysis and results > 1. Choose the data to use and set-up the code > Cache needed functions", "title": "Consistency of carbon dioxide satellite observations for the evaluation of Earth system models.", "chunk_index": 6, "token_count": 336, "text_raw": "In this section, we cached a list of functions used in the analyses.\n\n* The `convert_units` function rescales XCO$_2$ mole fraction to parts per million (ppm).\n\n* The `compute_coverage` function calculates the spatial fractional coverage of the available data and creates associated maps.\n\n* The `def area_weighted_series` function calculates the spatial averages over all grid cells in a region. It uses spatial weighting to account for the latitudinal dependence of the grid size in the lon/lat grids.\n\n* The `compute_GR_and_seasonal` funtion calculates the monthly resolved annual growth rates of XCO$_2$ and the annual seasonal cycles.\n\nnumerator: sum over lat,lon of x * weight (ignore NaNs in x)\navoid division by zero\nconvert to xarray DataArray with time coords\ncompute GR: GR(m) = x(t+6) - x(t-6)\nannual mean GR (used to substitute missing monthly GR)\nfill monthly GR missing with annual mean where possible\nmonthly increment (ppm/month) from yearly GR\ncumulative trend (ppm)\nseasonal cycle: xco2 - cumulative trend ; center to zero mean\n\n(template:section-2)=", "text_with_prefix": "EQC Quality Assessment: \"Consistency of carbon dioxide satellite observations for the evaluation of Earth system models.\"\nDataset: satellite-carbon-dioxide [CDS]\nAspect: consistency_q04 | Category: Satellite_ECVs\nSection: Consistency of carbon dioxide satellite observations for the evaluation of Earth system models. > Analysis and results > 1. Choose the data to use and set-up the code > Cache needed functions\n---\nIn this section, we cached a list of functions used in the analyses.\n\n* The `convert_units` function rescales XCO$_2$ mole fraction to parts per million (ppm).\n\n* The `compute_coverage` function calculates the spatial fractional coverage of the available data and creates associated maps.\n\n* The `def area_weighted_series` function calculates the spatial averages over all grid cells in a region. It uses spatial weighting to account for the latitudinal dependence of the grid size in the lon/lat grids.\n\n* The `compute_GR_and_seasonal` funtion calculates the monthly resolved annual growth rates of XCO$_2$ and the annual seasonal cycles.\n\nnumerator: sum over lat,lon of x * weight (ignore NaNs in x)\navoid division by zero\nconvert to xarray DataArray with time coords\ncompute GR: GR(m) = x(t+6) - x(t-6)\nannual mean GR (used to substitute missing monthly GR)\nfill monthly GR missing with annual mean where possible\nmonthly increment (ppm/month) from yearly GR\ncumulative trend (ppm)\nseasonal cycle: xco2 - cumulative trend ; center to zero mean\n\n(template:section-2)="} {"chunk_id": "satellite_satellite-carbon-dioxide_consistency_q04__97a92e777296", "report_id": "satellite_satellite-carbon-dioxide_consistency_q04", "dataset_id": "satellite-carbon-dioxide", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q04", "aspect_base": "consistency", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Consistency of carbon dioxide satellite observations for the evaluation of Earth system models. > Analysis and results > 2. Retrieve data > 2.1 Obs4MIPs", "title": "Consistency of carbon dioxide satellite observations for the evaluation of Earth system models.", "chunk_index": 7, "token_count": 135, "text_raw": "In this section, we define the data request to CDS (data product Obs4MIPs, Level 3, version 4.5, XCO$_2$) and download the full dataset.\n\n(template:section-3)=", "text_with_prefix": "EQC Quality Assessment: \"Consistency of carbon dioxide satellite observations for the evaluation of Earth system models.\"\nDataset: satellite-carbon-dioxide [CDS]\nAspect: consistency_q04 | Category: Satellite_ECVs\nSection: Consistency of carbon dioxide satellite observations for the evaluation of Earth system models. > Analysis and results > 2. Retrieve data > 2.1 Obs4MIPs\n---\nIn this section, we define the data request to CDS (data product Obs4MIPs, Level 3, version 4.5, XCO$_2$) and download the full dataset.\n\n(template:section-3)="} {"chunk_id": "satellite_satellite-carbon-dioxide_consistency_q04__f6650aa68ca4", "report_id": "satellite_satellite-carbon-dioxide_consistency_q04", "dataset_id": "satellite-carbon-dioxide", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q04", "aspect_base": "consistency", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Consistency of carbon dioxide satellite observations for the evaluation of Earth system models. > Analysis and results > 3. Data analysis > Global maps of monthly XCO$_2$ data coverage", "title": "Consistency of carbon dioxide satellite observations for the evaluation of Earth system models.", "chunk_index": 8, "token_count": 332, "text_raw": "In this section, we calculate and plot the mean monthly fractional coverage of data over the period 2003 - 2023. In agreement with the results by [[6]](https://doi.org/10.5194/bg-17-6115-2020), the number of available observations depends significantly on the location with low data coverage values over locations typically affected by high cloud coverage (like tropics) and low Sun elevation (like high latitude in winter months). Coverage over ocean is sparse as ocean retrievals are included from GOSAT, OCO-2 and GOSAT2 Sun-glint mode observations. [[6]](https://doi.org/10.5194/bg-17-6115-2020) demonstrated that taking into account this sampling coverage feature is essential for a proper comparison with outputs from ESMs.\n\n```text\n100%|██████████| 1/1 [00:00<00:00, 3.36it/s]\n```\n\n*The figure shows the monthly fractional XCO$_2$ data coverage from 2003 to 2023 derived from the XCO2_OBS4MIPS dataset (version 4.6).*", "text_with_prefix": "EQC Quality Assessment: \"Consistency of carbon dioxide satellite observations for the evaluation of Earth system models.\"\nDataset: satellite-carbon-dioxide [CDS]\nAspect: consistency_q04 | Category: Satellite_ECVs\nSection: Consistency of carbon dioxide satellite observations for the evaluation of Earth system models. > Analysis and results > 3. Data analysis > Global maps of monthly XCO$_2$ data coverage\n---\nIn this section, we calculate and plot the mean monthly fractional coverage of data over the period 2003 - 2023. In agreement with the results by [[6]](https://doi.org/10.5194/bg-17-6115-2020), the number of available observations depends significantly on the location with low data coverage values over locations typically affected by high cloud coverage (like tropics) and low Sun elevation (like high latitude in winter months). Coverage over ocean is sparse as ocean retrievals are included from GOSAT, OCO-2 and GOSAT2 Sun-glint mode observations. [[6]](https://doi.org/10.5194/bg-17-6115-2020) demonstrated that taking into account this sampling coverage feature is essential for a proper comparison with outputs from ESMs.\n\n```text\n100%|██████████| 1/1 [00:00<00:00, 3.36it/s]\n```\n\n*The figure shows the monthly fractional XCO$_2$ data coverage from 2003 to 2023 derived from the XCO2_OBS4MIPS dataset (version 4.6).*"} {"chunk_id": "satellite_satellite-carbon-dioxide_consistency_q04__0fea1c988ad4", "report_id": "satellite_satellite-carbon-dioxide_consistency_q04", "dataset_id": "satellite-carbon-dioxide", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q04", "aspect_base": "consistency", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Consistency of carbon dioxide satellite observations for the evaluation of Earth system models. > Analysis and results > 3. Data analysis > Global maps of XCO$_2$ data coverage for different periods", "title": "Consistency of carbon dioxide satellite observations for the evaluation of Earth system models.", "chunk_index": 9, "token_count": 555, "text_raw": "In this section, to assess the temporal consistency of spatial data coverage over the whole period covered by the Level-3 Obs4MIPs merged dataset, we calculate and plot the mean fractional coverage of CO$_2$ data for time periods characterised by the contribution of different satellite observations: i.e., 2003 - 2008 (only SCIAMACHY observations), 2009 - 2011 (combination of SCIAMACHY and GOSAT), 2012 - 2014 (GOSAT), 2015 - 2018 (combination of GOSAT and OCO-2), 2019 - 2023 (combination of GOSAT, OCO-2 and GOSAT2). A comprehensive analysis of the number of soundings contained within the EMMA database on a monthly basis, along with the comparative contribution of each individual Level 2 algorithm to the EMMA database, can be found in [[8]](https://confluence.ecmwf.int/pages/viewpage.action?pageId=567160577).\n\nBesides not including data over oceans, with respect to subsequent periods in which GOSAT measurements were used (2009 - 2011 and 2012 - 2014), for the period 2003 - 2008 more data are available for regions higher than 50°N as well as over tropical regions in South America and Africa. The introduction of OCO-2 and GOSAT2 observations (2015 – 2018 and 2019 – 2023, respectively) resulted in an increase of the data coverage over Europe and the oceans. A substantial proportion of these regions exhibited data coverage levels exceeding 0.5, with the data extending to higher latitudes.\n\nTo define temporal coverage\nTo categorise data coverage as a function of time periods\nadd custom labels\nAdd hatched pattern where coverage > 0.5\nconvert to 2D array (lat, lon)\nmask values <=0.5\nAdjust layout and move colorbar below all maps\n\n*The figure shows the monthly fractional XCO$_2$ data coverage for different time periods characterised by the use of different satellites/sensors to obtain obervations. The hatched areas denote regions with data coverage higher than 0.5*", "text_with_prefix": "EQC Quality Assessment: \"Consistency of carbon dioxide satellite observations for the evaluation of Earth system models.\"\nDataset: satellite-carbon-dioxide [CDS]\nAspect: consistency_q04 | Category: Satellite_ECVs\nSection: Consistency of carbon dioxide satellite observations for the evaluation of Earth system models. > Analysis and results > 3. Data analysis > Global maps of XCO$_2$ data coverage for different periods\n---\nIn this section, to assess the temporal consistency of spatial data coverage over the whole period covered by the Level-3 Obs4MIPs merged dataset, we calculate and plot the mean fractional coverage of CO$_2$ data for time periods characterised by the contribution of different satellite observations: i.e., 2003 - 2008 (only SCIAMACHY observations), 2009 - 2011 (combination of SCIAMACHY and GOSAT), 2012 - 2014 (GOSAT), 2015 - 2018 (combination of GOSAT and OCO-2), 2019 - 2023 (combination of GOSAT, OCO-2 and GOSAT2). A comprehensive analysis of the number of soundings contained within the EMMA database on a monthly basis, along with the comparative contribution of each individual Level 2 algorithm to the EMMA database, can be found in [[8]](https://confluence.ecmwf.int/pages/viewpage.action?pageId=567160577).\n\nBesides not including data over oceans, with respect to subsequent periods in which GOSAT measurements were used (2009 - 2011 and 2012 - 2014), for the period 2003 - 2008 more data are available for regions higher than 50°N as well as over tropical regions in South America and Africa. The introduction of OCO-2 and GOSAT2 observations (2015 – 2018 and 2019 – 2023, respectively) resulted in an increase of the data coverage over Europe and the oceans. A substantial proportion of these regions exhibited data coverage levels exceeding 0.5, with the data extending to higher latitudes.\n\nTo define temporal coverage\nTo categorise data coverage as a function of time periods\nadd custom labels\nAdd hatched pattern where coverage > 0.5\nconvert to 2D array (lat, lon)\nmask values <=0.5\nAdjust layout and move colorbar below all maps\n\n*The figure shows the monthly fractional XCO$_2$ data coverage for different time periods characterised by the use of different satellites/sensors to obtain obervations. The hatched areas denote regions with data coverage higher than 0.5*"} {"chunk_id": "satellite_satellite-carbon-dioxide_consistency_q04__a6c2bb73ff9e", "report_id": "satellite_satellite-carbon-dioxide_consistency_q04", "dataset_id": "satellite-carbon-dioxide", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q04", "aspect_base": "consistency", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Consistency of carbon dioxide satellite observations for the evaluation of Earth system models. > Analysis and results > 3. Data analysis > Calculation of growth rate and seasonal cycles", "title": "Consistency of carbon dioxide satellite observations for the evaluation of Earth system models.", "chunk_index": 10, "token_count": 863, "text_raw": "In this section, we calculate and report the time series of monthly global CO₂ values, the resolved annual growth rate, and the seasonal cycles.\nOver the period 2003–2014, our average growth rate of 2.02 ppm/year (with a standard deviation of 0.49 ppm/year) is in very good agreement with the values provided in [[4]](https://doi.org/10.5194/acp-18-17355-2018). The calculated seasonal cycles also completely agree with those calculated in [[6]](https://doi.org/10.5194/bg-17-6115-2020) (1.9 ± 0.4 ppm/year), which report a clear change in shape and amplitude when GOSAT observations began contributing to the calculation of global CO₂ in 2009. The low growth rate observed in 2009 is due to the introduction of GOSAT data, which altered the shape of the seasonal cycle. When considering only land-based data (orange lines), the low growth rate in 2009 disappears, and the seasonal cycles appear more consistent throughout the measurement period with an higher amplitude due to the higher carbon fluxes occurring over land regions than over oceans. This clearly shows that any changes to the sampling features that occur during the period covered by the dataset must be taken into account, particularly when evaluating uniform, gap-free datasets such as those provided by ESMs.\nThe high growth rate values observed in 2010, 2012, 2016, 2019 and 2023 are consistent with those reported by the in situ global network coordinated by [[1]](https://library.wmo.int/idurl/4/69057), and are related to El Niño conditions which favour net carbon emissions into the atmosphere on a global scale (see [[9]](https://doi.org/10.5194/acp-25-13053-2025) for a focus on the large growth rate on 2023).\n\n-------------------------------------------------------\nLoad dataset and basic preprocess\n-------------------------------------------------------\nds = download.download_and_transform(*request)\nReplace missing value 1.0e20 with NaN\nConvert xco2 to ppm\n-------------------------------------------------------\nPrepare area weights (cos(lat)) as 2D DataArray (lat, lon)\n-------------------------------------------------------\n-------------------------------------------------------\nland mask: keep only cells with land_fraction > 0.5\n-------------------------------------------------------\nweights masked\n-------------------------------------------------------\nCompute series for both cases: all gridcells and masked (land_fraction>0.5)\n-------------------------------------------------------\narea-weighted global series\ncompute monthly GR and seasonal cycle for both\n-------------------------------------------------------\nQuick diagnostics (optional prints)\n-------------------------------------------------------\nprint(\"Global (all) series length:\", len(global_all_series.dropna()))\nprint(\"Global (masked) series length:\", len(global_masked_series.dropna()))\nprint(\"Fraction positive GR (all):\", (monthly_GR_all.dropna() > 0).mean())\nprint(\"Fraction positive GR (masked):\", (monthly_GR_masked.dropna() > 0).mean())\n-------------------------------------------------------\nPlotting: three vertical plots with both lines\n-------------------------------------------------------\n(1) Global monthly mean XCO2\n(2) Monthly Growth Rate (ppm/yr)\nadd stats box for both series (means and stds)\n(3) Seasonal Cycle (detrended)\ncenter y-axis of seasonal cycle around zero (symmetric limits) using both series combined\n-------------------------\nFormat x-axis to show integer years\n-------------------------\n\n*The figure shows the global time series of monthly XCO$_2$ mean values (upper panel), monthly resolved annual growth rate (middle panel) and detrended seasonal cycles (bottom panel) from 2003 to 2023. Blue lines indicate results for the whole dataset, orange lines indicate results for land-masked data.*", "text_with_prefix": "EQC Quality Assessment: \"Consistency of carbon dioxide satellite observations for the evaluation of Earth system models.\"\nDataset: satellite-carbon-dioxide [CDS]\nAspect: consistency_q04 | Category: Satellite_ECVs\nSection: Consistency of carbon dioxide satellite observations for the evaluation of Earth system models. > Analysis and results > 3. Data analysis > Calculation of growth rate and seasonal cycles\n---\nIn this section, we calculate and report the time series of monthly global CO₂ values, the resolved annual growth rate, and the seasonal cycles.\nOver the period 2003–2014, our average growth rate of 2.02 ppm/year (with a standard deviation of 0.49 ppm/year) is in very good agreement with the values provided in [[4]](https://doi.org/10.5194/acp-18-17355-2018). The calculated seasonal cycles also completely agree with those calculated in [[6]](https://doi.org/10.5194/bg-17-6115-2020) (1.9 ± 0.4 ppm/year), which report a clear change in shape and amplitude when GOSAT observations began contributing to the calculation of global CO₂ in 2009. The low growth rate observed in 2009 is due to the introduction of GOSAT data, which altered the shape of the seasonal cycle. When considering only land-based data (orange lines), the low growth rate in 2009 disappears, and the seasonal cycles appear more consistent throughout the measurement period with an higher amplitude due to the higher carbon fluxes occurring over land regions than over oceans. This clearly shows that any changes to the sampling features that occur during the period covered by the dataset must be taken into account, particularly when evaluating uniform, gap-free datasets such as those provided by ESMs.\nThe high growth rate values observed in 2010, 2012, 2016, 2019 and 2023 are consistent with those reported by the in situ global network coordinated by [[1]](https://library.wmo.int/idurl/4/69057), and are related to El Niño conditions which favour net carbon emissions into the atmosphere on a global scale (see [[9]](https://doi.org/10.5194/acp-25-13053-2025) for a focus on the large growth rate on 2023).\n\n-------------------------------------------------------\nLoad dataset and basic preprocess\n-------------------------------------------------------\nds = download.download_and_transform(*request)\nReplace missing value 1.0e20 with NaN\nConvert xco2 to ppm\n-------------------------------------------------------\nPrepare area weights (cos(lat)) as 2D DataArray (lat, lon)\n-------------------------------------------------------\n-------------------------------------------------------\nland mask: keep only cells with land_fraction > 0.5\n-------------------------------------------------------\nweights masked\n-------------------------------------------------------\nCompute series for both cases: all gridcells and masked (land_fraction>0.5)\n-------------------------------------------------------\narea-weighted global series\ncompute monthly GR and seasonal cycle for both\n-------------------------------------------------------\nQuick diagnostics (optional prints)\n-------------------------------------------------------\nprint(\"Global (all) series length:\", len(global_all_series.dropna()))\nprint(\"Global (masked) series length:\", len(global_masked_series.dropna()))\nprint(\"Fraction positive GR (all):\", (monthly_GR_all.dropna() > 0).mean())\nprint(\"Fraction positive GR (masked):\", (monthly_GR_masked.dropna() > 0).mean())\n-------------------------------------------------------\nPlotting: three vertical plots with both lines\n-------------------------------------------------------\n(1) Global monthly mean XCO2\n(2) Monthly Growth Rate (ppm/yr)\nadd stats box for both series (means and stds)\n(3) Seasonal Cycle (detrended)\ncenter y-axis of seasonal cycle around zero (symmetric limits) using both series combined\n-------------------------\nFormat x-axis to show integer years\n-------------------------\n\n*The figure shows the global time series of monthly XCO$_2$ mean values (upper panel), monthly resolved annual growth rate (middle panel) and detrended seasonal cycles (bottom panel) from 2003 to 2023. Blue lines indicate results for the whole dataset, orange lines indicate results for land-masked data.*"} {"chunk_id": "satellite_satellite-carbon-dioxide_consistency_q04__6a79c177e5b5", "report_id": "satellite_satellite-carbon-dioxide_consistency_q04", "dataset_id": "satellite-carbon-dioxide", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q04", "aspect_base": "consistency", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Consistency of carbon dioxide satellite observations for the evaluation of Earth system models. > Analysis and results > 3. Data analysis > Calculation of the Seasonal Cycle Amplitude (SCA)", "title": "Consistency of carbon dioxide satellite observations for the evaluation of Earth system models.", "chunk_index": 11, "token_count": 316, "text_raw": "In this section, we show the time series of the annual global SCA calculated for the XCO$_2$ variables for the whole dataset (blue) as well as for the land data (orange). Overall, the global SCA shows a decreasing trend over the period 2004 - 2022, however it should be pointed out that discontinuities exist at the time when new satellite observations were added to retrieve XCO$_2$: especially GOSAT in 2009 and OCO-2 in 2015. When the land data only are considered, the negative tendency diappeared from 2008 onward in agreement with the ESM simulations reported by [[6]](https://doi.org/10.5194/bg-17-6115-2020).\n\n---------------------------------------------------------------------\nCompute the SCA (peak-to-trough amplitude in each year, min 7 months of data)\n---------------------------------------------------------------------\n------------------------------------------------------------------------------\n------------------------------------------------------------------------------\n(1) SCA plot\n\n*The figure shows the annual time series of SCA from 2004 to 2022 for the whole dataset (blue) and for land-masked data (orange).*", "text_with_prefix": "EQC Quality Assessment: \"Consistency of carbon dioxide satellite observations for the evaluation of Earth system models.\"\nDataset: satellite-carbon-dioxide [CDS]\nAspect: consistency_q04 | Category: Satellite_ECVs\nSection: Consistency of carbon dioxide satellite observations for the evaluation of Earth system models. > Analysis and results > 3. Data analysis > Calculation of the Seasonal Cycle Amplitude (SCA)\n---\nIn this section, we show the time series of the annual global SCA calculated for the XCO$_2$ variables for the whole dataset (blue) as well as for the land data (orange). Overall, the global SCA shows a decreasing trend over the period 2004 - 2022, however it should be pointed out that discontinuities exist at the time when new satellite observations were added to retrieve XCO$_2$: especially GOSAT in 2009 and OCO-2 in 2015. When the land data only are considered, the negative tendency diappeared from 2008 onward in agreement with the ESM simulations reported by [[6]](https://doi.org/10.5194/bg-17-6115-2020).\n\n---------------------------------------------------------------------\nCompute the SCA (peak-to-trough amplitude in each year, min 7 months of data)\n---------------------------------------------------------------------\n------------------------------------------------------------------------------\n------------------------------------------------------------------------------\n(1) SCA plot\n\n*The figure shows the annual time series of SCA from 2004 to 2022 for the whole dataset (blue) and for land-masked data (orange).*"} {"chunk_id": "satellite_satellite-carbon-dioxide_consistency_q04__2b55871e069e", "report_id": "satellite_satellite-carbon-dioxide_consistency_q04", "dataset_id": "satellite-carbon-dioxide", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q04", "aspect_base": "consistency", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Consistency of carbon dioxide satellite observations for the evaluation of Earth system models. > ℹ️ If you want to know more > Key resources", "title": "Consistency of carbon dioxide satellite observations for the evaluation of Earth system models.", "chunk_index": 12, "token_count": 344, "text_raw": "The CDS catalogue entries for the data used were:\n* Carbon dioxide data from 2002 to present derived from satellite observations: https://cds.climate.copernicus.eu/datasets/satellite-carbon-dioxide?tab=overview\n\nCode libraries used:\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)\n\nUsers interested in accessing official C3S products about long term trends of greenhouse gases from satellite observations are encouraged to access specifically designed Copernicus resources like the [European State of Climate](https://climate.copernicus.eu/esotc/2024/trends-climate-indicators) or the [Global Climate Highlights](https://climate.copernicus.eu/global-climate-highlights-2024).\n\nUsers interested in the investigation of CO$_2$ fluxes can consider to use specifically designed Copernicus resources like [CAMS global inversion-optimised greenhouse gas fluxes and concentrations](https://ads.atmosphere.copernicus.eu/datasets/cams-global-greenhouse-gas-inversion?tab=overview).", "text_with_prefix": "EQC Quality Assessment: \"Consistency of carbon dioxide satellite observations for the evaluation of Earth system models.\"\nDataset: satellite-carbon-dioxide [CDS]\nAspect: consistency_q04 | Category: Satellite_ECVs\nSection: Consistency of carbon dioxide satellite observations for the evaluation of Earth system models. > ℹ️ If you want to know more > Key resources\n---\nThe CDS catalogue entries for the data used were:\n* Carbon dioxide data from 2002 to present derived from satellite observations: https://cds.climate.copernicus.eu/datasets/satellite-carbon-dioxide?tab=overview\n\nCode libraries used:\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)\n\nUsers interested in accessing official C3S products about long term trends of greenhouse gases from satellite observations are encouraged to access specifically designed Copernicus resources like the [European State of Climate](https://climate.copernicus.eu/esotc/2024/trends-climate-indicators) or the [Global Climate Highlights](https://climate.copernicus.eu/global-climate-highlights-2024).\n\nUsers interested in the investigation of CO$_2$ fluxes can consider to use specifically designed Copernicus resources like [CAMS global inversion-optimised greenhouse gas fluxes and concentrations](https://ads.atmosphere.copernicus.eu/datasets/cams-global-greenhouse-gas-inversion?tab=overview)."} {"chunk_id": "satellite_satellite-carbon-dioxide_consistency_q04__773797617def", "report_id": "satellite_satellite-carbon-dioxide_consistency_q04", "dataset_id": "satellite-carbon-dioxide", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q04", "aspect_base": "consistency", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Consistency of carbon dioxide satellite observations for the evaluation of Earth system models. > ℹ️ If you want to know more > References", "title": "Consistency of carbon dioxide satellite observations for the evaluation of Earth system models.", "chunk_index": 13, "token_count": 1020, "text_raw": "[[1]](https://library.wmo.int/idurl/4/69057) World Meteorological Organization. (2024). WMO Greenhouse Gas Bulletin, 20, ISSN 2078-0796.\n\n[[2]](https://doi.org/10.5194/gmd-9-1937-2016) Eyring, V., Bony, S., Meehl, G. A., Senior, C. A., Stevens, B., Stouffer, R. J., and Taylor, K. E. (2016). Overview of the Coupled Model Intercomparison Project Phase 6 (CMIP6) experimental design and organization, Geoscience Model Development, 9, 1937–1958.\n\n[[3]](https://cds.climate.copernicus.eu/datasets/satellite-carbon-dioxide?tab=overview) Copernicus Climate Change Service, Climate Data Store. (2018). Carbon dioxide data from 2002 to present derived from satellite observations. Copernicus Climate Change Service (C3S) Climate Data Store (CDS) (Accessed on 30-Oct-2025).\n\n[[4]](https://doi.org/10.5194/acp-18-17355-2018) Buchwitz, M., Reuter, M., Schneising, O., Noël, S., Gier, B., Bovensmann, H., Burrows, J. P., Boesch, H., Anand, J., Parker, R. J., Somkuti, P., Detmers, R. G., Hasekamp, O. P., Aben, I., Butz, A., Kuze, A., Suto, H., Yoshida, Y., Crisp, D., and O'Dell, C. (2018). Computation and analysis of atmospheric carbon dioxide annual mean growth rates from satellite observations during 2003–2016, Atmospheric Chemistry and Physics, 18, 17355–17370.\n\n[[5]](https://doi.org/10.5194/amt-13-789-2020) Reuter, M., Buchwitz, M., Schneising, O., Noël, S., Bovensmann, H., Burrows, J. P., Boesch, H., Di Noia, A., Anand, J., Parker, R. J., Somkuti, P., Wu, L., Hasekamp, O. P., Aben, I., Kuze, A., Suto, H., Shiomi, K., Yoshida, Y., Morino, I., Crisp, D., O'Dell, C. W., Notholt, J., Petri, C., Warneke, T., Velazco, V. A., Deutscher, N. M., Griffith, D. W. T., Kivi, R., Pollard, D. F., Hase, F., Sussmann, R., Té, Y. V., Strong, K., Roche, S., Sha, M. K., De Mazière, M., Feist, D. G., Iraci, L. T., Roehl, C. M., Retscher, C., and Schepers, D. (2020). Ensemble-based satellite-derived carbon dioxide and methane column-averaged dry-air mole fraction data sets (2003–2018) for carbon and climate applications, Atmospheric Measurement Techniques, 13, 789–819.\n\n[[6]](https://doi.org/10.5194/bg-17-6115-2020) Gier, B. K., Buchwitz, M., Reuter, M., Cox, P. M., Friedlingstein, P., and Eyring, V. (2020). Spatially resolved evaluation of Earth system models with satellite column-averaged CO2, Biogeosciences, 17, 6115–6144.\n\n[[7]](https://confluence.ecmwf.int/pages/viewpage.action?pageId=567160605) Reuter, M., Fuentes Andrade, B., Buchwitz, M. (2025). C3S Greenhouse Gas (GHG:CO2 & CH4) v4.6: Product User Guide and Specification (PUGS), C3S2_313a_DLR_WP1-DDP-GHG-v1_PUGS_XGHG_v4.6.", "text_with_prefix": "EQC Quality Assessment: \"Consistency of carbon dioxide satellite observations for the evaluation of Earth system models.\"\nDataset: satellite-carbon-dioxide [CDS]\nAspect: consistency_q04 | Category: Satellite_ECVs\nSection: Consistency of carbon dioxide satellite observations for the evaluation of Earth system models. > ℹ️ If you want to know more > References\n---\n[[1]](https://library.wmo.int/idurl/4/69057) World Meteorological Organization. (2024). WMO Greenhouse Gas Bulletin, 20, ISSN 2078-0796.\n\n[[2]](https://doi.org/10.5194/gmd-9-1937-2016) Eyring, V., Bony, S., Meehl, G. A., Senior, C. A., Stevens, B., Stouffer, R. J., and Taylor, K. E. (2016). Overview of the Coupled Model Intercomparison Project Phase 6 (CMIP6) experimental design and organization, Geoscience Model Development, 9, 1937–1958.\n\n[[3]](https://cds.climate.copernicus.eu/datasets/satellite-carbon-dioxide?tab=overview) Copernicus Climate Change Service, Climate Data Store. (2018). Carbon dioxide data from 2002 to present derived from satellite observations. Copernicus Climate Change Service (C3S) Climate Data Store (CDS) (Accessed on 30-Oct-2025).\n\n[[4]](https://doi.org/10.5194/acp-18-17355-2018) Buchwitz, M., Reuter, M., Schneising, O., Noël, S., Gier, B., Bovensmann, H., Burrows, J. P., Boesch, H., Anand, J., Parker, R. J., Somkuti, P., Detmers, R. G., Hasekamp, O. P., Aben, I., Butz, A., Kuze, A., Suto, H., Yoshida, Y., Crisp, D., and O'Dell, C. (2018). Computation and analysis of atmospheric carbon dioxide annual mean growth rates from satellite observations during 2003–2016, Atmospheric Chemistry and Physics, 18, 17355–17370.\n\n[[5]](https://doi.org/10.5194/amt-13-789-2020) Reuter, M., Buchwitz, M., Schneising, O., Noël, S., Bovensmann, H., Burrows, J. P., Boesch, H., Di Noia, A., Anand, J., Parker, R. J., Somkuti, P., Wu, L., Hasekamp, O. P., Aben, I., Kuze, A., Suto, H., Shiomi, K., Yoshida, Y., Morino, I., Crisp, D., O'Dell, C. W., Notholt, J., Petri, C., Warneke, T., Velazco, V. A., Deutscher, N. M., Griffith, D. W. T., Kivi, R., Pollard, D. F., Hase, F., Sussmann, R., Té, Y. V., Strong, K., Roche, S., Sha, M. K., De Mazière, M., Feist, D. G., Iraci, L. T., Roehl, C. M., Retscher, C., and Schepers, D. (2020). Ensemble-based satellite-derived carbon dioxide and methane column-averaged dry-air mole fraction data sets (2003–2018) for carbon and climate applications, Atmospheric Measurement Techniques, 13, 789–819.\n\n[[6]](https://doi.org/10.5194/bg-17-6115-2020) Gier, B. K., Buchwitz, M., Reuter, M., Cox, P. M., Friedlingstein, P., and Eyring, V. (2020). Spatially resolved evaluation of Earth system models with satellite column-averaged CO2, Biogeosciences, 17, 6115–6144.\n\n[[7]](https://confluence.ecmwf.int/pages/viewpage.action?pageId=567160605) Reuter, M., Fuentes Andrade, B., Buchwitz, M. (2025). C3S Greenhouse Gas (GHG:CO2 & CH4) v4.6: Product User Guide and Specification (PUGS), C3S2_313a_DLR_WP1-DDP-GHG-v1_PUGS_XGHG_v4.6."} {"chunk_id": "satellite_satellite-carbon-dioxide_consistency_q04__d01d98dea466", "report_id": "satellite_satellite-carbon-dioxide_consistency_q04", "dataset_id": "satellite-carbon-dioxide", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q04", "aspect_base": "consistency", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Consistency of carbon dioxide satellite observations for the evaluation of Earth system models. > ℹ️ If you want to know more > References", "title": "Consistency of carbon dioxide satellite observations for the evaluation of Earth system models.", "chunk_index": 14, "token_count": 369, "text_raw": "& CH4) v4.6: Product User Guide and Specification (PUGS), C3S2_313a_DLR_WP1-DDP-GHG-v1_PUGS_XGHG_v4.6.\n\n[[8]](https://confluence.ecmwf.int/pages/viewpage.action?pageId=567160577) Reuter, M., Fuentes Andrade, B., Buchwitz, M. (2025). C3S Greenhouse Gas (GHG: CO2 & CH4) v4.6: Algorithm Theoretical Basis Document (ATBD), C3S2_313a_DLR_WP1-DDP-GHG-v1_ATBD_XGHG_v4.6.\n\n[[9]](https://doi.org/10.5194/acp-25-13053-2025) Feng, L., Palmer, P. I., Smallman, L., Xiao, J., Cristofanelli, P., Hermansen, O., Lee, J., Labuschagne, C., Montaguti, S., Noe, S. M., Platt, S. M., Ren, X., Steinbacher, M., and Xueref-Remy, I. (2025). The role of the tropical carbon balance in determining the large atmospheric CO2 growth rate in 2023, Atmospheric Chemistry and Physics, 25, 13053–13076.", "text_with_prefix": "EQC Quality Assessment: \"Consistency of carbon dioxide satellite observations for the evaluation of Earth system models.\"\nDataset: satellite-carbon-dioxide [CDS]\nAspect: consistency_q04 | Category: Satellite_ECVs\nSection: Consistency of carbon dioxide satellite observations for the evaluation of Earth system models. > ℹ️ If you want to know more > References\n---\n& CH4) v4.6: Product User Guide and Specification (PUGS), C3S2_313a_DLR_WP1-DDP-GHG-v1_PUGS_XGHG_v4.6.\n\n[[8]](https://confluence.ecmwf.int/pages/viewpage.action?pageId=567160577) Reuter, M., Fuentes Andrade, B., Buchwitz, M. (2025). C3S Greenhouse Gas (GHG: CO2 & CH4) v4.6: Algorithm Theoretical Basis Document (ATBD), C3S2_313a_DLR_WP1-DDP-GHG-v1_ATBD_XGHG_v4.6.\n\n[[9]](https://doi.org/10.5194/acp-25-13053-2025) Feng, L., Palmer, P. I., Smallman, L., Xiao, J., Cristofanelli, P., Hermansen, O., Lee, J., Labuschagne, C., Montaguti, S., Noe, S. M., Platt, S. M., Ren, X., Steinbacher, M., and Xueref-Remy, I. (2025). The role of the tropical carbon balance in determining the large atmospheric CO2 growth rate in 2023, Atmospheric Chemistry and Physics, 25, 13053–13076."} {"chunk_id": "satellite_satellite-carbon-dioxide_resolution_q03__8a751492c30f", "report_id": "satellite_satellite-carbon-dioxide_resolution_q03", "dataset_id": "satellite-carbon-dioxide", "store": "CDS", "doc_type": "EQC_QA", "aspect": "resolution_q03", "aspect_base": "resolution", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Spatial resolution and completeness of carbon dioxide satellite observations for quantifying carbon fluxes", "title": "Spatial resolution and completeness of carbon dioxide satellite observations for quantifying carbon fluxes", "chunk_index": 0, "token_count": 1007, "text_raw": "Production date: 09-04-2026\n\nProduced by: Paolo Cristofanelli (CNR)\n\nPlease note that this assessment also refers to the deprecated Level 2 XCO$_2$ data products which won't be extended after 2022.\n\n## 🌍 Use case: Using satellite observations for investigating the global carbon fluxes\n\n## ❓ Quality assessment question\n* **Are the CO$_2$ satellite observations suitable for deriving global and regional carbon fluxes in terms of resolution and spatial completeness?**\n\nCarbon dioxide (CO$_2$) is the most important anthropogenic greenhouse gas, accounting for almost 64% of the total radiative forcing by long-lived greenhouse gases [[1]](https://library.wmo.int/viewer/68532/?offset=#page=4&viewer=picture&o=bookmarks&n=0&q=). The atmospheric concentration of CO$_2$ in 2023 was 419.3$\\,\\pm\\,$0.1 ppm [[2]](https://essd.copernicus.org/articles/17/965/2025/).\n\nAccurate spatial and temporal quantification of carbon dioxide (CO$_2$) is central to understanding the carbon cycle and supporting policy decisions on climate change mitigation. Satellite retrievals (Level 2 data products) of column-averaged dry air mole fractions of CO$_2$ (XCO$_2$) have been widely used to infer regional and global variations in carbon fluxes through atmospheric inversion modelling (e.g., [[3]](https://doi.org/10.5194/acp-14-13739-2014), [[4]](https://doi.org/10.1175/BAMS-D-15-00310.1), [[5]](https://acp.copernicus.org/articles/19/14233/2019/), [[6]](https://doi.org/10.5194/acp-19-12067-2019), [[7]](https://essd.copernicus.org/articles/16/2857/2024/), [[8]](https://climate.copernicus.eu/esotc/2023/greenhouse-gas-fluxes)).\n\nAtmospheric inverse modelling [[9]](https://doi.org/10.5194/bg-10-6699-2013) links CO$_2$ fluxes to observed atmospheric CO$_2$ mole fractions (like those provided by satellite measurements) using atmospheric transport (and chemistry) models, and is often referred to as \"top-down\" approach. Typically \"top-down\" inverse systems estimate residual natural or fluxes not related to fossil fuel consumption and cement production (FFC) from land and ocean regions (e.g., [[10]](https://acp.copernicus.org/articles/22/9215/2022/) and [[7]](https://essd.copernicus.org/articles/16/2857/2024/)).\n\nAtmospheric inversion systems estimate carbon fluxes by adjusting modelled CO$_2$ fluxes to match atmospheric observations. The so-called \"a‑priori\" flux fields provide initial modelled CO$_2$ fluxes by combining, e.g., estimates of anthropogenic emissions from statistical inventories, terrestrial biosphere and oceanic fluxes from models or climatologies and emissions from vegetation fires. An example of a global-scale inversion system and adopted \"a-priori\" fields is represented by the CAMS \"PYVAR\" inversion system [[11]](https://confluence.ecmwf.int/display/CKB/Description+of+the+CO2+inversion+production+chain).\n\nDue to their large spatial coverage, satellite observations are used for constraining atmospheric inversions by complementing independent datasets provided by high-precision in-situ measurements like those provided by [[12]](https://doi.org/10.15138/9N0H-ZH07).\n\nIn this assessment, through a review of existing literature and use cases combined with specific data analyses, we investigate if the spatial resolution and coverage of the dataset \"Carbon dioxide data from 2002 to present derived from satellite observations\" provided by [[13]](https://cds.climate.copernicus.eu/datasets/satellite-carbon-dioxide?tab=overview) are suitable to constrain global and regional Earth's carbon fluxes in \"top-down\" inversion systems.\n\n![image.png](attachment:196b9642-4db0-4282-bb42-20bffbd84814.png)", "text_with_prefix": "EQC Quality Assessment: \"Spatial resolution and completeness of carbon dioxide satellite observations for quantifying carbon fluxes\"\nDataset: satellite-carbon-dioxide [CDS]\nAspect: resolution_q03 | Category: Satellite_ECVs\nSection: Spatial resolution and completeness of carbon dioxide satellite observations for quantifying carbon fluxes\n---\nProduction date: 09-04-2026\n\nProduced by: Paolo Cristofanelli (CNR)\n\nPlease note that this assessment also refers to the deprecated Level 2 XCO$_2$ data products which won't be extended after 2022.\n\n## 🌍 Use case: Using satellite observations for investigating the global carbon fluxes\n\n## ❓ Quality assessment question\n* **Are the CO$_2$ satellite observations suitable for deriving global and regional carbon fluxes in terms of resolution and spatial completeness?**\n\nCarbon dioxide (CO$_2$) is the most important anthropogenic greenhouse gas, accounting for almost 64% of the total radiative forcing by long-lived greenhouse gases [[1]](https://library.wmo.int/viewer/68532/?offset=#page=4&viewer=picture&o=bookmarks&n=0&q=). The atmospheric concentration of CO$_2$ in 2023 was 419.3$\\,\\pm\\,$0.1 ppm [[2]](https://essd.copernicus.org/articles/17/965/2025/).\n\nAccurate spatial and temporal quantification of carbon dioxide (CO$_2$) is central to understanding the carbon cycle and supporting policy decisions on climate change mitigation. Satellite retrievals (Level 2 data products) of column-averaged dry air mole fractions of CO$_2$ (XCO$_2$) have been widely used to infer regional and global variations in carbon fluxes through atmospheric inversion modelling (e.g., [[3]](https://doi.org/10.5194/acp-14-13739-2014), [[4]](https://doi.org/10.1175/BAMS-D-15-00310.1), [[5]](https://acp.copernicus.org/articles/19/14233/2019/), [[6]](https://doi.org/10.5194/acp-19-12067-2019), [[7]](https://essd.copernicus.org/articles/16/2857/2024/), [[8]](https://climate.copernicus.eu/esotc/2023/greenhouse-gas-fluxes)).\n\nAtmospheric inverse modelling [[9]](https://doi.org/10.5194/bg-10-6699-2013) links CO$_2$ fluxes to observed atmospheric CO$_2$ mole fractions (like those provided by satellite measurements) using atmospheric transport (and chemistry) models, and is often referred to as \"top-down\" approach. Typically \"top-down\" inverse systems estimate residual natural or fluxes not related to fossil fuel consumption and cement production (FFC) from land and ocean regions (e.g., [[10]](https://acp.copernicus.org/articles/22/9215/2022/) and [[7]](https://essd.copernicus.org/articles/16/2857/2024/)).\n\nAtmospheric inversion systems estimate carbon fluxes by adjusting modelled CO$_2$ fluxes to match atmospheric observations. The so-called \"a‑priori\" flux fields provide initial modelled CO$_2$ fluxes by combining, e.g., estimates of anthropogenic emissions from statistical inventories, terrestrial biosphere and oceanic fluxes from models or climatologies and emissions from vegetation fires. An example of a global-scale inversion system and adopted \"a-priori\" fields is represented by the CAMS \"PYVAR\" inversion system [[11]](https://confluence.ecmwf.int/display/CKB/Description+of+the+CO2+inversion+production+chain).\n\nDue to their large spatial coverage, satellite observations are used for constraining atmospheric inversions by complementing independent datasets provided by high-precision in-situ measurements like those provided by [[12]](https://doi.org/10.15138/9N0H-ZH07).\n\nIn this assessment, through a review of existing literature and use cases combined with specific data analyses, we investigate if the spatial resolution and coverage of the dataset \"Carbon dioxide data from 2002 to present derived from satellite observations\" provided by [[13]](https://cds.climate.copernicus.eu/datasets/satellite-carbon-dioxide?tab=overview) are suitable to constrain global and regional Earth's carbon fluxes in \"top-down\" inversion systems.\n\n![image.png](attachment:196b9642-4db0-4282-bb42-20bffbd84814.png)"} {"chunk_id": "satellite_satellite-carbon-dioxide_resolution_q03__1904a5b5399c", "report_id": "satellite_satellite-carbon-dioxide_resolution_q03", "dataset_id": "satellite-carbon-dioxide", "store": "CDS", "doc_type": "EQC_QA", "aspect": "resolution_q03", "aspect_base": "resolution", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Spatial resolution and completeness of carbon dioxide satellite observations for quantifying carbon fluxes", "title": "Spatial resolution and completeness of carbon dioxide satellite observations for quantifying carbon fluxes", "chunk_index": 1, "token_count": 1112, "text_raw": "suitable to constrain global and regional Earth's carbon fluxes in \"top-down\" inversion systems.\n\n![image.png](attachment:196b9642-4db0-4282-bb42-20bffbd84814.png)\n\n*The figure shows the global carbon budget estimated by the Global ObservatioN-based system for monitoring Greenhouse GAses (GONGGA) atmospheric inversion system through assimilation of Orbiting Carbon Observatory-2 (OCO-2) XCO$_2$ data and the global atmospheric growth rate by [[12]](https://doi.org/10.15138/9N0H-ZH07). $E_{\\mathrm{FOS}}$ denotes fossil fuel CO$_2$ emissions, $E_{\\mathrm{FIRE}}$ denotes biomass combustion emissions, $F_{\\mathrm{OCEAN}}$ denotes ocean-atmosphere carbon fluxes, NEE denotes net ecosystem exchange (i.e. the balance of photosynthesis and respiration from the terrestrial ecosystem) and $G_{\\mathrm{ATM}}$denotes the growth rate of atmospheric CO$_2$ concentration derived from GONGGA and [[12]](https://doi.org/10.15138/9N0H-ZH07). Image reproduced from [[7]](https://essd.copernicus.org/articles/16/2857/2024/) under CC-BY licence.*\n\n## 📢 Quality assessment statement\n\nThese are the key outcomes of this assessment\n\n* The XCO2_OBS4MIPS Level 3 product has been primarily generated for comparison with climate models and has also been used for computations of annual mean atmospheric growth rates. This data product in not recommended to derive global and regional carbon fluxes.\n* In terms of spatial resolution, the mid-tropospheric-averaged air mole fractions of CO$_2$ (MTCO2_OBS4MIPS, Level 3) data product appears appropriate for use in global and regional inversion system. However, users must carefully consider limitations related with spatial completeness (mostly limited to tropical regions) and vertical representativeness. As there are no known applications of this MT-CO$_2$ product, extreme caution should be exercised to use this data product to derive global and regional carbon fluxes by inversion modelling systems. \n* The deprecated Level 2 XCO$_2$ datasets can be used within inversion modelling systems to derive historical global and regional carbon fluxes. Users would use the data product derived by the algorithm XCO2_EMMA that optimise data availability over regions frequently affected by cloud cover or high aerosol loading.\n* The spatial coverage and resolution (spatial density) of the satellite data or a combination of them, affected the impact of satellite observations on carbon flux quantifications in different regions. \n* As the MTCO2_OBS4MIPS dataset is updated every six months and Level 2 XCO$_2$ won't be extended in time after 2022, users interested in more timely greenhouse gas fluxes quantification should consider other [Copernicus resources](https://ads.atmosphere.copernicus.eu/datasets/cams-global-greenhouse-gas-inversion?tab=overview).\n```\n\n## 📋 Methodology\n\nTo evaluate if the spatial coverage and resolution of satellite-based observations characterizing the Climate Data Store dataset “Carbon dioxide data from 2002 to present derived from satellite observations” [[13]](https://cds.climate.copernicus.eu/datasets/satellite-carbon-dioxide?tab=overview) are suitable for their usage in an atmospheric inversion system to investigate Earth's carbon fluxes, we inspected the related CDS Documentation [[14]](https://confluence.ecmwf.int/pages/viewpage.action?pageId=567160575) as well as external references like [[15]](http://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PUGS_GHG_main_latest.pdf), and case studies available from scientific literature ([[3]](https://doi.org/10.5194/acp-14-13739-2014), [[5]](https://acp.copernicus.org/articles/19/14233/2019/), [[6]](https://doi.org/10.5194/acp-19-12067-2019), [[7]](https://essd.copernicus.org/articles/16/2857/2024/)) and applications (e.g., [[8]](https://climate.copernicus.eu/esotc/2023/greenhouse-gas-fluxes), [[11]](https://confluence.ecmwf.int/display/CKB/Description+of+the+CO2+inversion+production+chain)). Moreover, we performed specific analyses on the spatial data coverage of the Level 3 data product Obs4MIPs (version 4.6) and IASI-MERGED-OBS4MIPS (version 10.1).", "text_with_prefix": "EQC Quality Assessment: \"Spatial resolution and completeness of carbon dioxide satellite observations for quantifying carbon fluxes\"\nDataset: satellite-carbon-dioxide [CDS]\nAspect: resolution_q03 | Category: Satellite_ECVs\nSection: Spatial resolution and completeness of carbon dioxide satellite observations for quantifying carbon fluxes\n---\nsuitable to constrain global and regional Earth's carbon fluxes in \"top-down\" inversion systems.\n\n![image.png](attachment:196b9642-4db0-4282-bb42-20bffbd84814.png)\n\n*The figure shows the global carbon budget estimated by the Global ObservatioN-based system for monitoring Greenhouse GAses (GONGGA) atmospheric inversion system through assimilation of Orbiting Carbon Observatory-2 (OCO-2) XCO$_2$ data and the global atmospheric growth rate by [[12]](https://doi.org/10.15138/9N0H-ZH07). $E_{\\mathrm{FOS}}$ denotes fossil fuel CO$_2$ emissions, $E_{\\mathrm{FIRE}}$ denotes biomass combustion emissions, $F_{\\mathrm{OCEAN}}$ denotes ocean-atmosphere carbon fluxes, NEE denotes net ecosystem exchange (i.e. the balance of photosynthesis and respiration from the terrestrial ecosystem) and $G_{\\mathrm{ATM}}$denotes the growth rate of atmospheric CO$_2$ concentration derived from GONGGA and [[12]](https://doi.org/10.15138/9N0H-ZH07). Image reproduced from [[7]](https://essd.copernicus.org/articles/16/2857/2024/) under CC-BY licence.*\n\n## 📢 Quality assessment statement\n\nThese are the key outcomes of this assessment\n\n* The XCO2_OBS4MIPS Level 3 product has been primarily generated for comparison with climate models and has also been used for computations of annual mean atmospheric growth rates. This data product in not recommended to derive global and regional carbon fluxes.\n* In terms of spatial resolution, the mid-tropospheric-averaged air mole fractions of CO$_2$ (MTCO2_OBS4MIPS, Level 3) data product appears appropriate for use in global and regional inversion system. However, users must carefully consider limitations related with spatial completeness (mostly limited to tropical regions) and vertical representativeness. As there are no known applications of this MT-CO$_2$ product, extreme caution should be exercised to use this data product to derive global and regional carbon fluxes by inversion modelling systems. \n* The deprecated Level 2 XCO$_2$ datasets can be used within inversion modelling systems to derive historical global and regional carbon fluxes. Users would use the data product derived by the algorithm XCO2_EMMA that optimise data availability over regions frequently affected by cloud cover or high aerosol loading.\n* The spatial coverage and resolution (spatial density) of the satellite data or a combination of them, affected the impact of satellite observations on carbon flux quantifications in different regions. \n* As the MTCO2_OBS4MIPS dataset is updated every six months and Level 2 XCO$_2$ won't be extended in time after 2022, users interested in more timely greenhouse gas fluxes quantification should consider other [Copernicus resources](https://ads.atmosphere.copernicus.eu/datasets/cams-global-greenhouse-gas-inversion?tab=overview).\n```\n\n## 📋 Methodology\n\nTo evaluate if the spatial coverage and resolution of satellite-based observations characterizing the Climate Data Store dataset “Carbon dioxide data from 2002 to present derived from satellite observations” [[13]](https://cds.climate.copernicus.eu/datasets/satellite-carbon-dioxide?tab=overview) are suitable for their usage in an atmospheric inversion system to investigate Earth's carbon fluxes, we inspected the related CDS Documentation [[14]](https://confluence.ecmwf.int/pages/viewpage.action?pageId=567160575) as well as external references like [[15]](http://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PUGS_GHG_main_latest.pdf), and case studies available from scientific literature ([[3]](https://doi.org/10.5194/acp-14-13739-2014), [[5]](https://acp.copernicus.org/articles/19/14233/2019/), [[6]](https://doi.org/10.5194/acp-19-12067-2019), [[7]](https://essd.copernicus.org/articles/16/2857/2024/)) and applications (e.g., [[8]](https://climate.copernicus.eu/esotc/2023/greenhouse-gas-fluxes), [[11]](https://confluence.ecmwf.int/display/CKB/Description+of+the+CO2+inversion+production+chain)). Moreover, we performed specific analyses on the spatial data coverage of the Level 3 data product Obs4MIPs (version 4.6) and IASI-MERGED-OBS4MIPS (version 10.1)."} {"chunk_id": "satellite_satellite-carbon-dioxide_resolution_q03__bab49ed02df5", "report_id": "satellite_satellite-carbon-dioxide_resolution_q03", "dataset_id": "satellite-carbon-dioxide", "store": "CDS", "doc_type": "EQC_QA", "aspect": "resolution_q03", "aspect_base": "resolution", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Spatial resolution and completeness of carbon dioxide satellite observations for quantifying carbon fluxes", "title": "Spatial resolution and completeness of carbon dioxide satellite observations for quantifying carbon fluxes", "chunk_index": 2, "token_count": 989, "text_raw": ")). Moreover, we performed specific analyses on the spatial data coverage of the Level 3 data product Obs4MIPs (version 4.6) and IASI-MERGED-OBS4MIPS (version 10.1).\n\nIn particular:\n * [[3]](https://doi.org/10.5194/acp-14-13739-2014) inferred regional estimates of the European land carbon sink by performing a regional surface flux inversion using earlier versions of the XCO$_2$ satellite measurements from SCIAMACHY (BESD v02.00.08 and) and GOSAT (ACOS v3.4r03 2010, UoLFP v4.0 2010, RemoTeC v2.11 2010, and NIES v02.xx 2010) than those available from the Climate Data Store dataset “Carbon dioxide data from 2002 to present derived from satellite observations” [[13]](https://cds.climate.copernicus.eu/datasets/satellite-carbon-dioxide?tab=overview).\n\n* [[5]](https://acp.copernicus.org/articles/19/14233/2019/) evaluated an ensemble of six multi-year global CO$_2$ atmospheric inversions by using a large dataset of accurate aircraft measurements in the free troposphere over the globe. For the inversion experiments, besides using OCO-2 data (ACOS bias-corrected retrievals, version 9), they also considered a previous version of a Level 2 data product available by the Climate Data Store (CO2_GOS_OCFP, v7.1).\n\n* [[6]](https://doi.org/10.5194/acp-19-12067-2019) compared the results of constraining the terrestrial ecosystem carbon fluxes by assimilating GOSAT and OCO-2 XCO$_2$ retrievals (both produced by the \"NASA Atmospheric CO$_2$ Observations from Space\" project, version b7.3) within the GEOS-Chem 4D-Var assimilation framework.\n\n* [[7]](https://essd.copernicus.org/articles/16/2857/2024/) presented a global spatially resolved terrestrial and ocean carbon flux dataset for years 2015–2022 generated by an atmospheric inversion system through the assimilation of OCO-2 XCO$_2$ retrievals (Level 2, Lite v11r data product).\n\n* [[8]](https://climate.copernicus.eu/esotc/2023/greenhouse-gas-fluxes) provide a case about the use of CO$_2$ satellite observations to estimate net CO$_2$ fluxes into the atmosphere at global scale and over specific land and oceans regions by using the [CAMS](https://confluence.ecmwf.int/display/CKB/Description+of+the+CO2+inversion+production+chain) global inversion systems. This inversion system uses in-situ atmospheric observations from large living databases, like [NOAA Earth System Research Laboratory Observation Package](https://www.esrl.noaa.gov/gmd/ccgg/obspack/), [Integrated Carbon Observation System](https://icos-atc.lsce.ipsl.fr/), and satellite retrievals from OCO-2.\n\nPlease note that the analysed references referred to the use of Level 2 data products within atmospheric inversion systems, for which there are example applications. Level 3 data products for satellite XCO$_2$ or MT-CO$_2$ have not yet been used in inversion systems to derive quantifications of surface CO$_2$ fluxes. However, a critical comparison of the resolution and completeness features of the Level 3 data products with respect to known applications based on Level 2 data products has been performed.\n\nIn the following, we report the outline of this notebook.\n\n**[](template:section-1)**\n * Import all relevant packages.\n * Cache needed functions.\n\n**[](template:section-2)**\n * Define the data request to CDS.\n\n**[](template:section-3)**\n * We compared the spatial resolution of the “Carbon dioxide data from 2002 to present derived from satellite observations” [[13]](https://cds.climate.copernicus.eu/datasets/satellite-carbon-dioxide?tab=overview) data products with those of the the satellite datasets used by atmospheric inversion experiments documented above.", "text_with_prefix": "EQC Quality Assessment: \"Spatial resolution and completeness of carbon dioxide satellite observations for quantifying carbon fluxes\"\nDataset: satellite-carbon-dioxide [CDS]\nAspect: resolution_q03 | Category: Satellite_ECVs\nSection: Spatial resolution and completeness of carbon dioxide satellite observations for quantifying carbon fluxes\n---\n)). Moreover, we performed specific analyses on the spatial data coverage of the Level 3 data product Obs4MIPs (version 4.6) and IASI-MERGED-OBS4MIPS (version 10.1).\n\nIn particular:\n * [[3]](https://doi.org/10.5194/acp-14-13739-2014) inferred regional estimates of the European land carbon sink by performing a regional surface flux inversion using earlier versions of the XCO$_2$ satellite measurements from SCIAMACHY (BESD v02.00.08 and) and GOSAT (ACOS v3.4r03 2010, UoLFP v4.0 2010, RemoTeC v2.11 2010, and NIES v02.xx 2010) than those available from the Climate Data Store dataset “Carbon dioxide data from 2002 to present derived from satellite observations” [[13]](https://cds.climate.copernicus.eu/datasets/satellite-carbon-dioxide?tab=overview).\n\n* [[5]](https://acp.copernicus.org/articles/19/14233/2019/) evaluated an ensemble of six multi-year global CO$_2$ atmospheric inversions by using a large dataset of accurate aircraft measurements in the free troposphere over the globe. For the inversion experiments, besides using OCO-2 data (ACOS bias-corrected retrievals, version 9), they also considered a previous version of a Level 2 data product available by the Climate Data Store (CO2_GOS_OCFP, v7.1).\n\n* [[6]](https://doi.org/10.5194/acp-19-12067-2019) compared the results of constraining the terrestrial ecosystem carbon fluxes by assimilating GOSAT and OCO-2 XCO$_2$ retrievals (both produced by the \"NASA Atmospheric CO$_2$ Observations from Space\" project, version b7.3) within the GEOS-Chem 4D-Var assimilation framework.\n\n* [[7]](https://essd.copernicus.org/articles/16/2857/2024/) presented a global spatially resolved terrestrial and ocean carbon flux dataset for years 2015–2022 generated by an atmospheric inversion system through the assimilation of OCO-2 XCO$_2$ retrievals (Level 2, Lite v11r data product).\n\n* [[8]](https://climate.copernicus.eu/esotc/2023/greenhouse-gas-fluxes) provide a case about the use of CO$_2$ satellite observations to estimate net CO$_2$ fluxes into the atmosphere at global scale and over specific land and oceans regions by using the [CAMS](https://confluence.ecmwf.int/display/CKB/Description+of+the+CO2+inversion+production+chain) global inversion systems. This inversion system uses in-situ atmospheric observations from large living databases, like [NOAA Earth System Research Laboratory Observation Package](https://www.esrl.noaa.gov/gmd/ccgg/obspack/), [Integrated Carbon Observation System](https://icos-atc.lsce.ipsl.fr/), and satellite retrievals from OCO-2.\n\nPlease note that the analysed references referred to the use of Level 2 data products within atmospheric inversion systems, for which there are example applications. Level 3 data products for satellite XCO$_2$ or MT-CO$_2$ have not yet been used in inversion systems to derive quantifications of surface CO$_2$ fluxes. However, a critical comparison of the resolution and completeness features of the Level 3 data products with respect to known applications based on Level 2 data products has been performed.\n\nIn the following, we report the outline of this notebook.\n\n**[](template:section-1)**\n * Import all relevant packages.\n * Cache needed functions.\n\n**[](template:section-2)**\n * Define the data request to CDS.\n\n**[](template:section-3)**\n * We compared the spatial resolution of the “Carbon dioxide data from 2002 to present derived from satellite observations” [[13]](https://cds.climate.copernicus.eu/datasets/satellite-carbon-dioxide?tab=overview) data products with those of the the satellite datasets used by atmospheric inversion experiments documented above."} {"chunk_id": "satellite_satellite-carbon-dioxide_resolution_q03__d79dcd8a9f0a", "report_id": "satellite_satellite-carbon-dioxide_resolution_q03", "dataset_id": "satellite-carbon-dioxide", "store": "CDS", "doc_type": "EQC_QA", "aspect": "resolution_q03", "aspect_base": "resolution", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Spatial resolution and completeness of carbon dioxide satellite observations for quantifying carbon fluxes", "title": "Spatial resolution and completeness of carbon dioxide satellite observations for quantifying carbon fluxes", "chunk_index": 3, "token_count": 789, "text_raw": "(https://cds.climate.copernicus.eu/datasets/satellite-carbon-dioxide?tab=overview) data products with those of the the satellite datasets used by atmospheric inversion experiments documented above.\n\n**[](template:section-4)**\n * We created plots of the monthly mean data coverage of the Level 3 data product Obs4MIPs (version 4.6) and IASI-MERGED-OBS4MIPS (version 10.1) and we compared them with those of the satellite datasets used by atmospheric inversion experiments documented above. Moreover, we also considered external documentation ([[15]](http://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PUGS_GHG_main_latest.pdf)) to assess the spatial coverage of the (deprecated) Level 2 data products part of the “Carbon dioxide data from 2002 to present derived from satellite observations” [[13]](https://cds.climate.copernicus.eu/datasets/satellite-carbon-dioxide?tab=overview) dataset.\n\n**[](template:section-5)**\n * We provide a summary of the main outcomes of this assessment.\n\n## 📈 Analysis and results\n\n(template:section-1)=\n### 1. Set-up the code and choose the data to use\n\n#### Import all relevant packages\nIn this section, we import all the relevant packages needed for running the notebook.\n\n#### Cache needed functions\nIn this section, we cached a list of functions used in the analyses.\n\n* The `convert_units` function rescales XCO$_2$ mole fraction to parts per million (ppm).\n\n* The `compute_coverage_obs4mips` function calculates the spatial fractional coverage of the available data for the Obs4MIPs Level 3 dataset.\n\n* The `compute_coverage_iasi` and `aggregate_coverage_iasi` functions calculate the spatial fractional coverage of the available data for the IASI-MERGED-OBS4MIPS Level 3 dataset.\n\n(template:section-2)=\n### 2. Retrieve data\n\n#### 2.1 Obs4MIPs\nIn this section, we define the data request to CDS (data product Obs4MIPs, Level 3, version 4.6, XCO$_2$) and download the full dataset.\n\n#### 2.2 IASI-MERGED-OBS4MIPS\nIn this section, we define the data request to CDS (data product IASI-MERGED-OBS4MIPS, Level 3, version 10.1, MT-CO$_2$) and download the full dataset (please note that users can modify the time period).\n\nTime parameters\n\n(template:section-3)=\n### 3. Spatial resolution assessment\nIdeally, the spatial resolution of the satellite observations used in an inversion system should match that of the transport model. In turn, the transport model should match the spatial scale of carbon fluxes, which is typically on the order of hundreds of metres. In practice, however, contemporary models generally have much coarser resolutions, typically on the order of degrees ([[16]](https://acp.copernicus.org/articles/19/14233/2019/)), and errors in the simulated atmospheric transport may also affect the inversion results ([[17]](https://doi.org/10.1029/2005JD006390#jgrd12458-bib-0008)).", "text_with_prefix": "EQC Quality Assessment: \"Spatial resolution and completeness of carbon dioxide satellite observations for quantifying carbon fluxes\"\nDataset: satellite-carbon-dioxide [CDS]\nAspect: resolution_q03 | Category: Satellite_ECVs\nSection: Spatial resolution and completeness of carbon dioxide satellite observations for quantifying carbon fluxes\n---\n(https://cds.climate.copernicus.eu/datasets/satellite-carbon-dioxide?tab=overview) data products with those of the the satellite datasets used by atmospheric inversion experiments documented above.\n\n**[](template:section-4)**\n * We created plots of the monthly mean data coverage of the Level 3 data product Obs4MIPs (version 4.6) and IASI-MERGED-OBS4MIPS (version 10.1) and we compared them with those of the satellite datasets used by atmospheric inversion experiments documented above. Moreover, we also considered external documentation ([[15]](http://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PUGS_GHG_main_latest.pdf)) to assess the spatial coverage of the (deprecated) Level 2 data products part of the “Carbon dioxide data from 2002 to present derived from satellite observations” [[13]](https://cds.climate.copernicus.eu/datasets/satellite-carbon-dioxide?tab=overview) dataset.\n\n**[](template:section-5)**\n * We provide a summary of the main outcomes of this assessment.\n\n## 📈 Analysis and results\n\n(template:section-1)=\n### 1. Set-up the code and choose the data to use\n\n#### Import all relevant packages\nIn this section, we import all the relevant packages needed for running the notebook.\n\n#### Cache needed functions\nIn this section, we cached a list of functions used in the analyses.\n\n* The `convert_units` function rescales XCO$_2$ mole fraction to parts per million (ppm).\n\n* The `compute_coverage_obs4mips` function calculates the spatial fractional coverage of the available data for the Obs4MIPs Level 3 dataset.\n\n* The `compute_coverage_iasi` and `aggregate_coverage_iasi` functions calculate the spatial fractional coverage of the available data for the IASI-MERGED-OBS4MIPS Level 3 dataset.\n\n(template:section-2)=\n### 2. Retrieve data\n\n#### 2.1 Obs4MIPs\nIn this section, we define the data request to CDS (data product Obs4MIPs, Level 3, version 4.6, XCO$_2$) and download the full dataset.\n\n#### 2.2 IASI-MERGED-OBS4MIPS\nIn this section, we define the data request to CDS (data product IASI-MERGED-OBS4MIPS, Level 3, version 10.1, MT-CO$_2$) and download the full dataset (please note that users can modify the time period).\n\nTime parameters\n\n(template:section-3)=\n### 3. Spatial resolution assessment\nIdeally, the spatial resolution of the satellite observations used in an inversion system should match that of the transport model. In turn, the transport model should match the spatial scale of carbon fluxes, which is typically on the order of hundreds of metres. In practice, however, contemporary models generally have much coarser resolutions, typically on the order of degrees ([[16]](https://acp.copernicus.org/articles/19/14233/2019/)), and errors in the simulated atmospheric transport may also affect the inversion results ([[17]](https://doi.org/10.1029/2005JD006390#jgrd12458-bib-0008))."} {"chunk_id": "satellite_satellite-carbon-dioxide_resolution_q03__b3c09ff33805", "report_id": "satellite_satellite-carbon-dioxide_resolution_q03", "dataset_id": "satellite-carbon-dioxide", "store": "CDS", "doc_type": "EQC_QA", "aspect": "resolution_q03", "aspect_base": "resolution", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Spatial resolution and completeness of carbon dioxide satellite observations for quantifying carbon fluxes", "title": "Spatial resolution and completeness of carbon dioxide satellite observations for quantifying carbon fluxes", "chunk_index": 4, "token_count": 840, "text_raw": "inversion results ([[17]](https://doi.org/10.1029/2005JD006390#jgrd12458-bib-0008)).\n\nThe documented use cases ([[5]](https://acp.copernicus.org/articles/19/14233/2019/), [[6]](https://doi.org/10.5194/acp-19-12067-2019),[[7]](https://essd.copernicus.org/articles/16/2857/2024/), [[8]](https://climate.copernicus.eu/esotc/2023/greenhouse-gas-fluxes)) used OCO-2 observations characterised by spatially dense data with a narrow swath (10.3 km) and with footprints of a few square kilometres (1.29 km × 2.25 km). [[5]](https://acp.copernicus.org/articles/19/14233/2019/) and [[6]](https://doi.org/10.5194/acp-19-12067-2019) also used a previous version of the CDS GOSAT Level 2 deprecated dataset, characterised by a coarser-resolution data (100 km$^2$ at nadir) with lower spatial density than OCO-2. A more recent application is the inversion system that generates the CAMS global CO$_2$ atmospheric inversion product (PyVAR, see [[11]](https://confluence.ecmwf.int/display/CKB/Description+of+the+CO2+inversion+production+chain)): it uses OCO-2 bias-corrected land retrievals of XCO$_2$ after averaging glint and nadir OCO-2 retrievals in 10-s bins (corresponding to a ground-track swaths of 67 km) to match the transport model horizontal spatial resolution of about 90 km$^{2}$.\n\n* **XCO2_OBS4MIPS Level 3**: the spatial aggregation at 5° × 5° resolution and the temporal aggregation at monthly scales inevitably reduce the amount of information available to the inversion system. Such levels of aggregation are not commonly adopted in global or regional inverse modeling studies (see, e.g., [[5]](https://acp.copernicus.org/articles/19/14233/2019/), [[6]](https://doi.org/10.5194/acp-19-12067-2019), [[11]](https://confluence.ecmwf.int/display/CKB/Description+of+the+CO2+inversion+production+chain)): indeed, this data product has primarily been generated for comparison with climate models and climate monitoring, rather than for the quantification of carbon fluxes through the ingestion-inversion system.\n\n* **MTCO2_OBS4MIPS Level 3** : this dataset is new and thus never used before. However, Level 2 MT-CO2 and MT-CH4 products have been used for flux inversions ([[18]](https://acp.copernicus.org/articles/14/577/2014/), [[19]](https://acp.copernicus.org/articles/24/10639/2024/)) and the current spatial (1° x 1°) and temporal (1 day) resolution of this Level 3 data product can potentially meet the requirements for the investigation of carbon fluxes at regional and global scales (see, e.g. [[11]](https://confluence.ecmwf.int/display/CKB/Description+of+the+CO2+inversion+production+chain)).", "text_with_prefix": "EQC Quality Assessment: \"Spatial resolution and completeness of carbon dioxide satellite observations for quantifying carbon fluxes\"\nDataset: satellite-carbon-dioxide [CDS]\nAspect: resolution_q03 | Category: Satellite_ECVs\nSection: Spatial resolution and completeness of carbon dioxide satellite observations for quantifying carbon fluxes\n---\ninversion results ([[17]](https://doi.org/10.1029/2005JD006390#jgrd12458-bib-0008)).\n\nThe documented use cases ([[5]](https://acp.copernicus.org/articles/19/14233/2019/), [[6]](https://doi.org/10.5194/acp-19-12067-2019),[[7]](https://essd.copernicus.org/articles/16/2857/2024/), [[8]](https://climate.copernicus.eu/esotc/2023/greenhouse-gas-fluxes)) used OCO-2 observations characterised by spatially dense data with a narrow swath (10.3 km) and with footprints of a few square kilometres (1.29 km × 2.25 km). [[5]](https://acp.copernicus.org/articles/19/14233/2019/) and [[6]](https://doi.org/10.5194/acp-19-12067-2019) also used a previous version of the CDS GOSAT Level 2 deprecated dataset, characterised by a coarser-resolution data (100 km$^2$ at nadir) with lower spatial density than OCO-2. A more recent application is the inversion system that generates the CAMS global CO$_2$ atmospheric inversion product (PyVAR, see [[11]](https://confluence.ecmwf.int/display/CKB/Description+of+the+CO2+inversion+production+chain)): it uses OCO-2 bias-corrected land retrievals of XCO$_2$ after averaging glint and nadir OCO-2 retrievals in 10-s bins (corresponding to a ground-track swaths of 67 km) to match the transport model horizontal spatial resolution of about 90 km$^{2}$.\n\n* **XCO2_OBS4MIPS Level 3**: the spatial aggregation at 5° × 5° resolution and the temporal aggregation at monthly scales inevitably reduce the amount of information available to the inversion system. Such levels of aggregation are not commonly adopted in global or regional inverse modeling studies (see, e.g., [[5]](https://acp.copernicus.org/articles/19/14233/2019/), [[6]](https://doi.org/10.5194/acp-19-12067-2019), [[11]](https://confluence.ecmwf.int/display/CKB/Description+of+the+CO2+inversion+production+chain)): indeed, this data product has primarily been generated for comparison with climate models and climate monitoring, rather than for the quantification of carbon fluxes through the ingestion-inversion system.\n\n* **MTCO2_OBS4MIPS Level 3** : this dataset is new and thus never used before. However, Level 2 MT-CO2 and MT-CH4 products have been used for flux inversions ([[18]](https://acp.copernicus.org/articles/14/577/2014/), [[19]](https://acp.copernicus.org/articles/24/10639/2024/)) and the current spatial (1° x 1°) and temporal (1 day) resolution of this Level 3 data product can potentially meet the requirements for the investigation of carbon fluxes at regional and global scales (see, e.g. [[11]](https://confluence.ecmwf.int/display/CKB/Description+of+the+CO2+inversion+production+chain))."} {"chunk_id": "satellite_satellite-carbon-dioxide_resolution_q03__c0d66ebb7630", "report_id": "satellite_satellite-carbon-dioxide_resolution_q03", "dataset_id": "satellite-carbon-dioxide", "store": "CDS", "doc_type": "EQC_QA", "aspect": "resolution_q03", "aspect_base": "resolution", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Spatial resolution and completeness of carbon dioxide satellite observations for quantifying carbon fluxes", "title": "Spatial resolution and completeness of carbon dioxide satellite observations for quantifying carbon fluxes", "chunk_index": 5, "token_count": 846, "text_raw": "e.g. [[11]](https://confluence.ecmwf.int/display/CKB/Description+of+the+CO2+inversion+production+chain)).\n\n* **Level 2 (deprecated)**: GOSAT and GOSAT-2 measurements (CO2_GOS_SRFP v2.3.8, CO2_GOS_OCFP v7.3, CO2_GO2_SRFP v2.1.0) are characterised by a 9.7 km (diameter) footprint with the single observations typically on the order of 100 km apart with a 6-day repeat cycle. [[5]](https://acp.copernicus.org/articles/19/14233/2019/) and [[6]](https://doi.org/10.5194/acp-19-12067-2019) used previous versions of the GOSAT Level 2 deprecated dataset. According to [[5]](https://acp.copernicus.org/articles/19/14233/2019/), the lower data density of GOSAT observations with respect to OCO-2 may potentially affect the quantification of carbon fluxes but no conclusive statements were made on this topic. For use with an inverse modelling system, interested users may also consider the deprecated Level 2 merged multi-sensor XCO2_EMMA data product that consists of individual Level 2 soundings retrieved by different sensors (including also OCO-2) and algorithms ([[20]](https://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_ATBD_GHG_D_latest.pdf)).\n\n(template:section-4)=\n### 4. Spatial completeness assessment\nThe impact of satellite data on the carbon fluxes diagnosed by the inversion system is affected by to the spatial coverage and by the amount of data provided by the satellite measurements. In general, for regions characterised by low data coverage, the carbon fluxes diagnosed by the inversion system are dominated by the information provided by the \"a-priori\" fields (see [[6]](https://doi.org/10.5194/acp-19-12067-2019)). As documented by [[5]](https://acp.copernicus.org/articles/19/14233/2019/) and [[6]](https://doi.org/10.5194/acp-19-12067-2019), both OCO-2 and GOSAT observations are not evenly distributed spatially.\n\nDue to cloud contamination, there are few retrievals in a large portion of tropical land. When clouds are present, they act as opaque or semi-opaque barriers to the solar radiation reflected at the Earth’s surface, meaning that the radiation reflected from the cloud top does not carry information about the gases below the clouds.\n\nMoreover, available satellite retrievals are very sparse in the northern high-latitude area, especially in boreal regions, due to the occurrence of low solar zenith angles (SZAs). This is because high SZAs, which are typical of regions at high latitudes, complicate retrievals due to weaker illumination. This reduces the satellite radiance signal, making retrievals noisier and more uncertain. Furthermore, the long slant paths under high SZAs amplify the influence of aerosol scattering and cirrus clouds, which can result in errors in the retrieved gas amount. Furthermore, they cause multiple scattering and path-length uncertainty, which can affect satellite radiance observations in the near infrared/short wave infrared (NIR/SWIR) spectral region.\nAs reported by [[21]](https://confluence.ecmwf.int/pages/viewpage.action?pageId=567160577), the specific retrieval difficulties depend on the retrieval algorithm adopted.", "text_with_prefix": "EQC Quality Assessment: \"Spatial resolution and completeness of carbon dioxide satellite observations for quantifying carbon fluxes\"\nDataset: satellite-carbon-dioxide [CDS]\nAspect: resolution_q03 | Category: Satellite_ECVs\nSection: Spatial resolution and completeness of carbon dioxide satellite observations for quantifying carbon fluxes\n---\ne.g. [[11]](https://confluence.ecmwf.int/display/CKB/Description+of+the+CO2+inversion+production+chain)).\n\n* **Level 2 (deprecated)**: GOSAT and GOSAT-2 measurements (CO2_GOS_SRFP v2.3.8, CO2_GOS_OCFP v7.3, CO2_GO2_SRFP v2.1.0) are characterised by a 9.7 km (diameter) footprint with the single observations typically on the order of 100 km apart with a 6-day repeat cycle. [[5]](https://acp.copernicus.org/articles/19/14233/2019/) and [[6]](https://doi.org/10.5194/acp-19-12067-2019) used previous versions of the GOSAT Level 2 deprecated dataset. According to [[5]](https://acp.copernicus.org/articles/19/14233/2019/), the lower data density of GOSAT observations with respect to OCO-2 may potentially affect the quantification of carbon fluxes but no conclusive statements were made on this topic. For use with an inverse modelling system, interested users may also consider the deprecated Level 2 merged multi-sensor XCO2_EMMA data product that consists of individual Level 2 soundings retrieved by different sensors (including also OCO-2) and algorithms ([[20]](https://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_ATBD_GHG_D_latest.pdf)).\n\n(template:section-4)=\n### 4. Spatial completeness assessment\nThe impact of satellite data on the carbon fluxes diagnosed by the inversion system is affected by to the spatial coverage and by the amount of data provided by the satellite measurements. In general, for regions characterised by low data coverage, the carbon fluxes diagnosed by the inversion system are dominated by the information provided by the \"a-priori\" fields (see [[6]](https://doi.org/10.5194/acp-19-12067-2019)). As documented by [[5]](https://acp.copernicus.org/articles/19/14233/2019/) and [[6]](https://doi.org/10.5194/acp-19-12067-2019), both OCO-2 and GOSAT observations are not evenly distributed spatially.\n\nDue to cloud contamination, there are few retrievals in a large portion of tropical land. When clouds are present, they act as opaque or semi-opaque barriers to the solar radiation reflected at the Earth’s surface, meaning that the radiation reflected from the cloud top does not carry information about the gases below the clouds.\n\nMoreover, available satellite retrievals are very sparse in the northern high-latitude area, especially in boreal regions, due to the occurrence of low solar zenith angles (SZAs). This is because high SZAs, which are typical of regions at high latitudes, complicate retrievals due to weaker illumination. This reduces the satellite radiance signal, making retrievals noisier and more uncertain. Furthermore, the long slant paths under high SZAs amplify the influence of aerosol scattering and cirrus clouds, which can result in errors in the retrieved gas amount. Furthermore, they cause multiple scattering and path-length uncertainty, which can affect satellite radiance observations in the near infrared/short wave infrared (NIR/SWIR) spectral region.\nAs reported by [[21]](https://confluence.ecmwf.int/pages/viewpage.action?pageId=567160577), the specific retrieval difficulties depend on the retrieval algorithm adopted."} {"chunk_id": "satellite_satellite-carbon-dioxide_resolution_q03__b3bdaa40a8a0", "report_id": "satellite_satellite-carbon-dioxide_resolution_q03", "dataset_id": "satellite-carbon-dioxide", "store": "CDS", "doc_type": "EQC_QA", "aspect": "resolution_q03", "aspect_base": "resolution", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Spatial resolution and completeness of carbon dioxide satellite observations for quantifying carbon fluxes", "title": "Spatial resolution and completeness of carbon dioxide satellite observations for quantifying carbon fluxes", "chunk_index": 6, "token_count": 1041, "text_raw": "As reported by [[21]](https://confluence.ecmwf.int/pages/viewpage.action?pageId=567160577), the specific retrieval difficulties depend on the retrieval algorithm adopted.\n\n* **XCO2_OBS4MIPS Level 3**: this is a global dataset, obtained by gridding the merged Level 2 products generated with the ensemble median algorithm (XCO2_EMMA, see [[21]](https://confluence.ecmwf.int/pages/viewpage.action?pageId=567160577)). EMMA combines several different XCO$_2$ level-2 satellite data products: SCIAMACHY/Envisat (2003–2012), TANSO-FTS/GOSAT (2009–2022), OCO-2 (2014-2022), TANSO-FTS-2/GOSAT2 (2019-2023). The completeness of the spatial data varies with the seasons due to the above-described limitations in retrieving XCO$_2$ (i.e. low reflected sunlight during periods characterised by low SZAs, and surface shielding due to clouds). Data for the Southern Hemisphere (SH) mid and high latitudes is available from September to March, whereas data for the Northern Hemisphere (NH) mid and high latitudes is mostly available from April to August. Furthermore, the number of available observations depends significantly on location, with low coverage over areas typically affected by high cloud (such as the tropics).\n\n```text\n100%|██████████| 1/1 [00:00<00:00, 1.95it/s]\n```\n\n*The figure shows the monthly fractional XCO$_2$ data coverage from 2003 to 2023 derived from the XCO2_OBS4MIPS dataset (version 4.6). Please note that the colour scale has been set to optimise visualisation of the spatial path of the data coverage.*\n\n* **MTCO2_OBS4MIPS Level 3**: as described in the Documentation [[22]](https://confluence.ecmwf.int/pages/viewpage.action?pageId=567160640), the dataset is mostly limited to tropical air masses (between about 30°N and 30°S) due to the fact that decorrelation between CO$_2$ and temperature signals in the IASI radiances is more complex outside of the tropical region, yielding too high retrieval uncertainty. Moreover, it should be stressed that this is a mid‑troposphere product with the highest sensitivity of the measurements around 250 hPa (see [[28]](https://confluence.ecmwf.int/pages/viewpage.action?pageId=567160673)), while surface CO$_2$ fluxes imprint their strongest signals in the boundary layer (0–2 km). This implies potential limitations in the ability of this dataset in constraining surface CO$_2$ fluxes.\n\n```text\n100%|██████████| 210/210 [01:31<00:00, 2.30it/s]\n```\n\n*The figure shows the monthly fractional XCO$_2$ data coverage from 2007 to 2024 derived from the IASI-MERGED-OBS4MIPS (version 10.1). Please note that the colour scale has been set to optimise visualisation of the spatial path of the data coverage.*\n\n* **Level 2 (deprecated)**: with respect to spatial completeness GOSAT data products are characterised by low data coverage at high latitudes (due to unfavourable illumination conditions) and over the tropics (due to frequent cloud contamination). For use with an inverse modelling system to obtain information on CO$_2$ surface fluxes, interested users may also consider the Level 2 merged multi-sensor XCO2_EMMA data product ([[20]](https://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_ATBD_GHG_D_latest.pdf)) that optimises the spatial and the temporal data coverage by merging XCO$_2$ data from different sensors and algorithms. The XCO2_EMMA data product consists of individual Level 2 soundings retrieved by algorithms that may vary from gridbox to gridbox and from month to month. As recommended by [[23]](https://amt.copernicus.org/articles/13/789/2020/), it is important to note that these merged products are not necessarily the most optimal products for all applications, as they do not include all data from a given satellite sensor.\n\n![Figure_GOS_XCO2.png](attachment:f1d30092-e175-41f4-b8c8-8e1a5b5e895d.png)", "text_with_prefix": "EQC Quality Assessment: \"Spatial resolution and completeness of carbon dioxide satellite observations for quantifying carbon fluxes\"\nDataset: satellite-carbon-dioxide [CDS]\nAspect: resolution_q03 | Category: Satellite_ECVs\nSection: Spatial resolution and completeness of carbon dioxide satellite observations for quantifying carbon fluxes\n---\nAs reported by [[21]](https://confluence.ecmwf.int/pages/viewpage.action?pageId=567160577), the specific retrieval difficulties depend on the retrieval algorithm adopted.\n\n* **XCO2_OBS4MIPS Level 3**: this is a global dataset, obtained by gridding the merged Level 2 products generated with the ensemble median algorithm (XCO2_EMMA, see [[21]](https://confluence.ecmwf.int/pages/viewpage.action?pageId=567160577)). EMMA combines several different XCO$_2$ level-2 satellite data products: SCIAMACHY/Envisat (2003–2012), TANSO-FTS/GOSAT (2009–2022), OCO-2 (2014-2022), TANSO-FTS-2/GOSAT2 (2019-2023). The completeness of the spatial data varies with the seasons due to the above-described limitations in retrieving XCO$_2$ (i.e. low reflected sunlight during periods characterised by low SZAs, and surface shielding due to clouds). Data for the Southern Hemisphere (SH) mid and high latitudes is available from September to March, whereas data for the Northern Hemisphere (NH) mid and high latitudes is mostly available from April to August. Furthermore, the number of available observations depends significantly on location, with low coverage over areas typically affected by high cloud (such as the tropics).\n\n```text\n100%|██████████| 1/1 [00:00<00:00, 1.95it/s]\n```\n\n*The figure shows the monthly fractional XCO$_2$ data coverage from 2003 to 2023 derived from the XCO2_OBS4MIPS dataset (version 4.6). Please note that the colour scale has been set to optimise visualisation of the spatial path of the data coverage.*\n\n* **MTCO2_OBS4MIPS Level 3**: as described in the Documentation [[22]](https://confluence.ecmwf.int/pages/viewpage.action?pageId=567160640), the dataset is mostly limited to tropical air masses (between about 30°N and 30°S) due to the fact that decorrelation between CO$_2$ and temperature signals in the IASI radiances is more complex outside of the tropical region, yielding too high retrieval uncertainty. Moreover, it should be stressed that this is a mid‑troposphere product with the highest sensitivity of the measurements around 250 hPa (see [[28]](https://confluence.ecmwf.int/pages/viewpage.action?pageId=567160673)), while surface CO$_2$ fluxes imprint their strongest signals in the boundary layer (0–2 km). This implies potential limitations in the ability of this dataset in constraining surface CO$_2$ fluxes.\n\n```text\n100%|██████████| 210/210 [01:31<00:00, 2.30it/s]\n```\n\n*The figure shows the monthly fractional XCO$_2$ data coverage from 2007 to 2024 derived from the IASI-MERGED-OBS4MIPS (version 10.1). Please note that the colour scale has been set to optimise visualisation of the spatial path of the data coverage.*\n\n* **Level 2 (deprecated)**: with respect to spatial completeness GOSAT data products are characterised by low data coverage at high latitudes (due to unfavourable illumination conditions) and over the tropics (due to frequent cloud contamination). For use with an inverse modelling system to obtain information on CO$_2$ surface fluxes, interested users may also consider the Level 2 merged multi-sensor XCO2_EMMA data product ([[20]](https://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_ATBD_GHG_D_latest.pdf)) that optimises the spatial and the temporal data coverage by merging XCO$_2$ data from different sensors and algorithms. The XCO2_EMMA data product consists of individual Level 2 soundings retrieved by algorithms that may vary from gridbox to gridbox and from month to month. As recommended by [[23]](https://amt.copernicus.org/articles/13/789/2020/), it is important to note that these merged products are not necessarily the most optimal products for all applications, as they do not include all data from a given satellite sensor.\n\n![Figure_GOS_XCO2.png](attachment:f1d30092-e175-41f4-b8c8-8e1a5b5e895d.png)"} {"chunk_id": "satellite_satellite-carbon-dioxide_resolution_q03__a8dfb01fa53d", "report_id": "satellite_satellite-carbon-dioxide_resolution_q03", "dataset_id": "satellite-carbon-dioxide", "store": "CDS", "doc_type": "EQC_QA", "aspect": "resolution_q03", "aspect_base": "resolution", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Spatial resolution and completeness of carbon dioxide satellite observations for quantifying carbon fluxes", "title": "Spatial resolution and completeness of carbon dioxide satellite observations for quantifying carbon fluxes", "chunk_index": 7, "token_count": 974, "text_raw": "all data from a given satellite sensor.\n\n![Figure_GOS_XCO2.png](attachment:f1d30092-e175-41f4-b8c8-8e1a5b5e895d.png)\n\n*Figure showing global seasonal maps of GOSAT XCO2 \"CO2_GOS_OCFP\" retrieved between December 2021 and November 2022. XCO$_2$ values are expresssed as parts per million (ppm). Image adapted from [[24]](http://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PUGS_GHG_A_latest.pdf) under CC-BY license.*\n\n![Figure_GO2_XCO2.png](attachment:b763ef8a-3c34-4635-90a8-accdd2de12a6.png)\n\n*Figure showing global XCO$_2$ map of GOSAT XCO2 \"CO2_GO2_SRFP\" for the 2019-2022 period on a 1° X 1° resolution for both land (top) and ocean (bottom) retrievals. XCO$_2$ values are expresssed as parts per million (ppm). Image adapted from [[25]](http://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PUGS_GHG_B_latest.pdf) under CC-BY license.*\n\n(template:section-5)=\n### 5. How suitable are these datasets for ingestion into inversion systems?\nIn this section, we summarise the features of the \"Carbon dioxide data from 2002 to present derived from satellite observations\" dataset ([13](https://cds.climate.copernicus.eu/datasets/satellite-carbon-dioxide?tab=overview)) in terms of spatial resolution and completeness, in order to constrain the \"top-down\" inversion systems of global and regional Earth's carbon fluxes. Please note that Level 3 XCO$_2$ or MT-CO$_2$ satellite data products have not yet been used in inversion systems to derive surface CO$_2$ flux quantifications. Thus, the outcomes of this assessment, especially for the MT-CO2_OBS4MIPS data product, may change in the future if applications become available.\n\n* **XCO2_OBS4MIPS Level 3**: This data product is characterised by temporal (monthly) and spatial (5° x 5°) resolutions higher that those typically used to quantify surface CO$_2$ fluxes by inversion systems. Also considering that no applications currently exist for using this XCO$_2$ data product, it cannot be recommended for quantifying surface CO$_2$ fluxes by inversion modelling systems.\n\n* **MTCO2_OBS4MIPS Level 3**: This data product is characterised by temporal (daily) and spatial (1° x 1°) resolutions consistent with those reported in the scientific literature and use cases for use in an inversion system. The data product is primarily limited to tropical air masses (between approximately 30°N and 30°S), with the highest measurement sensitivity around 250 hPa. This implies potential limitations for this data product to constrain surface CO$_2$ fluxes. As there are currently no applications for using this MT-CO$_2$ data product, extreme caution should be exercised when using it to quantify surface CO$_2$ fluxes through inversion modelling systems.\n\n* **Level 2 (deprecated)**: these XCO$_2$ data products are suitable to be ingested in inversion modelling systems for the quantification of historical (2002 - 2022) surface CO$_2$ fluxes. In particular, users may consider the XCO2_EMMA data product that optimises the spatial and the temporal data coverage.\n\nPlease note that this assessment was not produced to rank the distinctive performance of the OCO-2 and GOSAT datasets in atmospheric inversion systems. Differences in the results of inversion experiments can be related to different factors, some of which are related to the satellite observations used (such as their data coverage, data precision, data accuracy including the implemented bias correction), while others can be related to other components of the inversion system (e.g. the transport model used, the \"a-priori\" field, the inversion method used).", "text_with_prefix": "EQC Quality Assessment: \"Spatial resolution and completeness of carbon dioxide satellite observations for quantifying carbon fluxes\"\nDataset: satellite-carbon-dioxide [CDS]\nAspect: resolution_q03 | Category: Satellite_ECVs\nSection: Spatial resolution and completeness of carbon dioxide satellite observations for quantifying carbon fluxes\n---\nall data from a given satellite sensor.\n\n![Figure_GOS_XCO2.png](attachment:f1d30092-e175-41f4-b8c8-8e1a5b5e895d.png)\n\n*Figure showing global seasonal maps of GOSAT XCO2 \"CO2_GOS_OCFP\" retrieved between December 2021 and November 2022. XCO$_2$ values are expresssed as parts per million (ppm). Image adapted from [[24]](http://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PUGS_GHG_A_latest.pdf) under CC-BY license.*\n\n![Figure_GO2_XCO2.png](attachment:b763ef8a-3c34-4635-90a8-accdd2de12a6.png)\n\n*Figure showing global XCO$_2$ map of GOSAT XCO2 \"CO2_GO2_SRFP\" for the 2019-2022 period on a 1° X 1° resolution for both land (top) and ocean (bottom) retrievals. XCO$_2$ values are expresssed as parts per million (ppm). Image adapted from [[25]](http://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PUGS_GHG_B_latest.pdf) under CC-BY license.*\n\n(template:section-5)=\n### 5. How suitable are these datasets for ingestion into inversion systems?\nIn this section, we summarise the features of the \"Carbon dioxide data from 2002 to present derived from satellite observations\" dataset ([13](https://cds.climate.copernicus.eu/datasets/satellite-carbon-dioxide?tab=overview)) in terms of spatial resolution and completeness, in order to constrain the \"top-down\" inversion systems of global and regional Earth's carbon fluxes. Please note that Level 3 XCO$_2$ or MT-CO$_2$ satellite data products have not yet been used in inversion systems to derive surface CO$_2$ flux quantifications. Thus, the outcomes of this assessment, especially for the MT-CO2_OBS4MIPS data product, may change in the future if applications become available.\n\n* **XCO2_OBS4MIPS Level 3**: This data product is characterised by temporal (monthly) and spatial (5° x 5°) resolutions higher that those typically used to quantify surface CO$_2$ fluxes by inversion systems. Also considering that no applications currently exist for using this XCO$_2$ data product, it cannot be recommended for quantifying surface CO$_2$ fluxes by inversion modelling systems.\n\n* **MTCO2_OBS4MIPS Level 3**: This data product is characterised by temporal (daily) and spatial (1° x 1°) resolutions consistent with those reported in the scientific literature and use cases for use in an inversion system. The data product is primarily limited to tropical air masses (between approximately 30°N and 30°S), with the highest measurement sensitivity around 250 hPa. This implies potential limitations for this data product to constrain surface CO$_2$ fluxes. As there are currently no applications for using this MT-CO$_2$ data product, extreme caution should be exercised when using it to quantify surface CO$_2$ fluxes through inversion modelling systems.\n\n* **Level 2 (deprecated)**: these XCO$_2$ data products are suitable to be ingested in inversion modelling systems for the quantification of historical (2002 - 2022) surface CO$_2$ fluxes. In particular, users may consider the XCO2_EMMA data product that optimises the spatial and the temporal data coverage.\n\nPlease note that this assessment was not produced to rank the distinctive performance of the OCO-2 and GOSAT datasets in atmospheric inversion systems. Differences in the results of inversion experiments can be related to different factors, some of which are related to the satellite observations used (such as their data coverage, data precision, data accuracy including the implemented bias correction), while others can be related to other components of the inversion system (e.g. the transport model used, the \"a-priori\" field, the inversion method used)."} {"chunk_id": "satellite_satellite-carbon-dioxide_resolution_q03__1a7592b9f796", "report_id": "satellite_satellite-carbon-dioxide_resolution_q03", "dataset_id": "satellite-carbon-dioxide", "store": "CDS", "doc_type": "EQC_QA", "aspect": "resolution_q03", "aspect_base": "resolution", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Spatial resolution and completeness of carbon dioxide satellite observations for quantifying carbon fluxes > ℹ️ If you want to know more > Key resources", "title": "Spatial resolution and completeness of carbon dioxide satellite observations for quantifying carbon fluxes", "chunk_index": 8, "token_count": 197, "text_raw": "The CDS catalogue entries for the data used were:\n* Carbon dioxide data from 2002 to present derived from satellite observations: https://cds.climate.copernicus.eu/datasets/satellite-carbon-dioxide?tab=overview\n\nUsers interested in accessing CO$_2$ fluxes from inversion systems can consider to use specifically designed Copernicus resources like [CAMS global inversion-optimised greenhouse gas fluxes and concentrations](https://ads.atmosphere.copernicus.eu/datasets/cams-global-greenhouse-gas-inversion?tab=overview).", "text_with_prefix": "EQC Quality Assessment: \"Spatial resolution and completeness of carbon dioxide satellite observations for quantifying carbon fluxes\"\nDataset: satellite-carbon-dioxide [CDS]\nAspect: resolution_q03 | Category: Satellite_ECVs\nSection: Spatial resolution and completeness of carbon dioxide satellite observations for quantifying carbon fluxes > ℹ️ If you want to know more > Key resources\n---\nThe CDS catalogue entries for the data used were:\n* Carbon dioxide data from 2002 to present derived from satellite observations: https://cds.climate.copernicus.eu/datasets/satellite-carbon-dioxide?tab=overview\n\nUsers interested in accessing CO$_2$ fluxes from inversion systems can consider to use specifically designed Copernicus resources like [CAMS global inversion-optimised greenhouse gas fluxes and concentrations](https://ads.atmosphere.copernicus.eu/datasets/cams-global-greenhouse-gas-inversion?tab=overview)."} {"chunk_id": "satellite_satellite-carbon-dioxide_resolution_q03__be6f945b9bc3", "report_id": "satellite_satellite-carbon-dioxide_resolution_q03", "dataset_id": "satellite-carbon-dioxide", "store": "CDS", "doc_type": "EQC_QA", "aspect": "resolution_q03", "aspect_base": "resolution", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Spatial resolution and completeness of carbon dioxide satellite observations for quantifying carbon fluxes > ℹ️ If you want to know more > References", "title": "Spatial resolution and completeness of carbon dioxide satellite observations for quantifying carbon fluxes", "chunk_index": 9, "token_count": 901, "text_raw": "[[1]](https://library.wmo.int/viewer/68532/?offset=#page=4&viewer=picture&o=bookmarks&n=0&q=) World Meteorological Organization. (2023). WMO Greenhouse Gas Bulletin, 19, ISSN 2078-0796.\n\n[[2]](https://essd.copernicus.org/articles/17/965/2025/) Friedlingstein, P., O'Sullivan, M., Jones, M. W., Andrew, R. M., Hauck, J., Landschützer, P., Le Quéré, C., Li, H., Luijkx, I. T., Olsen, A., Peters, G. P., Peters, W., Pongratz, J., Schwingshackl, C., Sitch, S., Canadell, J. G., Ciais, P., Jackson, R. B., Alin, S. R., Arneth, A., Arora, V., Bates, N. R., Becker, M., Bellouin, N., Berghoff, C. F., Bittig, H. C., Bopp, L., Cadule, P., Campbell, K., Chamberlain, M. A., Chandra, N., Chevallier, F., Chini, L. P., Colligan, T., Decayeux, J., Djeutchouang, L. M., Dou, X., Duran Rojas, C., Enyo, K., Evans, W., Fay, A. R., Feely, R. A., Ford, D. J., Foster, A., Gasser, T., Gehlen, M., Gkritzalis, T., Grassi, G., Gregor, L., Gruber, N., Gürses, Ö., Harris, I., Hefner, M., Heinke, J., Hurtt, G. C., Iida, Y., Ilyina, T., Jacobson, A. R., Jain, A. K., Jarníková, T., Jersild, A., Jiang, F., Jin, Z., Kato, E., Keeling, R. F., Klein Goldewijk, K., Knauer, J., Korsbakken, J. I., Lan, X., Lauvset, S. K., Lefèvre, N., Liu, Z., Liu, J., Ma, L., Maksyutov, S., Marland, G., Mayot, N., McGuire, P. C., Metzl, N., Monacci, N. M., Morgan, E. J., Nakaoka, S.-I., Neill, C., Niwa, Y., Nützel, T., Olivier, L., Ono, T., Palmer, P. I., Pierrot, D., Qin, Z., Resplandy, L., Roobaert, A., Rosan, T. M., Rödenbeck, C., Schwinger, J., Smallman, T. L., Smith, S. M., Sospedra-Alfonso, R., Steinhoff, T., Sun, Q., Sutton, A. J., Séférian, R., Takao, S., Tatebe, H., Tian, H., Tilbrook, B., Torres, O., Tourigny, E., Tsujino, H., Tubiello, F., van der Werf, G., Wanninkhof, R., Wang, X., Yang, D., Yang, X., Yu, Z., Yuan, W., Yue, X., Zaehle, S., Zeng, N., and Zeng, J. (2025). Global Carbon Budget 2024, Earth System Science Data, 17, 965-1039. https://doi.org/10.5194/essd-17-965-2025", "text_with_prefix": "EQC Quality Assessment: \"Spatial resolution and completeness of carbon dioxide satellite observations for quantifying carbon fluxes\"\nDataset: satellite-carbon-dioxide [CDS]\nAspect: resolution_q03 | Category: Satellite_ECVs\nSection: Spatial resolution and completeness of carbon dioxide satellite observations for quantifying carbon fluxes > ℹ️ If you want to know more > References\n---\n[[1]](https://library.wmo.int/viewer/68532/?offset=#page=4&viewer=picture&o=bookmarks&n=0&q=) World Meteorological Organization. (2023). WMO Greenhouse Gas Bulletin, 19, ISSN 2078-0796.\n\n[[2]](https://essd.copernicus.org/articles/17/965/2025/) Friedlingstein, P., O'Sullivan, M., Jones, M. W., Andrew, R. M., Hauck, J., Landschützer, P., Le Quéré, C., Li, H., Luijkx, I. T., Olsen, A., Peters, G. P., Peters, W., Pongratz, J., Schwingshackl, C., Sitch, S., Canadell, J. G., Ciais, P., Jackson, R. B., Alin, S. R., Arneth, A., Arora, V., Bates, N. R., Becker, M., Bellouin, N., Berghoff, C. F., Bittig, H. C., Bopp, L., Cadule, P., Campbell, K., Chamberlain, M. A., Chandra, N., Chevallier, F., Chini, L. P., Colligan, T., Decayeux, J., Djeutchouang, L. M., Dou, X., Duran Rojas, C., Enyo, K., Evans, W., Fay, A. R., Feely, R. A., Ford, D. J., Foster, A., Gasser, T., Gehlen, M., Gkritzalis, T., Grassi, G., Gregor, L., Gruber, N., Gürses, Ö., Harris, I., Hefner, M., Heinke, J., Hurtt, G. C., Iida, Y., Ilyina, T., Jacobson, A. R., Jain, A. K., Jarníková, T., Jersild, A., Jiang, F., Jin, Z., Kato, E., Keeling, R. F., Klein Goldewijk, K., Knauer, J., Korsbakken, J. I., Lan, X., Lauvset, S. K., Lefèvre, N., Liu, Z., Liu, J., Ma, L., Maksyutov, S., Marland, G., Mayot, N., McGuire, P. C., Metzl, N., Monacci, N. M., Morgan, E. J., Nakaoka, S.-I., Neill, C., Niwa, Y., Nützel, T., Olivier, L., Ono, T., Palmer, P. I., Pierrot, D., Qin, Z., Resplandy, L., Roobaert, A., Rosan, T. M., Rödenbeck, C., Schwinger, J., Smallman, T. L., Smith, S. M., Sospedra-Alfonso, R., Steinhoff, T., Sun, Q., Sutton, A. J., Séférian, R., Takao, S., Tatebe, H., Tian, H., Tilbrook, B., Torres, O., Tourigny, E., Tsujino, H., Tubiello, F., van der Werf, G., Wanninkhof, R., Wang, X., Yang, D., Yang, X., Yu, Z., Yuan, W., Yue, X., Zaehle, S., Zeng, N., and Zeng, J. (2025). Global Carbon Budget 2024, Earth System Science Data, 17, 965-1039. https://doi.org/10.5194/essd-17-965-2025"} {"chunk_id": "satellite_satellite-carbon-dioxide_resolution_q03__95f6d02fb1b7", "report_id": "satellite_satellite-carbon-dioxide_resolution_q03", "dataset_id": "satellite-carbon-dioxide", "store": "CDS", "doc_type": "EQC_QA", "aspect": "resolution_q03", "aspect_base": "resolution", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Spatial resolution and completeness of carbon dioxide satellite observations for quantifying carbon fluxes > ℹ️ If you want to know more > References", "title": "Spatial resolution and completeness of carbon dioxide satellite observations for quantifying carbon fluxes", "chunk_index": 10, "token_count": 995, "text_raw": "). Global Carbon Budget 2024, Earth System Science Data, 17, 965-1039. https://doi.org/10.5194/essd-17-965-2025\n\n[[3]](https://doi.org/10.5194/acp-14-13739-2014) Reuter, M., Buchwitz, M., Hilker, M., Heymann, J., Schneising, O., Pillai, D., Bovensmann, H., Burrows, J. P., Bösch, H., Parker, R., Butz, A., Hasekamp, O., O'Dell, C. W., Yoshida, Y., Gerbig, C., Nehrkorn, T., Deutscher, N. M., Warneke, T., Notholt, J., Hase, F., Kivi, R., Sussmann, R., Machida, T., Matsueda, H., and Sawa, Y. (2013). Satellite-inferred European carbon sink larger than expected. Atmospheric Chemistry and Physics, 14, 13739–13753. https://doi.org/10.5194/acp-14-13739-2014\n\n[[4]](https://doi.org/10.1175/BAMS-D-15-00310.1) Reuter, M., and Coauthors. (2017). How much CO$_2$ is taken up by the European terrestrial biosphere?. Bullettin of American Meteorology Society, 98, 665–671. https://doi.org/10.1175/BAMS-D-15-00310.1\n\n[[5]](https://acp.copernicus.org/articles/19/14233/2019/) Chevallier, F., Remaud, M., O'Dell, C. W., Baker, D., Peylin, P., and Cozic, A. (2019). Objective evaluation of surface- and satellite-driven carbon dioxide atmospheric inversions. Atmospheric Chemistry and Physics, 19, 14233–14251. https://doi.org/10.5194/acp-19-14233-2019\n\n[[6]](https://doi.org/10.5194/acp-19-12067-2019) Wang, H., Jiang, F., Wang, J., Ju, W., and Chen, J. M. (2019). Terrestrial ecosystem carbon flux estimated using GOSAT and OCO-2 XCO2 retrievals. Atmospheric Chemistry and Physics, 19, 12067–12082. https://doi.org/10.5194/acp-19-12067-2019\n\n[[7]](https://essd.copernicus.org/articles/16/2857/2024/) Jin, Z., Tian, X., Wang, Y., Zhang, H., Zhao, M., Wang, T., Ding, J., and Piao, S. (2024). A global surface CO$_2$ flux dataset (2015–2022) inferred from OCO-2 retrievals using the GONGGA inversion system. Earth System Science Data, 16, 2857–2876. https://doi.org/10.5194/essd-16-2857-2024\n\n[[8]](https://climate.copernicus.eu/esotc/2023/greenhouse-gas-fluxes) Copernicus Climate Change Service (C3S). (2024). European State of the Climate 2023, Web site.\n\n[[9]](https://doi.org/10.5194/bg-10-6699-2013) Peylin, P., Law, R. M., Gurney, K. R., Chevallier, F., Jacobson, A. R., Maki, T., Niwa, Y., Patra, P. K., Peters, W., Rayner, P. J., Rödenbeck, C., van der Laan-Luijkx, I. T., and Zhang, X. (2013). Global atmospheric carbon budget: results from an ensemble of atmospheric CO$_2$ inversions. Biogeosciences, 10, 6699–6720. https://doi.org/10.5194/bg-10-6699-2013", "text_with_prefix": "EQC Quality Assessment: \"Spatial resolution and completeness of carbon dioxide satellite observations for quantifying carbon fluxes\"\nDataset: satellite-carbon-dioxide [CDS]\nAspect: resolution_q03 | Category: Satellite_ECVs\nSection: Spatial resolution and completeness of carbon dioxide satellite observations for quantifying carbon fluxes > ℹ️ If you want to know more > References\n---\n). Global Carbon Budget 2024, Earth System Science Data, 17, 965-1039. https://doi.org/10.5194/essd-17-965-2025\n\n[[3]](https://doi.org/10.5194/acp-14-13739-2014) Reuter, M., Buchwitz, M., Hilker, M., Heymann, J., Schneising, O., Pillai, D., Bovensmann, H., Burrows, J. P., Bösch, H., Parker, R., Butz, A., Hasekamp, O., O'Dell, C. W., Yoshida, Y., Gerbig, C., Nehrkorn, T., Deutscher, N. M., Warneke, T., Notholt, J., Hase, F., Kivi, R., Sussmann, R., Machida, T., Matsueda, H., and Sawa, Y. (2013). Satellite-inferred European carbon sink larger than expected. Atmospheric Chemistry and Physics, 14, 13739–13753. https://doi.org/10.5194/acp-14-13739-2014\n\n[[4]](https://doi.org/10.1175/BAMS-D-15-00310.1) Reuter, M., and Coauthors. (2017). How much CO$_2$ is taken up by the European terrestrial biosphere?. Bullettin of American Meteorology Society, 98, 665–671. https://doi.org/10.1175/BAMS-D-15-00310.1\n\n[[5]](https://acp.copernicus.org/articles/19/14233/2019/) Chevallier, F., Remaud, M., O'Dell, C. W., Baker, D., Peylin, P., and Cozic, A. (2019). Objective evaluation of surface- and satellite-driven carbon dioxide atmospheric inversions. Atmospheric Chemistry and Physics, 19, 14233–14251. https://doi.org/10.5194/acp-19-14233-2019\n\n[[6]](https://doi.org/10.5194/acp-19-12067-2019) Wang, H., Jiang, F., Wang, J., Ju, W., and Chen, J. M. (2019). Terrestrial ecosystem carbon flux estimated using GOSAT and OCO-2 XCO2 retrievals. Atmospheric Chemistry and Physics, 19, 12067–12082. https://doi.org/10.5194/acp-19-12067-2019\n\n[[7]](https://essd.copernicus.org/articles/16/2857/2024/) Jin, Z., Tian, X., Wang, Y., Zhang, H., Zhao, M., Wang, T., Ding, J., and Piao, S. (2024). A global surface CO$_2$ flux dataset (2015–2022) inferred from OCO-2 retrievals using the GONGGA inversion system. Earth System Science Data, 16, 2857–2876. https://doi.org/10.5194/essd-16-2857-2024\n\n[[8]](https://climate.copernicus.eu/esotc/2023/greenhouse-gas-fluxes) Copernicus Climate Change Service (C3S). (2024). European State of the Climate 2023, Web site.\n\n[[9]](https://doi.org/10.5194/bg-10-6699-2013) Peylin, P., Law, R. M., Gurney, K. R., Chevallier, F., Jacobson, A. R., Maki, T., Niwa, Y., Patra, P. K., Peters, W., Rayner, P. J., Rödenbeck, C., van der Laan-Luijkx, I. T., and Zhang, X. (2013). Global atmospheric carbon budget: results from an ensemble of atmospheric CO$_2$ inversions. Biogeosciences, 10, 6699–6720. https://doi.org/10.5194/bg-10-6699-2013"} {"chunk_id": "satellite_satellite-carbon-dioxide_resolution_q03__b24467d4428a", "report_id": "satellite_satellite-carbon-dioxide_resolution_q03", "dataset_id": "satellite-carbon-dioxide", "store": "CDS", "doc_type": "EQC_QA", "aspect": "resolution_q03", "aspect_base": "resolution", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Spatial resolution and completeness of carbon dioxide satellite observations for quantifying carbon fluxes > ℹ️ If you want to know more > References", "title": "Spatial resolution and completeness of carbon dioxide satellite observations for quantifying carbon fluxes", "chunk_index": 11, "token_count": 982, "text_raw": "ensemble of atmospheric CO$_2$ inversions. Biogeosciences, 10, 6699–6720. https://doi.org/10.5194/bg-10-6699-2013\n\n[[10]](https://acp.copernicus.org/articles/22/9215/2022/) Chandra, N., Patra, P. K., Niwa, Y., Ito, A., Iida, Y., Goto, D., Morimoto, S., Kondo, M., Takigawa, M., Hajima, T., and Watanabe, M. (2022). Estimated regional CO2 flux and uncertainty based on an ensemble of atmospheric CO2 inversions. Atmospheric Chemistry and Physics, 22, 9215–9243. https://doi.org/10.5194/acp-22-9215-2022\n\n[[11]](https://confluence.ecmwf.int/display/CKB/Description+of+the+CO2+inversion+production+chain) European Centre for Medium-Range Weather Forecasts (ECMWF). Description of the CO2 inversion production chain. Version: 12-December-2025 11:25 (Accessed on 7-January-2026).\n\n[[12]](https://doi.org/10.15138/9N0H-ZH07) Lan, X., Tans, P. and K.W., Thoning. Trends in globally-averaged CO$_2$ determined from NOAA Global Monitoring Laboratory measurements. Version: Monday, 05-May-2025 16:38:58 MDT. https://doi.org/10.15138/9N0H-ZH07\n\n[[13]](https://cds.climate.copernicus.eu/datasets/satellite-carbon-dioxide?tab=overview) Copernicus Climate Change Service, Climate Data Store, (2018). Carbon dioxide data from 2002 to present derived from satellite observations. Copernicus Climate Change Service (C3S) Climate Data Store (CDS). https://doi.org/10.24381/cds.f74805c8 (Accessed on 7-January-2026).\n\n[[14]](https://confluence.ecmwf.int/pages/viewpage.action?pageId=567160575) European Centre for Medium-Range Weather Forecasts (ECMWF). C3S Greenhouse Gas (GHG). Version: 5-September-2025 10:55 (Accessed on 7-January-2026).\n\n[[15]](http://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PUGS_GHG_main_latest.pdf) Buchwitz, M. (2024). Product User Guide and Specification (PUGS) – Main document for Greenhouse Gas (GHG: CO$_2$ & CH$_4$) data set CDR7 (01.2003-12.2022), C3S project 2021/C3S2_312a_Lot2_DLR/SC1, v7.3.\n\n[[16]](https://acp.copernicus.org/articles/19/14233/2019/) Chevallier, F., Remaud, M., O'Dell, C. W., Baker, D., Peylin, P., and Cozic, A. (2019). Objective evaluation of surface- and satellite-driven carbon dioxide atmospheric inversions. Atmospheric Chemistry and Physics, 19, 14233–14251, https://doi.org/10.5194/acp-19-14233-2019\n\n[[17]](https://doi.org/10.1029/2005JD006390#jgrd12458-bib-0008) Chevallier, F., Fisher, M. , Peylin, P., Serrar, S., Bousquet, P., Bréon, F.-M., Chédin, A., and Ciais P. (2005). Inferring CO2 sources and sinks from satellite observations: Method and application to TOVS data. Journal of Geophysical Research, 110, D24309. https://doi.org/10.1029/2005JD006390", "text_with_prefix": "EQC Quality Assessment: \"Spatial resolution and completeness of carbon dioxide satellite observations for quantifying carbon fluxes\"\nDataset: satellite-carbon-dioxide [CDS]\nAspect: resolution_q03 | Category: Satellite_ECVs\nSection: Spatial resolution and completeness of carbon dioxide satellite observations for quantifying carbon fluxes > ℹ️ If you want to know more > References\n---\nensemble of atmospheric CO$_2$ inversions. Biogeosciences, 10, 6699–6720. https://doi.org/10.5194/bg-10-6699-2013\n\n[[10]](https://acp.copernicus.org/articles/22/9215/2022/) Chandra, N., Patra, P. K., Niwa, Y., Ito, A., Iida, Y., Goto, D., Morimoto, S., Kondo, M., Takigawa, M., Hajima, T., and Watanabe, M. (2022). Estimated regional CO2 flux and uncertainty based on an ensemble of atmospheric CO2 inversions. Atmospheric Chemistry and Physics, 22, 9215–9243. https://doi.org/10.5194/acp-22-9215-2022\n\n[[11]](https://confluence.ecmwf.int/display/CKB/Description+of+the+CO2+inversion+production+chain) European Centre for Medium-Range Weather Forecasts (ECMWF). Description of the CO2 inversion production chain. Version: 12-December-2025 11:25 (Accessed on 7-January-2026).\n\n[[12]](https://doi.org/10.15138/9N0H-ZH07) Lan, X., Tans, P. and K.W., Thoning. Trends in globally-averaged CO$_2$ determined from NOAA Global Monitoring Laboratory measurements. Version: Monday, 05-May-2025 16:38:58 MDT. https://doi.org/10.15138/9N0H-ZH07\n\n[[13]](https://cds.climate.copernicus.eu/datasets/satellite-carbon-dioxide?tab=overview) Copernicus Climate Change Service, Climate Data Store, (2018). Carbon dioxide data from 2002 to present derived from satellite observations. Copernicus Climate Change Service (C3S) Climate Data Store (CDS). https://doi.org/10.24381/cds.f74805c8 (Accessed on 7-January-2026).\n\n[[14]](https://confluence.ecmwf.int/pages/viewpage.action?pageId=567160575) European Centre for Medium-Range Weather Forecasts (ECMWF). C3S Greenhouse Gas (GHG). Version: 5-September-2025 10:55 (Accessed on 7-January-2026).\n\n[[15]](http://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PUGS_GHG_main_latest.pdf) Buchwitz, M. (2024). Product User Guide and Specification (PUGS) – Main document for Greenhouse Gas (GHG: CO$_2$ & CH$_4$) data set CDR7 (01.2003-12.2022), C3S project 2021/C3S2_312a_Lot2_DLR/SC1, v7.3.\n\n[[16]](https://acp.copernicus.org/articles/19/14233/2019/) Chevallier, F., Remaud, M., O'Dell, C. W., Baker, D., Peylin, P., and Cozic, A. (2019). Objective evaluation of surface- and satellite-driven carbon dioxide atmospheric inversions. Atmospheric Chemistry and Physics, 19, 14233–14251, https://doi.org/10.5194/acp-19-14233-2019\n\n[[17]](https://doi.org/10.1029/2005JD006390#jgrd12458-bib-0008) Chevallier, F., Fisher, M. , Peylin, P., Serrar, S., Bousquet, P., Bréon, F.-M., Chédin, A., and Ciais P. (2005). Inferring CO2 sources and sinks from satellite observations: Method and application to TOVS data. Journal of Geophysical Research, 110, D24309. https://doi.org/10.1029/2005JD006390"} {"chunk_id": "satellite_satellite-carbon-dioxide_resolution_q03__620b58cad42b", "report_id": "satellite_satellite-carbon-dioxide_resolution_q03", "dataset_id": "satellite-carbon-dioxide", "store": "CDS", "doc_type": "EQC_QA", "aspect": "resolution_q03", "aspect_base": "resolution", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Spatial resolution and completeness of carbon dioxide satellite observations for quantifying carbon fluxes > ℹ️ If you want to know more > References", "title": "Spatial resolution and completeness of carbon dioxide satellite observations for quantifying carbon fluxes", "chunk_index": 12, "token_count": 1103, "text_raw": "and sinks from satellite observations: Method and application to TOVS data. Journal of Geophysical Research, 110, D24309. https://doi.org/10.1029/2005JD006390\n\n[[18]](https://acp.copernicus.org/articles/14/577/2014/)Cressot, C., Chevallier, F., Bousquet, P., Crevoisier, C., Dlugokencky, E. J., Fortems-Cheiney, A., Frankenberg, C., Parker, R., Pison, I., Scheepmaker, R. A., Montzka, S. A., Krummel, P. B., Steele, L. P., and Langenfelds, R. L. (2014). On the consistency between global and regional methane emissions inferred from SCIAMACHY, TANSO-FTS, IASI and surface measurements. Atmospheric Chemistry and Physics, 14, 577–592, https://doi.org/10.5194/acp-14-577-2014\n\n[[19]](https://acp.copernicus.org/articles/24/10639/2024/) Wilson, C., Kerridge, B. J., Siddans, R., Moore, D. P., Ventress, L. J., Dowd, E., Feng, W., Chipperfield, M. P., and Remedios, J. J. (2024). Quantifying large methane emissions from the Nord Stream pipeline gas leak of September 2022 using IASI satellite observations and inverse modelling. Atmospheric Chemistry and Physics, 24, 10639–10653, https://doi.org/10.5194/acp-24-10639-2024\n\n[[20]](https://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_ATBD_GHG_D_latest.pdf) Reuter, M.,Fuentes Andrade, B., Buchwitz, M. (2025). C3S Greenhouse Gas (GHG: CO2 & CH4) v4.6: Algorithm Theoretical Basis Document (ATBD), C3S2_313a_DLR_WP1-DDP-GHG-v1_ATBD_XGHG_v4.6.\n\n[[21]](https://confluence.ecmwf.int/pages/viewpage.action?pageId=567160577) European Centre for Medium-Range Weather Forecasts (ECMWF). C3S Greenhouse Gas (GHG: CO2 & CH4) v4.6: Algorithm Theoretical Basis Document (ATBD). Version: 10-September-2025 8:42 (Accessed on 7-January-2026).\n\n[[22]](https://confluence.ecmwf.int/pages/viewpage.action?pageId=567160640) European Centre for Medium-Range Weather Forecasts (ECMWF). C3S Greenhouse Gas (GHG: MTCO2 v10.1 & MTCH4 v10.2): Product User Guide and Specification (PUGS). Version: 8-December-2025 10:49 (Accessed on 7-January-2026).\n\n[[23]](https://amt.copernicus.org/articles/13/789/2020/) Reuter, M., Buchwitz, M., Schneising, O., Noël, S., Bovensmann, H., Burrows, J. P., Boesch, H., Di Noia, A., Anand, J., Parker, R. J., Somkuti, P., Wu, L., Hasekamp, O. P., Aben, I., Kuze, A., Suto, H., Shiomi, K., Yoshida, Y., Morino, I., Crisp, D., O'Dell, C. W., Notholt, J., Petri, C., Warneke, T., Velazco, V. A., Deutscher, N. M., Griffith, D. W. T., Kivi, R., Pollard, D. F., Hase, F., Sussmann, R., Té, Y. V., Strong, K., Roche, S., Sha, M. K., De Mazière, M., Feist, D. G., Iraci, L. T., Roehl, C. M., Retscher, C., and Schepers, D. (2020). Ensemble-based satellite-derived carbon dioxide and methane column-averaged dry-air mole fraction data sets (2003–2018) for carbon and climate applications. Atmospheric Chemistry and Physics, 13, 789–819. https://doi.org/10.5194/amt-13-789-2020, 2020.", "text_with_prefix": "EQC Quality Assessment: \"Spatial resolution and completeness of carbon dioxide satellite observations for quantifying carbon fluxes\"\nDataset: satellite-carbon-dioxide [CDS]\nAspect: resolution_q03 | Category: Satellite_ECVs\nSection: Spatial resolution and completeness of carbon dioxide satellite observations for quantifying carbon fluxes > ℹ️ If you want to know more > References\n---\nand sinks from satellite observations: Method and application to TOVS data. Journal of Geophysical Research, 110, D24309. https://doi.org/10.1029/2005JD006390\n\n[[18]](https://acp.copernicus.org/articles/14/577/2014/)Cressot, C., Chevallier, F., Bousquet, P., Crevoisier, C., Dlugokencky, E. J., Fortems-Cheiney, A., Frankenberg, C., Parker, R., Pison, I., Scheepmaker, R. A., Montzka, S. A., Krummel, P. B., Steele, L. P., and Langenfelds, R. L. (2014). On the consistency between global and regional methane emissions inferred from SCIAMACHY, TANSO-FTS, IASI and surface measurements. Atmospheric Chemistry and Physics, 14, 577–592, https://doi.org/10.5194/acp-14-577-2014\n\n[[19]](https://acp.copernicus.org/articles/24/10639/2024/) Wilson, C., Kerridge, B. J., Siddans, R., Moore, D. P., Ventress, L. J., Dowd, E., Feng, W., Chipperfield, M. P., and Remedios, J. J. (2024). Quantifying large methane emissions from the Nord Stream pipeline gas leak of September 2022 using IASI satellite observations and inverse modelling. Atmospheric Chemistry and Physics, 24, 10639–10653, https://doi.org/10.5194/acp-24-10639-2024\n\n[[20]](https://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_ATBD_GHG_D_latest.pdf) Reuter, M.,Fuentes Andrade, B., Buchwitz, M. (2025). C3S Greenhouse Gas (GHG: CO2 & CH4) v4.6: Algorithm Theoretical Basis Document (ATBD), C3S2_313a_DLR_WP1-DDP-GHG-v1_ATBD_XGHG_v4.6.\n\n[[21]](https://confluence.ecmwf.int/pages/viewpage.action?pageId=567160577) European Centre for Medium-Range Weather Forecasts (ECMWF). C3S Greenhouse Gas (GHG: CO2 & CH4) v4.6: Algorithm Theoretical Basis Document (ATBD). Version: 10-September-2025 8:42 (Accessed on 7-January-2026).\n\n[[22]](https://confluence.ecmwf.int/pages/viewpage.action?pageId=567160640) European Centre for Medium-Range Weather Forecasts (ECMWF). C3S Greenhouse Gas (GHG: MTCO2 v10.1 & MTCH4 v10.2): Product User Guide and Specification (PUGS). Version: 8-December-2025 10:49 (Accessed on 7-January-2026).\n\n[[23]](https://amt.copernicus.org/articles/13/789/2020/) Reuter, M., Buchwitz, M., Schneising, O., Noël, S., Bovensmann, H., Burrows, J. P., Boesch, H., Di Noia, A., Anand, J., Parker, R. J., Somkuti, P., Wu, L., Hasekamp, O. P., Aben, I., Kuze, A., Suto, H., Shiomi, K., Yoshida, Y., Morino, I., Crisp, D., O'Dell, C. W., Notholt, J., Petri, C., Warneke, T., Velazco, V. A., Deutscher, N. M., Griffith, D. W. T., Kivi, R., Pollard, D. F., Hase, F., Sussmann, R., Té, Y. V., Strong, K., Roche, S., Sha, M. K., De Mazière, M., Feist, D. G., Iraci, L. T., Roehl, C. M., Retscher, C., and Schepers, D. (2020). Ensemble-based satellite-derived carbon dioxide and methane column-averaged dry-air mole fraction data sets (2003–2018) for carbon and climate applications. Atmospheric Chemistry and Physics, 13, 789–819. https://doi.org/10.5194/amt-13-789-2020, 2020."} {"chunk_id": "satellite_satellite-carbon-dioxide_resolution_q03__f2a5d4d013ab", "report_id": "satellite_satellite-carbon-dioxide_resolution_q03", "dataset_id": "satellite-carbon-dioxide", "store": "CDS", "doc_type": "EQC_QA", "aspect": "resolution_q03", "aspect_base": "resolution", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Spatial resolution and completeness of carbon dioxide satellite observations for quantifying carbon fluxes > ℹ️ If you want to know more > References", "title": "Spatial resolution and completeness of carbon dioxide satellite observations for quantifying carbon fluxes", "chunk_index": 13, "token_count": 388, "text_raw": "2003–2018) for carbon and climate applications. Atmospheric Chemistry and Physics, 13, 789–819. https://doi.org/10.5194/amt-13-789-2020, 2020.\n\n[[24]](http://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PUGS_GHG_A_latest.pdf) Boesch, A.,and Di Noia, A. (2024). Product User Guide and Specification (PUGS) – ANNEX A for products CO2_GOS_OCFP, CH4_GOS_OCFP (v7.3, 2009-2021) & CH4_GOS_OCPR (v9.0, 2009- 2021), C3S2_312a_Lot2_DLR – Atmosphere.\n\n[[25]](http://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PUGS_GHG_B_latest.pdf) Barr A., and Borsdorff, T. (2024). Product User Guide and Specification (PUGS) – ANNEX B for products CO2_GO2_SRFP, CH4_GO2_SRFP (v2.0.0, 2019-2022), C3S2_312a_Lot2_DLR – Atmosphere.", "text_with_prefix": "EQC Quality Assessment: \"Spatial resolution and completeness of carbon dioxide satellite observations for quantifying carbon fluxes\"\nDataset: satellite-carbon-dioxide [CDS]\nAspect: resolution_q03 | Category: Satellite_ECVs\nSection: Spatial resolution and completeness of carbon dioxide satellite observations for quantifying carbon fluxes > ℹ️ If you want to know more > References\n---\n2003–2018) for carbon and climate applications. Atmospheric Chemistry and Physics, 13, 789–819. https://doi.org/10.5194/amt-13-789-2020, 2020.\n\n[[24]](http://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PUGS_GHG_A_latest.pdf) Boesch, A.,and Di Noia, A. (2024). Product User Guide and Specification (PUGS) – ANNEX A for products CO2_GOS_OCFP, CH4_GOS_OCFP (v7.3, 2009-2021) & CH4_GOS_OCPR (v9.0, 2009- 2021), C3S2_312a_Lot2_DLR – Atmosphere.\n\n[[25]](http://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PUGS_GHG_B_latest.pdf) Barr A., and Borsdorff, T. (2024). Product User Guide and Specification (PUGS) – ANNEX B for products CO2_GO2_SRFP, CH4_GO2_SRFP (v2.0.0, 2019-2022), C3S2_312a_Lot2_DLR – Atmosphere."} {"chunk_id": "satellite_satellite-carbon-dioxide_trend-assessment_q01__a53310b2b694", "report_id": "satellite_satellite-carbon-dioxide_trend-assessment_q01", "dataset_id": "satellite-carbon-dioxide", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q01", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Carbon dioxide satellite observations completeness assessment for greenhouse gas monitoring", "title": "Carbon dioxide satellite observations completeness assessment for greenhouse gas monitoring", "chunk_index": 0, "token_count": 95, "text_raw": "Production date: 03-09-2024\n\nProduced by: Consiglio Nazionale delle Ricerche ([CNR](https://www.cnr.it/en))", "text_with_prefix": "EQC Quality Assessment: \"Carbon dioxide satellite observations completeness assessment for greenhouse gas monitoring\"\nDataset: satellite-carbon-dioxide [CDS]\nAspect: trend-assessment_q01 | Category: Satellite_ECVs\nSection: Carbon dioxide satellite observations completeness assessment for greenhouse gas monitoring\n---\nProduction date: 03-09-2024\n\nProduced by: Consiglio Nazionale delle Ricerche ([CNR](https://www.cnr.it/en))"} {"chunk_id": "satellite_satellite-carbon-dioxide_trend-assessment_q01__ee5317c1e32c", "report_id": "satellite_satellite-carbon-dioxide_trend-assessment_q01", "dataset_id": "satellite-carbon-dioxide", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q01", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Carbon dioxide satellite observations completeness assessment for greenhouse gas monitoring > Quality assessment question", "title": "Carbon dioxide satellite observations completeness assessment for greenhouse gas monitoring", "chunk_index": 1, "token_count": 386, "text_raw": "* **How the variability of spatial and temporal data coverage can affect the quantification of the long-term atmospheric carbon dioxide trends by satellite measurements (XCO2 Level 3 gridded product)?**\n\nCarbon dioxide (CO2) is the most important anthropogenic greenhouse gas, representing about 64% of the total radiative forcing by long living greenhouse gases [[1]](https://library.wmo.int/viewer/68532/?offset=#page=4&viewer=picture&o=bookmarks&n=0&q=): in 2022, atmospheric CO2 represented the 150% of the pre-industrial level. In this assessment, atmospheric CO2 spatial seasonal means and trends are analysed using the XCO2 v4.4 Level 3 gridded product (OBS4MIPS)[[2]](http://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PUGS_GHG_main_latest.pdf), by adopting an approach similar to [[3]](https://amt.copernicus.org/articles/13/789/2020/).\n\nCode is included for transparency but also learning purposes and gives users the chance to adapt the code used for the assesment as they wish. Since this assessment is an experimental product, users are referred to official sources for accurate reporting (e.g., [[4]](https://climate.copernicus.eu/esotc/2023), [[5]](https://climate.copernicus.eu/global-climate-highlights-2023)).", "text_with_prefix": "EQC Quality Assessment: \"Carbon dioxide satellite observations completeness assessment for greenhouse gas monitoring\"\nDataset: satellite-carbon-dioxide [CDS]\nAspect: trend-assessment_q01 | Category: Satellite_ECVs\nSection: Carbon dioxide satellite observations completeness assessment for greenhouse gas monitoring > Quality assessment question\n---\n* **How the variability of spatial and temporal data coverage can affect the quantification of the long-term atmospheric carbon dioxide trends by satellite measurements (XCO2 Level 3 gridded product)?**\n\nCarbon dioxide (CO2) is the most important anthropogenic greenhouse gas, representing about 64% of the total radiative forcing by long living greenhouse gases [[1]](https://library.wmo.int/viewer/68532/?offset=#page=4&viewer=picture&o=bookmarks&n=0&q=): in 2022, atmospheric CO2 represented the 150% of the pre-industrial level. In this assessment, atmospheric CO2 spatial seasonal means and trends are analysed using the XCO2 v4.4 Level 3 gridded product (OBS4MIPS)[[2]](http://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PUGS_GHG_main_latest.pdf), by adopting an approach similar to [[3]](https://amt.copernicus.org/articles/13/789/2020/).\n\nCode is included for transparency but also learning purposes and gives users the chance to adapt the code used for the assesment as they wish. Since this assessment is an experimental product, users are referred to official sources for accurate reporting (e.g., [[4]](https://climate.copernicus.eu/esotc/2023), [[5]](https://climate.copernicus.eu/global-climate-highlights-2023))."} {"chunk_id": "satellite_satellite-carbon-dioxide_trend-assessment_q01__8634bd67dc86", "report_id": "satellite_satellite-carbon-dioxide_trend-assessment_q01", "dataset_id": "satellite-carbon-dioxide", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q01", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Carbon dioxide satellite observations completeness assessment for greenhouse gas monitoring > Quality assessment statements", "title": "Carbon dioxide satellite observations completeness assessment for greenhouse gas monitoring", "chunk_index": 2, "token_count": 227, "text_raw": "These are the key outcomes of this assessment\n\n* The dataset \"Carbon dioxide data from 2002 to present derived from satellite observations\" can be used to evaluate CO2 mean values, climatology and growth rate over the globe, hemispheres or large regions\n* Caution should be exercised in certain regions (high latitudes, regions with frequent cloud cover, oceans) where data availability varies along the temporal coverage of the data set: this must be carefully considered when evaluating global or hemispheric information\n* For data in high latitude regions or in regions with frequent cloud cover, users should consult uncertainty and quality flags as appropriate for their applications. In addition, for the years 2003-2008 only values over land are available, which has to be taken into account for possible applications of this dataset\n```", "text_with_prefix": "EQC Quality Assessment: \"Carbon dioxide satellite observations completeness assessment for greenhouse gas monitoring\"\nDataset: satellite-carbon-dioxide [CDS]\nAspect: trend-assessment_q01 | Category: Satellite_ECVs\nSection: Carbon dioxide satellite observations completeness assessment for greenhouse gas monitoring > Quality assessment statements\n---\nThese are the key outcomes of this assessment\n\n* The dataset \"Carbon dioxide data from 2002 to present derived from satellite observations\" can be used to evaluate CO2 mean values, climatology and growth rate over the globe, hemispheres or large regions\n* Caution should be exercised in certain regions (high latitudes, regions with frequent cloud cover, oceans) where data availability varies along the temporal coverage of the data set: this must be carefully considered when evaluating global or hemispheric information\n* For data in high latitude regions or in regions with frequent cloud cover, users should consult uncertainty and quality flags as appropriate for their applications. In addition, for the years 2003-2008 only values over land are available, which has to be taken into account for possible applications of this dataset\n```"} {"chunk_id": "satellite_satellite-carbon-dioxide_trend-assessment_q01__32abc9fc45ce", "report_id": "satellite_satellite-carbon-dioxide_trend-assessment_q01", "dataset_id": "satellite-carbon-dioxide", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q01", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Carbon dioxide satellite observations completeness assessment for greenhouse gas monitoring > Methodology", "title": "Carbon dioxide satellite observations completeness assessment for greenhouse gas monitoring", "chunk_index": 3, "token_count": 680, "text_raw": "**Spatial seasonal means and trends** are presented and assessed using the new **XCO2 v4.4 Level 3 gridded product (OBS4MIPS)** [[6]](https://cds.climate.copernicus.eu/datasets/satellite-carbon-dioxide?tab=overview), which has been generated using the Level 2 EMMA products [[2]](http://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PUGS_GHG_main_latest.pdf) as input.\n\n**To show how data coverage varies between years and seasons**, we have calculated and plotted the **average XCO2 values for singular seasons and years**.\n\n**Spatial trends** are calculated using a linear model (i.e. Theil-Sen slope estimator) over monthly anomalies (i.e. actual monthly values minus climatological monthly means). **This should be treated with caution as the long-term trend of atmospheric CO2 is not strictly linear**. The statistical significance of the trends is assessed using the Mann-Kendall test. Similar to [[3]](https://amt.copernicus.org/articles/13/789/2020/), only land pixels are considered to avoid artefacts related to different data availability over oceans (the data product is land only for 2003-2008).\n\n'The analysis and results are organised in the following steps, which are detailed in the sections below:'\n\n**[](template:section-1)**\n * Import all the relevant packages.\n * Choose the temporal and spatial coverage, land mask.\n * Cache needed functions.\n\n**[](template:section-2)**\n * In this section, we define the data request to CDS. *Please note that users have to accept the GHG-CCI Licence for getting the dataset (see [[6]](https://cds.climate.copernicus.eu/datasets/satellite-carbon-dioxide?tab=overview)).*\n\n**[](template:section-3)**\n * To show how data coverage varies between years and seasons, we have plotted spatial average XCO2 values for seasons and years. Seasons are defined as December - February (DJF), March - May (MAM), June - August (JJA) and September – November (SON).\n\n**[](template:section-4)**\n * Trends are calculated using a linear model (i.e. Theil-Sen slope estimator) over monthly anomalies (i.e. actual monthly values minus climatological monthly means). The statistical significance of the trends is assessed using the Mann-Kendall test.\n\n**[](template:section-5)**\n * Here main results are summarised\n\nPlease note that **no data sub-setting or selection has been performed**, neither as a function of the number of individual Level 2 observations leading to the reported Level 3 values (\"co2_nobs\"), nor as a function of the standard error (\"xco2_stderr\") provided in the data files.", "text_with_prefix": "EQC Quality Assessment: \"Carbon dioxide satellite observations completeness assessment for greenhouse gas monitoring\"\nDataset: satellite-carbon-dioxide [CDS]\nAspect: trend-assessment_q01 | Category: Satellite_ECVs\nSection: Carbon dioxide satellite observations completeness assessment for greenhouse gas monitoring > Methodology\n---\n**Spatial seasonal means and trends** are presented and assessed using the new **XCO2 v4.4 Level 3 gridded product (OBS4MIPS)** [[6]](https://cds.climate.copernicus.eu/datasets/satellite-carbon-dioxide?tab=overview), which has been generated using the Level 2 EMMA products [[2]](http://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PUGS_GHG_main_latest.pdf) as input.\n\n**To show how data coverage varies between years and seasons**, we have calculated and plotted the **average XCO2 values for singular seasons and years**.\n\n**Spatial trends** are calculated using a linear model (i.e. Theil-Sen slope estimator) over monthly anomalies (i.e. actual monthly values minus climatological monthly means). **This should be treated with caution as the long-term trend of atmospheric CO2 is not strictly linear**. The statistical significance of the trends is assessed using the Mann-Kendall test. Similar to [[3]](https://amt.copernicus.org/articles/13/789/2020/), only land pixels are considered to avoid artefacts related to different data availability over oceans (the data product is land only for 2003-2008).\n\n'The analysis and results are organised in the following steps, which are detailed in the sections below:'\n\n**[](template:section-1)**\n * Import all the relevant packages.\n * Choose the temporal and spatial coverage, land mask.\n * Cache needed functions.\n\n**[](template:section-2)**\n * In this section, we define the data request to CDS. *Please note that users have to accept the GHG-CCI Licence for getting the dataset (see [[6]](https://cds.climate.copernicus.eu/datasets/satellite-carbon-dioxide?tab=overview)).*\n\n**[](template:section-3)**\n * To show how data coverage varies between years and seasons, we have plotted spatial average XCO2 values for seasons and years. Seasons are defined as December - February (DJF), March - May (MAM), June - August (JJA) and September – November (SON).\n\n**[](template:section-4)**\n * Trends are calculated using a linear model (i.e. Theil-Sen slope estimator) over monthly anomalies (i.e. actual monthly values minus climatological monthly means). The statistical significance of the trends is assessed using the Mann-Kendall test.\n\n**[](template:section-5)**\n * Here main results are summarised\n\nPlease note that **no data sub-setting or selection has been performed**, neither as a function of the number of individual Level 2 observations leading to the reported Level 3 values (\"co2_nobs\"), nor as a function of the standard error (\"xco2_stderr\") provided in the data files."} {"chunk_id": "satellite_satellite-carbon-dioxide_trend-assessment_q01__5d1de1591127", "report_id": "satellite_satellite-carbon-dioxide_trend-assessment_q01", "dataset_id": "satellite-carbon-dioxide", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q01", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Carbon dioxide satellite observations completeness assessment for greenhouse gas monitoring > Analysis and results > 1. Choose the data to use and set-up the code > Choose temporal and spatial coverage, land mask", "title": "Carbon dioxide satellite observations completeness assessment for greenhouse gas monitoring", "chunk_index": 4, "token_count": 239, "text_raw": "In this section, we define the parameters to be ingested by the code (users can change the temporal periods of analyses, the region of interest and activate/deactivate the land masking). With the aim to compare our results with existing literature [[7]](https://library.wmo.int/idurl/4/58743), the analysis is limited to time period from 2012 to 2021. However, specific data subsetting can be set by the users, so that data can be also retrived for other specific periods or regions.\n\nChoose variable (xch4 or xco2)\nChoose a time period\nMinimum value of land fraction used for masking\nmin_land_fraction = None # None: Do not apply mask\nDefine region for analysis", "text_with_prefix": "EQC Quality Assessment: \"Carbon dioxide satellite observations completeness assessment for greenhouse gas monitoring\"\nDataset: satellite-carbon-dioxide [CDS]\nAspect: trend-assessment_q01 | Category: Satellite_ECVs\nSection: Carbon dioxide satellite observations completeness assessment for greenhouse gas monitoring > Analysis and results > 1. Choose the data to use and set-up the code > Choose temporal and spatial coverage, land mask\n---\nIn this section, we define the parameters to be ingested by the code (users can change the temporal periods of analyses, the region of interest and activate/deactivate the land masking). With the aim to compare our results with existing literature [[7]](https://library.wmo.int/idurl/4/58743), the analysis is limited to time period from 2012 to 2021. However, specific data subsetting can be set by the users, so that data can be also retrived for other specific periods or regions.\n\nChoose variable (xch4 or xco2)\nChoose a time period\nMinimum value of land fraction used for masking\nmin_land_fraction = None # None: Do not apply mask\nDefine region for analysis"} {"chunk_id": "satellite_satellite-carbon-dioxide_trend-assessment_q01__2162ca8b4bb1", "report_id": "satellite_satellite-carbon-dioxide_trend-assessment_q01", "dataset_id": "satellite-carbon-dioxide", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q01", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Carbon dioxide satellite observations completeness assessment for greenhouse gas monitoring > Analysis and results > 1. Choose the data to use and set-up the code > Chache needed functions", "title": "Carbon dioxide satellite observations completeness assessment for greenhouse gas monitoring", "chunk_index": 5, "token_count": 366, "text_raw": "In this section, we cached a list of functions used in the analyses.\n\n- The function `get_da` (`get_da_nomask`) is used to subset the data for the defined time period and spatial region by applying (not appliyng) land mask (as a function of *min_land_fraction*).\n\n- The function `convert_units` rescale xco2 mole fraction to part per millions (ppm).\n\n- The `seasonal_weighted_mean` function extracts the regional means over the selected domains. It uses spatial weighting to account for the latitudinal dependence of the grid size in the lon-lat grids used for the reanalysis and for the forecast models. It is used by the function `compute_seasonal_timeseries_nomask` to provide the seasonal xco2 averaged value for each single year (Figure 1).\n\n- The function `compute_anomaly_trends` is used to calculate the trend and the related statistical significance.\n\n- The `compute_monthly_anomalies` function is used to derive the monthly xco2 anomalies before calculating the trends.\n\nShift years (shift -1 to get D(year-1)J(year)F(year))\nGet rid of 1st JF and last D, so it become [MAM, JJA, SON, DJF, ..., SON]\nMann-Kendall\nDetrended anomalies\n\n(template:section-2)=", "text_with_prefix": "EQC Quality Assessment: \"Carbon dioxide satellite observations completeness assessment for greenhouse gas monitoring\"\nDataset: satellite-carbon-dioxide [CDS]\nAspect: trend-assessment_q01 | Category: Satellite_ECVs\nSection: Carbon dioxide satellite observations completeness assessment for greenhouse gas monitoring > Analysis and results > 1. Choose the data to use and set-up the code > Chache needed functions\n---\nIn this section, we cached a list of functions used in the analyses.\n\n- The function `get_da` (`get_da_nomask`) is used to subset the data for the defined time period and spatial region by applying (not appliyng) land mask (as a function of *min_land_fraction*).\n\n- The function `convert_units` rescale xco2 mole fraction to part per millions (ppm).\n\n- The `seasonal_weighted_mean` function extracts the regional means over the selected domains. It uses spatial weighting to account for the latitudinal dependence of the grid size in the lon-lat grids used for the reanalysis and for the forecast models. It is used by the function `compute_seasonal_timeseries_nomask` to provide the seasonal xco2 averaged value for each single year (Figure 1).\n\n- The function `compute_anomaly_trends` is used to calculate the trend and the related statistical significance.\n\n- The `compute_monthly_anomalies` function is used to derive the monthly xco2 anomalies before calculating the trends.\n\nShift years (shift -1 to get D(year-1)J(year)F(year))\nGet rid of 1st JF and last D, so it become [MAM, JJA, SON, DJF, ..., SON]\nMann-Kendall\nDetrended anomalies\n\n(template:section-2)="} {"chunk_id": "satellite_satellite-carbon-dioxide_trend-assessment_q01__425bbd85f48c", "report_id": "satellite_satellite-carbon-dioxide_trend-assessment_q01", "dataset_id": "satellite-carbon-dioxide", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q01", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Carbon dioxide satellite observations completeness assessment for greenhouse gas monitoring > Analysis and results > 2. Retrieve XCO2 data (OBS4MIPS)", "title": "Carbon dioxide satellite observations completeness assessment for greenhouse gas monitoring", "chunk_index": 6, "token_count": 118, "text_raw": "In this section, we define the data request to CDS (data product OBS4MIPS, Level 3, version 4.4, XCO2).\n\n(template:section-3)=", "text_with_prefix": "EQC Quality Assessment: \"Carbon dioxide satellite observations completeness assessment for greenhouse gas monitoring\"\nDataset: satellite-carbon-dioxide [CDS]\nAspect: trend-assessment_q01 | Category: Satellite_ECVs\nSection: Carbon dioxide satellite observations completeness assessment for greenhouse gas monitoring > Analysis and results > 2. Retrieve XCO2 data (OBS4MIPS)\n---\nIn this section, we define the data request to CDS (data product OBS4MIPS, Level 3, version 4.4, XCO2).\n\n(template:section-3)="} {"chunk_id": "satellite_satellite-carbon-dioxide_trend-assessment_q01__f7f6eb781d34", "report_id": "satellite_satellite-carbon-dioxide_trend-assessment_q01", "dataset_id": "satellite-carbon-dioxide", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q01", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Carbon dioxide satellite observations completeness assessment for greenhouse gas monitoring > Analysis and results > 3. Compute and plot the global variability of seasonal XCO2", "title": "Carbon dioxide satellite observations completeness assessment for greenhouse gas monitoring", "chunk_index": 7, "token_count": 262, "text_raw": "To show how data coverage varies between years and seasons, in this section we plot the spatial average XCO2 values for seasons and years. Seasons are defined as December - February (DJF), March - May (MAM), June - August (JJA) and September – November (SON).\n\nAnalysis of global variability of seasonal XCO2\nTo invalidate the cache you can pass the argument invalidate_cache=True\n\n```text\n100%|██████████| 1/1 [00:00<00:00, 2.95it/s]\n```\n\n*Figure 1. Global annual variability of seasonal XCO2. Different panels represent the individual seasons (column) and years (rows). Seasons are defined as December - February (DJF), March - May (MAM), June - August (JJA) and September - November (SON).*\n\n(template:section-4)=", "text_with_prefix": "EQC Quality Assessment: \"Carbon dioxide satellite observations completeness assessment for greenhouse gas monitoring\"\nDataset: satellite-carbon-dioxide [CDS]\nAspect: trend-assessment_q01 | Category: Satellite_ECVs\nSection: Carbon dioxide satellite observations completeness assessment for greenhouse gas monitoring > Analysis and results > 3. Compute and plot the global variability of seasonal XCO2\n---\nTo show how data coverage varies between years and seasons, in this section we plot the spatial average XCO2 values for seasons and years. Seasons are defined as December - February (DJF), March - May (MAM), June - August (JJA) and September – November (SON).\n\nAnalysis of global variability of seasonal XCO2\nTo invalidate the cache you can pass the argument invalidate_cache=True\n\n```text\n100%|██████████| 1/1 [00:00<00:00, 2.95it/s]\n```\n\n*Figure 1. Global annual variability of seasonal XCO2. Different panels represent the individual seasons (column) and years (rows). Seasons are defined as December - February (DJF), March - May (MAM), June - August (JJA) and September - November (SON).*\n\n(template:section-4)="} {"chunk_id": "satellite_satellite-carbon-dioxide_trend-assessment_q01__f92a7d172183", "report_id": "satellite_satellite-carbon-dioxide_trend-assessment_q01", "dataset_id": "satellite-carbon-dioxide", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q01", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Carbon dioxide satellite observations completeness assessment for greenhouse gas monitoring > Analysis and results > 4. Compute global and spatial trends", "title": "Carbon dioxide satellite observations completeness assessment for greenhouse gas monitoring", "chunk_index": 8, "token_count": 285, "text_raw": "In this section, XCO2 spatial trends are calculated and statistical significances is assessed for land pixels (“land_fraction > 0.5”). Global statistics for individual pixels are also reported as ppm/month. Please note that the low growth rates observed in the tropics are probably a result of sparse data availability due to the frequent presence of clouds.\n\nCalculation of global and pixel trends\nTo invalidate the cache you can pass the argument invalidate_cache=True\n\n```text\n100%|██████████| 1/1 [00:00<00:00, 3.43it/s]\n```\n\n*Figure 2. Trends of XCO2 given in ppm/month calculated by using linear model (Theil-Sen) and Mann-Kendall test for statistical significance. Over each pixel, trends are calculated for monthly anomalies over land (\"land_fraction > 0.5\"). The shaded areas indicate pixels that did not pass the Mann-Kendall significance test. Global statistics for individual pixels are shown on the left.*\n\n(template:section-5)=", "text_with_prefix": "EQC Quality Assessment: \"Carbon dioxide satellite observations completeness assessment for greenhouse gas monitoring\"\nDataset: satellite-carbon-dioxide [CDS]\nAspect: trend-assessment_q01 | Category: Satellite_ECVs\nSection: Carbon dioxide satellite observations completeness assessment for greenhouse gas monitoring > Analysis and results > 4. Compute global and spatial trends\n---\nIn this section, XCO2 spatial trends are calculated and statistical significances is assessed for land pixels (“land_fraction > 0.5”). Global statistics for individual pixels are also reported as ppm/month. Please note that the low growth rates observed in the tropics are probably a result of sparse data availability due to the frequent presence of clouds.\n\nCalculation of global and pixel trends\nTo invalidate the cache you can pass the argument invalidate_cache=True\n\n```text\n100%|██████████| 1/1 [00:00<00:00, 3.43it/s]\n```\n\n*Figure 2. Trends of XCO2 given in ppm/month calculated by using linear model (Theil-Sen) and Mann-Kendall test for statistical significance. Over each pixel, trends are calculated for monthly anomalies over land (\"land_fraction > 0.5\"). The shaded areas indicate pixels that did not pass the Mann-Kendall significance test. Global statistics for individual pixels are shown on the left.*\n\n(template:section-5)="} {"chunk_id": "satellite_satellite-carbon-dioxide_trend-assessment_q01__b5059773d6e3", "report_id": "satellite_satellite-carbon-dioxide_trend-assessment_q01", "dataset_id": "satellite-carbon-dioxide", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q01", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Carbon dioxide satellite observations completeness assessment for greenhouse gas monitoring > Analysis and results > 5. Results", "title": "Carbon dioxide satellite observations completeness assessment for greenhouse gas monitoring", "chunk_index": 9, "token_count": 281, "text_raw": "The seasonality of the availability of measurements is evident, with values available in the Southern Hemisphere (SH) mid and high latitudes from September to March, whereas values for the Northern Hemisphere (NH) mid and high latitudes are mostly available from April to August (Figure 1).\n\nFor 2012-2021, the mean global XCO2 growth rate is 2.28 ppm/yr (Figure 2). Taking into account the associated uncertainties and the different calculation methods, this value is in good agreement with the value of 2.46 ppm/yr provided by the WMO global in-situ observations [[7]](https://library.wmo.int/idurl/4/58743). The tropics and the middle and high latitudes are the regions that are not covered by measurements throughout the year or are characterised by higher uncertainty (see also [[3]](https://amt.copernicus.org/articles/13/789/2020/)), and therefore the trend values are more variable and in some cases not statistically significant.", "text_with_prefix": "EQC Quality Assessment: \"Carbon dioxide satellite observations completeness assessment for greenhouse gas monitoring\"\nDataset: satellite-carbon-dioxide [CDS]\nAspect: trend-assessment_q01 | Category: Satellite_ECVs\nSection: Carbon dioxide satellite observations completeness assessment for greenhouse gas monitoring > Analysis and results > 5. Results\n---\nThe seasonality of the availability of measurements is evident, with values available in the Southern Hemisphere (SH) mid and high latitudes from September to March, whereas values for the Northern Hemisphere (NH) mid and high latitudes are mostly available from April to August (Figure 1).\n\nFor 2012-2021, the mean global XCO2 growth rate is 2.28 ppm/yr (Figure 2). Taking into account the associated uncertainties and the different calculation methods, this value is in good agreement with the value of 2.46 ppm/yr provided by the WMO global in-situ observations [[7]](https://library.wmo.int/idurl/4/58743). The tropics and the middle and high latitudes are the regions that are not covered by measurements throughout the year or are characterised by higher uncertainty (see also [[3]](https://amt.copernicus.org/articles/13/789/2020/)), and therefore the trend values are more variable and in some cases not statistically significant."} {"chunk_id": "satellite_satellite-carbon-dioxide_trend-assessment_q01__69baeb5b0281", "report_id": "satellite_satellite-carbon-dioxide_trend-assessment_q01", "dataset_id": "satellite-carbon-dioxide", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q01", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Carbon dioxide satellite observations completeness assessment for greenhouse gas monitoring > ℹ️ If you want to know more > Key resources", "title": "Carbon dioxide satellite observations completeness assessment for greenhouse gas monitoring", "chunk_index": 10, "token_count": 189, "text_raw": "The CDS catalogue entries for the data used were:\n* Carbon dioxide data from 2002 to present derived from satellite observations: https://cds.climate.copernicus.eu/datasets/satellite-carbon-dioxide?tab=overview\n\nCode libraries used:\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)", "text_with_prefix": "EQC Quality Assessment: \"Carbon dioxide satellite observations completeness assessment for greenhouse gas monitoring\"\nDataset: satellite-carbon-dioxide [CDS]\nAspect: trend-assessment_q01 | Category: Satellite_ECVs\nSection: Carbon dioxide satellite observations completeness assessment for greenhouse gas monitoring > ℹ️ If you want to know more > Key resources\n---\nThe CDS catalogue entries for the data used were:\n* Carbon dioxide data from 2002 to present derived from satellite observations: https://cds.climate.copernicus.eu/datasets/satellite-carbon-dioxide?tab=overview\n\nCode libraries used:\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)"} {"chunk_id": "satellite_satellite-carbon-dioxide_trend-assessment_q01__31405265f9b5", "report_id": "satellite_satellite-carbon-dioxide_trend-assessment_q01", "dataset_id": "satellite-carbon-dioxide", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q01", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Carbon dioxide satellite observations completeness assessment for greenhouse gas monitoring > ℹ️ If you want to know more > Example of application", "title": "Carbon dioxide satellite observations completeness assessment for greenhouse gas monitoring", "chunk_index": 11, "token_count": 151, "text_raw": "Copernicus Climate Change Service (C3S). (2024). [European State of the Climate 2023](https://climate.copernicus.eu/ESOTC/2023), Full report. Specifically, [the section on greenhouse gas concentrations](https://climate.copernicus.eu/esotc/2023/greenhouse-gas-concentrations).", "text_with_prefix": "EQC Quality Assessment: \"Carbon dioxide satellite observations completeness assessment for greenhouse gas monitoring\"\nDataset: satellite-carbon-dioxide [CDS]\nAspect: trend-assessment_q01 | Category: Satellite_ECVs\nSection: Carbon dioxide satellite observations completeness assessment for greenhouse gas monitoring > ℹ️ If you want to know more > Example of application\n---\nCopernicus Climate Change Service (C3S). (2024). [European State of the Climate 2023](https://climate.copernicus.eu/ESOTC/2023), Full report. Specifically, [the section on greenhouse gas concentrations](https://climate.copernicus.eu/esotc/2023/greenhouse-gas-concentrations)."} {"chunk_id": "satellite_satellite-carbon-dioxide_trend-assessment_q01__93ab30f81d1c", "report_id": "satellite_satellite-carbon-dioxide_trend-assessment_q01", "dataset_id": "satellite-carbon-dioxide", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q01", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Carbon dioxide satellite observations completeness assessment for greenhouse gas monitoring > ℹ️ If you want to know more > References", "title": "Carbon dioxide satellite observations completeness assessment for greenhouse gas monitoring", "chunk_index": 12, "token_count": 805, "text_raw": "[[1]](https://library.wmo.int/viewer/68532/?offset=#page=4&viewer=picture&o=bookmarks&n=0&q=) World Meteorological Organization. (2023). WMO Greenhouse Gas Bullettin, 19, ISSN 2078-0796.\n\n[[2]](http://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PUGS_GHG_main_latest.pdf) Buchwitz, M. (2023). Product User Guide and Specification (PUGS) – Main document for Greenhouse Gas (GHG: CO2 & CH4) data set CDR 6 (01.2003-12.2021), C3S project 2021/C3S2_312b_Lot2_DLR/SC1, v6.3.\n\n[[3]](https://amt.copernicus.org/articles/13/789/2020/) Reuter, M., Buchwitz, M., Schneising, O., Noël, S., Bovensmann, H., Burrows, J. P., Boesch, H., Di Noia, A., Anand, J., Parker, R. J., Somkuti, P., Wu, L., Hasekamp, O. P., Aben, I., Kuze, A., Suto, H., Shiomi, K., Yoshida, Y., Morino, I., Crisp, D., O'Dell, C. W., Notholt, J., Petri, C., Warneke, T., Velazco, V. A., Deutscher, N. M., Griffith, D. W. T., Kivi, R., Pollard, D. F., Hase, F., Sussmann, R., Té, Y. V., Strong, K., Roche, S., Sha, M. K., De Mazière, M., Feist, D. G., Iraci, L. T., Roehl, C. M., Retscher, C., and Schepers, D. (2020). Ensemble-based satellite-derived carbon dioxide and methane column-averaged dry-air mole fraction data sets (2003–2018) for carbon and climate applications. Atmospheric Measurement Techniques, 13, 789–819.\n\n[[4]](https://climate.copernicus.eu/esotc/2023) Copernicus Climate Change Service (C3S). (2024). European State of the Climate 2023, Full report.\n\n[[5]](https://climate.copernicus.eu/global-climate-highlights-2023) Copernicus Climate Change Service (C3S). (2024). Global Climate Highlights 2023.\n\n[[6]](https://cds.climate.copernicus.eu/datasets/satellite-carbon-dioxide?tab=overview) Copernicus Climate Change Service, Climate Data Store, (2018): \"Carbon dioxide data from 2002 to present derived from satellite observations\". Copernicus Climate Change Service (C3S) Climate Data Store (CDS). DOI: 10.24381/cds.f74805c8.\n\n[[7]](https://library.wmo.int/idurl/4/58743) World Meteorological Organization. (2022). WMO Greenhouse Gas Bullettin, 18, ISSN 2078-0796.", "text_with_prefix": "EQC Quality Assessment: \"Carbon dioxide satellite observations completeness assessment for greenhouse gas monitoring\"\nDataset: satellite-carbon-dioxide [CDS]\nAspect: trend-assessment_q01 | Category: Satellite_ECVs\nSection: Carbon dioxide satellite observations completeness assessment for greenhouse gas monitoring > ℹ️ If you want to know more > References\n---\n[[1]](https://library.wmo.int/viewer/68532/?offset=#page=4&viewer=picture&o=bookmarks&n=0&q=) World Meteorological Organization. (2023). WMO Greenhouse Gas Bullettin, 19, ISSN 2078-0796.\n\n[[2]](http://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PUGS_GHG_main_latest.pdf) Buchwitz, M. (2023). Product User Guide and Specification (PUGS) – Main document for Greenhouse Gas (GHG: CO2 & CH4) data set CDR 6 (01.2003-12.2021), C3S project 2021/C3S2_312b_Lot2_DLR/SC1, v6.3.\n\n[[3]](https://amt.copernicus.org/articles/13/789/2020/) Reuter, M., Buchwitz, M., Schneising, O., Noël, S., Bovensmann, H., Burrows, J. P., Boesch, H., Di Noia, A., Anand, J., Parker, R. J., Somkuti, P., Wu, L., Hasekamp, O. P., Aben, I., Kuze, A., Suto, H., Shiomi, K., Yoshida, Y., Morino, I., Crisp, D., O'Dell, C. W., Notholt, J., Petri, C., Warneke, T., Velazco, V. A., Deutscher, N. M., Griffith, D. W. T., Kivi, R., Pollard, D. F., Hase, F., Sussmann, R., Té, Y. V., Strong, K., Roche, S., Sha, M. K., De Mazière, M., Feist, D. G., Iraci, L. T., Roehl, C. M., Retscher, C., and Schepers, D. (2020). Ensemble-based satellite-derived carbon dioxide and methane column-averaged dry-air mole fraction data sets (2003–2018) for carbon and climate applications. Atmospheric Measurement Techniques, 13, 789–819.\n\n[[4]](https://climate.copernicus.eu/esotc/2023) Copernicus Climate Change Service (C3S). (2024). European State of the Climate 2023, Full report.\n\n[[5]](https://climate.copernicus.eu/global-climate-highlights-2023) Copernicus Climate Change Service (C3S). (2024). Global Climate Highlights 2023.\n\n[[6]](https://cds.climate.copernicus.eu/datasets/satellite-carbon-dioxide?tab=overview) Copernicus Climate Change Service, Climate Data Store, (2018): \"Carbon dioxide data from 2002 to present derived from satellite observations\". Copernicus Climate Change Service (C3S) Climate Data Store (CDS). DOI: 10.24381/cds.f74805c8.\n\n[[7]](https://library.wmo.int/idurl/4/58743) World Meteorological Organization. (2022). WMO Greenhouse Gas Bullettin, 18, ISSN 2078-0796."} {"chunk_id": "satellite_satellite-carbon-dioxide_uncertainty-quality-flags_q02__6c9622b8d800", "report_id": "satellite_satellite-carbon-dioxide_uncertainty-quality-flags_q02", "dataset_id": "satellite-carbon-dioxide", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty-quality-flags_q02", "aspect_base": "uncertainty-quality-flags", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Spatial and temporal completeness and uncertainties of carbon dioxide satellite observations for quantifying atmospheric greenhouse gas variability", "title": "Spatial and temporal completeness and uncertainties of carbon dioxide satellite observations for quantifying atmospheric greenhouse gas variability", "chunk_index": 0, "token_count": 114, "text_raw": "Production date: 09-07-2025\n\nProduced by: National Research Council of Italy - Institute of Atmospheric Sciences and Climate, Paolo Cristofanelli and Davide Putero", "text_with_prefix": "EQC Quality Assessment: \"Spatial and temporal completeness and uncertainties of carbon dioxide satellite observations for quantifying atmospheric greenhouse gas variability\"\nDataset: satellite-carbon-dioxide [CDS]\nAspect: uncertainty-quality-flags_q02 | Category: Satellite_ECVs\nSection: Spatial and temporal completeness and uncertainties of carbon dioxide satellite observations for quantifying atmospheric greenhouse gas variability\n---\nProduction date: 09-07-2025\n\nProduced by: National Research Council of Italy - Institute of Atmospheric Sciences and Climate, Paolo Cristofanelli and Davide Putero"} {"chunk_id": "satellite_satellite-carbon-dioxide_uncertainty-quality-flags_q02__acb528a72755", "report_id": "satellite_satellite-carbon-dioxide_uncertainty-quality-flags_q02", "dataset_id": "satellite-carbon-dioxide", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty-quality-flags_q02", "aspect_base": "uncertainty-quality-flags", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Spatial and temporal completeness and uncertainties of carbon dioxide satellite observations for quantifying atmospheric greenhouse gas variability > Quality assessment question", "title": "Spatial and temporal completeness and uncertainties of carbon dioxide satellite observations for quantifying atmospheric greenhouse gas variability", "chunk_index": 1, "token_count": 529, "text_raw": "* **Are the XCO$_2$ Level 3 data suitable for identifying regions characterized by high XCO$_2$ values?**\n* **What are the uncertainties related to the XCO$_2$ satellite observations?**\n\nCarbon dioxide (CO$_2$) is the most important anthropogenic greenhouse gas, accounting for nearly 64% of the total radiative forcing by long-lived greenhouse gases [[1]](https://library.wmo.int/idurl/4/69057). The concentration of CO$_2$ in the atmosphere in 2023 was 420.0$\\,\\pm\\,$0.1 ppm [[1]](https://library.wmo.int/idurl/4/69057). Monitoring the long-term CO$_2$ variability is crucial for understanding the global carbon sources and sinks, improving [[2]](https://doi.org/10.5194/essd-17-965-2025) our knowledge of the carbon cycle, supporting the development of effective climate policies, and providing more reliable projections for future climate change.\n\nIn this assessment, we analyze the spatial variability of atmospheric column-averaged mixing ratios of CO$_2$ (XCO$_2$) over different regions using the Level 3 XCO$_2$ gridded data products (Obs4MIPs, version 4.5 [[3]](https://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PUGS_GHG_main_latest.pdf)), whose main application is related to the comparison and validation of global climate models (e.g., [[4]](https://doi.org/10.5194/bg-17-6115-2020)).\n\nIn particular, by comparison with an independent study based on a gap-free and fine-scale CO$_2$ dataset [[5]](https://doi.org/10.1016/j.envint.2023.108057), we assess the ability of this dataset to reproduce temporal and spatial anomalies in XCO$_2$, also quantifying the uncertainty provided with the dataset.\n\nThe code is included for transparency and to allow users to adapt it to their needs.", "text_with_prefix": "EQC Quality Assessment: \"Spatial and temporal completeness and uncertainties of carbon dioxide satellite observations for quantifying atmospheric greenhouse gas variability\"\nDataset: satellite-carbon-dioxide [CDS]\nAspect: uncertainty-quality-flags_q02 | Category: Satellite_ECVs\nSection: Spatial and temporal completeness and uncertainties of carbon dioxide satellite observations for quantifying atmospheric greenhouse gas variability > Quality assessment question\n---\n* **Are the XCO$_2$ Level 3 data suitable for identifying regions characterized by high XCO$_2$ values?**\n* **What are the uncertainties related to the XCO$_2$ satellite observations?**\n\nCarbon dioxide (CO$_2$) is the most important anthropogenic greenhouse gas, accounting for nearly 64% of the total radiative forcing by long-lived greenhouse gases [[1]](https://library.wmo.int/idurl/4/69057). The concentration of CO$_2$ in the atmosphere in 2023 was 420.0$\\,\\pm\\,$0.1 ppm [[1]](https://library.wmo.int/idurl/4/69057). Monitoring the long-term CO$_2$ variability is crucial for understanding the global carbon sources and sinks, improving [[2]](https://doi.org/10.5194/essd-17-965-2025) our knowledge of the carbon cycle, supporting the development of effective climate policies, and providing more reliable projections for future climate change.\n\nIn this assessment, we analyze the spatial variability of atmospheric column-averaged mixing ratios of CO$_2$ (XCO$_2$) over different regions using the Level 3 XCO$_2$ gridded data products (Obs4MIPs, version 4.5 [[3]](https://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PUGS_GHG_main_latest.pdf)), whose main application is related to the comparison and validation of global climate models (e.g., [[4]](https://doi.org/10.5194/bg-17-6115-2020)).\n\nIn particular, by comparison with an independent study based on a gap-free and fine-scale CO$_2$ dataset [[5]](https://doi.org/10.1016/j.envint.2023.108057), we assess the ability of this dataset to reproduce temporal and spatial anomalies in XCO$_2$, also quantifying the uncertainty provided with the dataset.\n\nThe code is included for transparency and to allow users to adapt it to their needs."} {"chunk_id": "satellite_satellite-carbon-dioxide_uncertainty-quality-flags_q02__163b90efc6b2", "report_id": "satellite_satellite-carbon-dioxide_uncertainty-quality-flags_q02", "dataset_id": "satellite-carbon-dioxide", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty-quality-flags_q02", "aspect_base": "uncertainty-quality-flags", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Spatial and temporal completeness and uncertainties of carbon dioxide satellite observations for quantifying atmospheric greenhouse gas variability > Quality assessment statement", "title": "Spatial and temporal completeness and uncertainties of carbon dioxide satellite observations for quantifying atmospheric greenhouse gas variability", "chunk_index": 2, "token_count": 354, "text_raw": "These are the key outcomes of this assessment\n\n* The XCO$_2$ satellite observations (Level 3, Obs4MIPs, version 4.5) provide a consistent global description of the spatial XCO$_2$ pattern and can be used to detect areas characterized by elevated XCO$_2$ values. Care should be excercised when considering data from regions affected by high uncertainties. The highest uncertainty values were found over high latitudes, the Himalayas and the tropical rainforest zone due to the occurrence of large solar zenith angles or frequent cloud cover. \n* The sampling characteristics of the dataset must be carefully considered when computing annual averages or anomalies, given the low or non-available data over high latitudes, especially during winter months.\n* As data availability and uncertainty are not constant over space and time, users are advised to consider the evolution of their values over the spatial regions and time periods of interest. \n* Users must be aware that the 5$^\\circ$x5$^\\circ$ averaging can introduce a smoothing effect in the spatial XCO$_2$ field with respect to the Level 2 datasets.\n* The main use of this data product is for comparison with climate models. Users interested in investigating the CO$_2$ fluxes occurring at the Earth's surface are advised to use products specifically designed for this purpose.\n```", "text_with_prefix": "EQC Quality Assessment: \"Spatial and temporal completeness and uncertainties of carbon dioxide satellite observations for quantifying atmospheric greenhouse gas variability\"\nDataset: satellite-carbon-dioxide [CDS]\nAspect: uncertainty-quality-flags_q02 | Category: Satellite_ECVs\nSection: Spatial and temporal completeness and uncertainties of carbon dioxide satellite observations for quantifying atmospheric greenhouse gas variability > Quality assessment statement\n---\nThese are the key outcomes of this assessment\n\n* The XCO$_2$ satellite observations (Level 3, Obs4MIPs, version 4.5) provide a consistent global description of the spatial XCO$_2$ pattern and can be used to detect areas characterized by elevated XCO$_2$ values. Care should be excercised when considering data from regions affected by high uncertainties. The highest uncertainty values were found over high latitudes, the Himalayas and the tropical rainforest zone due to the occurrence of large solar zenith angles or frequent cloud cover. \n* The sampling characteristics of the dataset must be carefully considered when computing annual averages or anomalies, given the low or non-available data over high latitudes, especially during winter months.\n* As data availability and uncertainty are not constant over space and time, users are advised to consider the evolution of their values over the spatial regions and time periods of interest. \n* Users must be aware that the 5$^\\circ$x5$^\\circ$ averaging can introduce a smoothing effect in the spatial XCO$_2$ field with respect to the Level 2 datasets.\n* The main use of this data product is for comparison with climate models. Users interested in investigating the CO$_2$ fluxes occurring at the Earth's surface are advised to use products specifically designed for this purpose.\n```"} {"chunk_id": "satellite_satellite-carbon-dioxide_uncertainty-quality-flags_q02__72340572299e", "report_id": "satellite_satellite-carbon-dioxide_uncertainty-quality-flags_q02", "dataset_id": "satellite-carbon-dioxide", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty-quality-flags_q02", "aspect_base": "uncertainty-quality-flags", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Spatial and temporal completeness and uncertainties of carbon dioxide satellite observations for quantifying atmospheric greenhouse gas variability > Methodology", "title": "Spatial and temporal completeness and uncertainties of carbon dioxide satellite observations for quantifying atmospheric greenhouse gas variability", "chunk_index": 3, "token_count": 661, "text_raw": "This notebook aims to provide users with information on the availability of XCO$_2$ Level 3 products, both globally and for specific user-defined regions. Additionally, it offers insights into the uncertainties associated with XCO$_2$ measurements and how these variables vary over time.\n\nMore specifically, this notebook aims to:\n* Produce global maps of yearly XCO$_2$ anomalies for comparison with an external dataset [[5]](https://doi.org/10.1016/j.envint.2023.108057). Please note that, in order to minimise the impact of non-uniform spatial coverage over the different seasons of the Obs4MIPs data product [[3]](https://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PUGS_GHG_main_latest.pdf), only data from 60°N to 60°S was considered in global spatial analyses. This approach is similar to that in [[4]](https://doi.org/10.5194/bg-17-6115-2020). \n* Generate global maps for the corresponding XCO$_2$ uncertainty (\"xco2_stderr\").\n* Produce maps of XCO$_2$ for specific regions over land [[5]](https://doi.org/10.1016/j.envint.2023.108057).\n* Create time series of selected XCO$_2$ variables (\"xco2\", \"xco2_stderr\" and \"xco2_nobs\"), for specific regions. Please note that this analysis can be customized by the user to include one additional variable (not shown here): \"xco2_stddev\", which represents the standard deviation of the XCO$_2$ Level 2 observations within each grid box.\n\nThe analysis and results are organized in the following steps, which are detailed in the sections below:\n\n**[](template:section-1)**\n * Import all relevant packages.\n * Define the temporal and spatial coverage, land mask, and spatial regions for the analysis.\n * Cache needed functions.\n\n**[](template:section-2)**\n * Define the data request to CDS.\n * Define the data request for the external dataset.\n\n**[](template:section-3)**\n\nThis section presents several results for the XCO$_2$ products, i.e.:\n * Annual global maps of the XCO$_2$ anomalies for 2015-2020 and comparison between CDS and the external dataset\n * Annual global maps of the XCO$_2$ uncertainties for 2015-2020\n * A focus of the XCO$_2$ maps over specific regions, and considering a single month (January 2016)\n * Global and regional time series of XCO$_2$, related uncertainties and data availability over the entire data period (2003-2022)", "text_with_prefix": "EQC Quality Assessment: \"Spatial and temporal completeness and uncertainties of carbon dioxide satellite observations for quantifying atmospheric greenhouse gas variability\"\nDataset: satellite-carbon-dioxide [CDS]\nAspect: uncertainty-quality-flags_q02 | Category: Satellite_ECVs\nSection: Spatial and temporal completeness and uncertainties of carbon dioxide satellite observations for quantifying atmospheric greenhouse gas variability > Methodology\n---\nThis notebook aims to provide users with information on the availability of XCO$_2$ Level 3 products, both globally and for specific user-defined regions. Additionally, it offers insights into the uncertainties associated with XCO$_2$ measurements and how these variables vary over time.\n\nMore specifically, this notebook aims to:\n* Produce global maps of yearly XCO$_2$ anomalies for comparison with an external dataset [[5]](https://doi.org/10.1016/j.envint.2023.108057). Please note that, in order to minimise the impact of non-uniform spatial coverage over the different seasons of the Obs4MIPs data product [[3]](https://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PUGS_GHG_main_latest.pdf), only data from 60°N to 60°S was considered in global spatial analyses. This approach is similar to that in [[4]](https://doi.org/10.5194/bg-17-6115-2020). \n* Generate global maps for the corresponding XCO$_2$ uncertainty (\"xco2_stderr\").\n* Produce maps of XCO$_2$ for specific regions over land [[5]](https://doi.org/10.1016/j.envint.2023.108057).\n* Create time series of selected XCO$_2$ variables (\"xco2\", \"xco2_stderr\" and \"xco2_nobs\"), for specific regions. Please note that this analysis can be customized by the user to include one additional variable (not shown here): \"xco2_stddev\", which represents the standard deviation of the XCO$_2$ Level 2 observations within each grid box.\n\nThe analysis and results are organized in the following steps, which are detailed in the sections below:\n\n**[](template:section-1)**\n * Import all relevant packages.\n * Define the temporal and spatial coverage, land mask, and spatial regions for the analysis.\n * Cache needed functions.\n\n**[](template:section-2)**\n * Define the data request to CDS.\n * Define the data request for the external dataset.\n\n**[](template:section-3)**\n\nThis section presents several results for the XCO$_2$ products, i.e.:\n * Annual global maps of the XCO$_2$ anomalies for 2015-2020 and comparison between CDS and the external dataset\n * Annual global maps of the XCO$_2$ uncertainties for 2015-2020\n * A focus of the XCO$_2$ maps over specific regions, and considering a single month (January 2016)\n * Global and regional time series of XCO$_2$, related uncertainties and data availability over the entire data period (2003-2022)"} {"chunk_id": "satellite_satellite-carbon-dioxide_uncertainty-quality-flags_q02__9d9eb86c24d9", "report_id": "satellite_satellite-carbon-dioxide_uncertainty-quality-flags_q02", "dataset_id": "satellite-carbon-dioxide", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty-quality-flags_q02", "aspect_base": "uncertainty-quality-flags", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Spatial and temporal completeness and uncertainties of carbon dioxide satellite observations for quantifying atmospheric greenhouse gas variability > Analysis and results > 1. Choose the data to use and set-up the code > Define temporal and spatial coverage, land mask, and spatial regions for the analysis", "title": "Spatial and temporal completeness and uncertainties of carbon dioxide satellite observations for quantifying atmospheric greenhouse gas variability", "chunk_index": 4, "token_count": 185, "text_raw": "In this section, we define the parameters to be ingested by the code (that can be customized by the user), i.e.: \n* the temporal period of analysis;\n* the activation/deactivation of the land masking;\n* the regions of interest;\n* the link to the external dataset.\n\nSingle time to display\nRange for 2015-2020\nExternal data", "text_with_prefix": "EQC Quality Assessment: \"Spatial and temporal completeness and uncertainties of carbon dioxide satellite observations for quantifying atmospheric greenhouse gas variability\"\nDataset: satellite-carbon-dioxide [CDS]\nAspect: uncertainty-quality-flags_q02 | Category: Satellite_ECVs\nSection: Spatial and temporal completeness and uncertainties of carbon dioxide satellite observations for quantifying atmospheric greenhouse gas variability > Analysis and results > 1. Choose the data to use and set-up the code > Define temporal and spatial coverage, land mask, and spatial regions for the analysis\n---\nIn this section, we define the parameters to be ingested by the code (that can be customized by the user), i.e.: \n* the temporal period of analysis;\n* the activation/deactivation of the land masking;\n* the regions of interest;\n* the link to the external dataset.\n\nSingle time to display\nRange for 2015-2020\nExternal data"} {"chunk_id": "satellite_satellite-carbon-dioxide_uncertainty-quality-flags_q02__7a893ed8c7e6", "report_id": "satellite_satellite-carbon-dioxide_uncertainty-quality-flags_q02", "dataset_id": "satellite-carbon-dioxide", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty-quality-flags_q02", "aspect_base": "uncertainty-quality-flags", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Spatial and temporal completeness and uncertainties of carbon dioxide satellite observations for quantifying atmospheric greenhouse gas variability > Analysis and results > 1. Choose the data to use and set-up the code > Cache needed functions", "title": "Spatial and temporal completeness and uncertainties of carbon dioxide satellite observations for quantifying atmospheric greenhouse gas variability", "chunk_index": 5, "token_count": 245, "text_raw": "In this section, we cached a list of functions used in the analyses.\n\n* The `convert_units` function rescales XCO$_2$ mole fraction (and related variables, except \"xco2_nobs\") to parts per million (ppm).\n\n* The `mask_scale_and_regionalise` function extracts the XCO$_2$ data over the selected spatial region. It uses spatial weighting to account for the latitudinal dependence of the grid size in the lon/lat grids used for the reanalysis and for the forecast models. It uses the `convert_units` function for rescaling the values to ppm, and it applies the threshold (if any) on the minimum land fraction.\n\n(template:section-2)=", "text_with_prefix": "EQC Quality Assessment: \"Spatial and temporal completeness and uncertainties of carbon dioxide satellite observations for quantifying atmospheric greenhouse gas variability\"\nDataset: satellite-carbon-dioxide [CDS]\nAspect: uncertainty-quality-flags_q02 | Category: Satellite_ECVs\nSection: Spatial and temporal completeness and uncertainties of carbon dioxide satellite observations for quantifying atmospheric greenhouse gas variability > Analysis and results > 1. Choose the data to use and set-up the code > Cache needed functions\n---\nIn this section, we cached a list of functions used in the analyses.\n\n* The `convert_units` function rescales XCO$_2$ mole fraction (and related variables, except \"xco2_nobs\") to parts per million (ppm).\n\n* The `mask_scale_and_regionalise` function extracts the XCO$_2$ data over the selected spatial region. It uses spatial weighting to account for the latitudinal dependence of the grid size in the lon/lat grids used for the reanalysis and for the forecast models. It uses the `convert_units` function for rescaling the values to ppm, and it applies the threshold (if any) on the minimum land fraction.\n\n(template:section-2)="} {"chunk_id": "satellite_satellite-carbon-dioxide_uncertainty-quality-flags_q02__320d06f5897c", "report_id": "satellite_satellite-carbon-dioxide_uncertainty-quality-flags_q02", "dataset_id": "satellite-carbon-dioxide", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty-quality-flags_q02", "aspect_base": "uncertainty-quality-flags", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Spatial and temporal completeness and uncertainties of carbon dioxide satellite observations for quantifying atmospheric greenhouse gas variability > Analysis and results > 2. Retrieve data > 2.1 Obs4MIPs", "title": "Spatial and temporal completeness and uncertainties of carbon dioxide satellite observations for quantifying atmospheric greenhouse gas variability", "chunk_index": 6, "token_count": 604, "text_raw": "In this section, we define the data request to CDS (data product Obs4MIPs, Level 3, version 4.5, XCO$_2$) and download the datasets (i.e., the 2015-2020 period, and the whole dataset).\n\nRetrieve the 2015-2020 dataset\nRetrieve the whole dataset, for plotting the time series\n\n```text\nregion='global'\n```\n\n```text\n100%|██████████| 1/1 [00:00<00:00, 27.82it/s]\n```\n\n```text\nregion='global_subset'\n```\n\n```text\n100%|██████████| 1/1 [00:00<00:00, 12.37it/s]\n```\n\n```text\nregion='north_america'\n```\n\n```text\n100%|██████████| 1/1 [00:00<00:00, 31.65it/s]\n```\n\n```text\nregion='europe_africa'\n```\n\n```text\n100%|██████████| 1/1 [00:00<00:00, 11.27it/s]\n```\n\n```text\nregion='east_asia'\n```\n\n```text\n100%|██████████| 1/1 [00:00<00:00, 43.42it/s]\n```\n\n```text\nregion='global'\n```\n\n```text\n100%|██████████| 1/1 [00:00<00:00, 34.60it/s]\n```\n\n```text\nregion='global_subset'\n```\n\n```text\n100%|██████████| 1/1 [00:00<00:00, 38.29it/s]\n```\n\n```text\nregion='north_america'\n```\n\n```text\n100%|██████████| 1/1 [00:00<00:00, 6.98it/s]\n```\n\n```text\nregion='europe_africa'\n```\n\n```text\n100%|██████████| 1/1 [00:00<00:00, 23.72it/s]\n```\n\n```text\nregion='east_asia'\n```\n\n```text\n100%|██████████| 1/1 [00:00<00:00, 44.67it/s]\n```", "text_with_prefix": "EQC Quality Assessment: \"Spatial and temporal completeness and uncertainties of carbon dioxide satellite observations for quantifying atmospheric greenhouse gas variability\"\nDataset: satellite-carbon-dioxide [CDS]\nAspect: uncertainty-quality-flags_q02 | Category: Satellite_ECVs\nSection: Spatial and temporal completeness and uncertainties of carbon dioxide satellite observations for quantifying atmospheric greenhouse gas variability > Analysis and results > 2. Retrieve data > 2.1 Obs4MIPs\n---\nIn this section, we define the data request to CDS (data product Obs4MIPs, Level 3, version 4.5, XCO$_2$) and download the datasets (i.e., the 2015-2020 period, and the whole dataset).\n\nRetrieve the 2015-2020 dataset\nRetrieve the whole dataset, for plotting the time series\n\n```text\nregion='global'\n```\n\n```text\n100%|██████████| 1/1 [00:00<00:00, 27.82it/s]\n```\n\n```text\nregion='global_subset'\n```\n\n```text\n100%|██████████| 1/1 [00:00<00:00, 12.37it/s]\n```\n\n```text\nregion='north_america'\n```\n\n```text\n100%|██████████| 1/1 [00:00<00:00, 31.65it/s]\n```\n\n```text\nregion='europe_africa'\n```\n\n```text\n100%|██████████| 1/1 [00:00<00:00, 11.27it/s]\n```\n\n```text\nregion='east_asia'\n```\n\n```text\n100%|██████████| 1/1 [00:00<00:00, 43.42it/s]\n```\n\n```text\nregion='global'\n```\n\n```text\n100%|██████████| 1/1 [00:00<00:00, 34.60it/s]\n```\n\n```text\nregion='global_subset'\n```\n\n```text\n100%|██████████| 1/1 [00:00<00:00, 38.29it/s]\n```\n\n```text\nregion='north_america'\n```\n\n```text\n100%|██████████| 1/1 [00:00<00:00, 6.98it/s]\n```\n\n```text\nregion='europe_africa'\n```\n\n```text\n100%|██████████| 1/1 [00:00<00:00, 23.72it/s]\n```\n\n```text\nregion='east_asia'\n```\n\n```text\n100%|██████████| 1/1 [00:00<00:00, 44.67it/s]\n```"} {"chunk_id": "satellite_satellite-carbon-dioxide_uncertainty-quality-flags_q02__61eeffcfcab4", "report_id": "satellite_satellite-carbon-dioxide_uncertainty-quality-flags_q02", "dataset_id": "satellite-carbon-dioxide", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty-quality-flags_q02", "aspect_base": "uncertainty-quality-flags", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Spatial and temporal completeness and uncertainties of carbon dioxide satellite observations for quantifying atmospheric greenhouse gas variability > Analysis and results > 2. Retrieve data > 2.2 Zhang et al. (2023) dataset", "title": "Spatial and temporal completeness and uncertainties of carbon dioxide satellite observations for quantifying atmospheric greenhouse gas variability", "chunk_index": 7, "token_count": 160, "text_raw": "In this section, we define the data request to the external dataset used for the comparison excercise.\n\n```text\nregion='global'\nregion='global_subset'\nregion='north_america'\nregion='europe_africa'\nregion='east_asia'\n```\n\n(template:section-3)=", "text_with_prefix": "EQC Quality Assessment: \"Spatial and temporal completeness and uncertainties of carbon dioxide satellite observations for quantifying atmospheric greenhouse gas variability\"\nDataset: satellite-carbon-dioxide [CDS]\nAspect: uncertainty-quality-flags_q02 | Category: Satellite_ECVs\nSection: Spatial and temporal completeness and uncertainties of carbon dioxide satellite observations for quantifying atmospheric greenhouse gas variability > Analysis and results > 2. Retrieve data > 2.2 Zhang et al. (2023) dataset\n---\nIn this section, we define the data request to the external dataset used for the comparison excercise.\n\n```text\nregion='global'\nregion='global_subset'\nregion='north_america'\nregion='europe_africa'\nregion='east_asia'\n```\n\n(template:section-3)="} {"chunk_id": "satellite_satellite-carbon-dioxide_uncertainty-quality-flags_q02__fb5c205f86ba", "report_id": "satellite_satellite-carbon-dioxide_uncertainty-quality-flags_q02", "dataset_id": "satellite-carbon-dioxide", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty-quality-flags_q02", "aspect_base": "uncertainty-quality-flags", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Spatial and temporal completeness and uncertainties of carbon dioxide satellite observations for quantifying atmospheric greenhouse gas variability > Analysis and results > 3. Data analysis for the different variables > Global maps of XCO$_2$ anomalies", "title": "Spatial and temporal completeness and uncertainties of carbon dioxide satellite observations for quantifying atmospheric greenhouse gas variability", "chunk_index": 8, "token_count": 1009, "text_raw": "In this section, global maps of XCO$_2$ anomalies for the years 2015–2020 are presented for the latitudes spanning from 60°N to 60°S. The calculation of anomalies is achieved by the subtraction of the annual global average from the global XCO$_2$ concentrations, in accordance with the approach outlined in [[5]](https://doi.org/10.1016/j.envint.2023.108057). Prior to the execution of the calculations, a land mask was implemented. A comparison was made between the XCO$_2$ anomalies from the Obs4MIPs (version 4.5) dataset and the results from the gap-free and fine-scale XCO$_2$ dataset created by [[5]](https://doi.org/10.1016/j.envint.2023.108057). Prior to the comparison, the external dataset was regridded to a 5° x 5° regular grid by means of an arithmetic unweighted averaging process.\n\nThroughout all the considered years, regions within the 10$^\\circ\\,$S–40$^\\circ\\,$N latitude band exhibit positive XCO$_2$ anomalies, consistent with the dataset produced by [[5]](https://doi.org/10.1016/j.envint.2023.108057). Both datasets reported similar features in the interannual variability of XCO$_2$ anomalies, like the higher values over China in 2020 compared to 2019, positive anomalies over Southeast Asia in 2016, and over tropical Africa in 2016, 2017 and 2018.\n\nFocusing on the yearly bias (obtained by subtracting the Zhang et al. [[5]](https://doi.org/10.1016/j.envint.2023.108057) anomalies from the Obs4MIPs anomalies), the absolute differences in the yearly anomalies were mostly related to the high latitudes and the tropical regions. Please note that the horizontal coverage of the XCO2_OBS4MIPS dataset (version 4.5) extends approximately from 60$^\\circ\\,$S to 70$^\\circ\\,$N [[6]](https://amt.copernicus.org/articles/13/789/2020/). Even though the XCO2_OBS4MIPS dataset was subsetted in this exercise to exclude high latitudes (i.e., higher than 60°N and 60°S), comparisons with the gap-free external dataset by Zhang et al. [[5]](https://doi.org/10.1016/j.envint.2023.108057) should be treated with extreme caution at latitudes above 50°N. This is due to non-uniform data coverage across different seasons for Obs4MIPs (version 4.5) at high latitudes, which can lead to sampling bias when data is aggregated on a yearly timescale. During the investigated period, the external dataset [[5]](https://doi.org/10.1016/j.envint.2023.108057) provided XCO$_2$ also for Greenland, a region over which Obs4MIPs (version 4.5) did not provide data. This is due to the high surface albedo, which hinders the reliability of satellite retrievals, even during periods of illumination.\n\nMoreover, the dataset [[5]](https://doi.org/10.1016/j.envint.2023.108057) appears to provide a smoother spatial field of XCO$_2$ anomalies with respect to Obs4MIPs (version 4.5), probably due to the use of a learning machine algorithm to produce this gap filled dataset.\n\nUnisci tutti e 3 i set\nStack delle dimensioni (year, product) in \"panel\"\nAssegna etichette leggibili ai pannelli\n\n*The figure shows the multi-year global distribution of XCO$_2$ anomalies (from 2015 to 2020) derived from the XCO2_OBS4MIPS dataset (version 4.5) as well as for the gap-filled dataset produced by Zhang et al. (2023) and the bias (Obs4MIPs - Zhang et al.) between the two datasets. Please be aware that the data displayed here includes only locations with latitudes below 60°N and 60°S.*", "text_with_prefix": "EQC Quality Assessment: \"Spatial and temporal completeness and uncertainties of carbon dioxide satellite observations for quantifying atmospheric greenhouse gas variability\"\nDataset: satellite-carbon-dioxide [CDS]\nAspect: uncertainty-quality-flags_q02 | Category: Satellite_ECVs\nSection: Spatial and temporal completeness and uncertainties of carbon dioxide satellite observations for quantifying atmospheric greenhouse gas variability > Analysis and results > 3. Data analysis for the different variables > Global maps of XCO$_2$ anomalies\n---\nIn this section, global maps of XCO$_2$ anomalies for the years 2015–2020 are presented for the latitudes spanning from 60°N to 60°S. The calculation of anomalies is achieved by the subtraction of the annual global average from the global XCO$_2$ concentrations, in accordance with the approach outlined in [[5]](https://doi.org/10.1016/j.envint.2023.108057). Prior to the execution of the calculations, a land mask was implemented. A comparison was made between the XCO$_2$ anomalies from the Obs4MIPs (version 4.5) dataset and the results from the gap-free and fine-scale XCO$_2$ dataset created by [[5]](https://doi.org/10.1016/j.envint.2023.108057). Prior to the comparison, the external dataset was regridded to a 5° x 5° regular grid by means of an arithmetic unweighted averaging process.\n\nThroughout all the considered years, regions within the 10$^\\circ\\,$S–40$^\\circ\\,$N latitude band exhibit positive XCO$_2$ anomalies, consistent with the dataset produced by [[5]](https://doi.org/10.1016/j.envint.2023.108057). Both datasets reported similar features in the interannual variability of XCO$_2$ anomalies, like the higher values over China in 2020 compared to 2019, positive anomalies over Southeast Asia in 2016, and over tropical Africa in 2016, 2017 and 2018.\n\nFocusing on the yearly bias (obtained by subtracting the Zhang et al. [[5]](https://doi.org/10.1016/j.envint.2023.108057) anomalies from the Obs4MIPs anomalies), the absolute differences in the yearly anomalies were mostly related to the high latitudes and the tropical regions. Please note that the horizontal coverage of the XCO2_OBS4MIPS dataset (version 4.5) extends approximately from 60$^\\circ\\,$S to 70$^\\circ\\,$N [[6]](https://amt.copernicus.org/articles/13/789/2020/). Even though the XCO2_OBS4MIPS dataset was subsetted in this exercise to exclude high latitudes (i.e., higher than 60°N and 60°S), comparisons with the gap-free external dataset by Zhang et al. [[5]](https://doi.org/10.1016/j.envint.2023.108057) should be treated with extreme caution at latitudes above 50°N. This is due to non-uniform data coverage across different seasons for Obs4MIPs (version 4.5) at high latitudes, which can lead to sampling bias when data is aggregated on a yearly timescale. During the investigated period, the external dataset [[5]](https://doi.org/10.1016/j.envint.2023.108057) provided XCO$_2$ also for Greenland, a region over which Obs4MIPs (version 4.5) did not provide data. This is due to the high surface albedo, which hinders the reliability of satellite retrievals, even during periods of illumination.\n\nMoreover, the dataset [[5]](https://doi.org/10.1016/j.envint.2023.108057) appears to provide a smoother spatial field of XCO$_2$ anomalies with respect to Obs4MIPs (version 4.5), probably due to the use of a learning machine algorithm to produce this gap filled dataset.\n\nUnisci tutti e 3 i set\nStack delle dimensioni (year, product) in \"panel\"\nAssegna etichette leggibili ai pannelli\n\n*The figure shows the multi-year global distribution of XCO$_2$ anomalies (from 2015 to 2020) derived from the XCO2_OBS4MIPS dataset (version 4.5) as well as for the gap-filled dataset produced by Zhang et al. (2023) and the bias (Obs4MIPs - Zhang et al.) between the two datasets. Please be aware that the data displayed here includes only locations with latitudes below 60°N and 60°S.*"} {"chunk_id": "satellite_satellite-carbon-dioxide_uncertainty-quality-flags_q02__e3145475966c", "report_id": "satellite_satellite-carbon-dioxide_uncertainty-quality-flags_q02", "dataset_id": "satellite-carbon-dioxide", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty-quality-flags_q02", "aspect_base": "uncertainty-quality-flags", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Spatial and temporal completeness and uncertainties of carbon dioxide satellite observations for quantifying atmospheric greenhouse gas variability > Analysis and results > 3. Data analysis for the different variables > Global maps of XCO$_2$ uncertainties", "title": "Spatial and temporal completeness and uncertainties of carbon dioxide satellite observations for quantifying atmospheric greenhouse gas variability", "chunk_index": 9, "token_count": 399, "text_raw": "In this section, we show the multi-year global maps of the XCO$_2$ reported uncertainties, i.e., the \"xco2_stderr\" variable. This parameter is defined as the standard error of the average including single sounding noise and potential seasonal and regional biases [[3]](https://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PUGS_GHG_main_latest.pdf). The yearly averages are calculated by averaging the monthly values provided by the considered dataset. Throughout all the considered years some features emerge, with higher uncertainties over high latitudes, the Himalayas, and the tropical rainforest zone, which are the regions where we observed larger deviations between Obs4MIPs (version 4.5) and the external dataset [[5]](https://doi.org/10.1016/j.envint.2023.108057). This is mainly due to the sparseness in sampling because of frequent cloud cover, as well as large solar zenith angles in high latitudes, which is a challenge for accurate XCO$_2$ retrievals [[6]](https://amt.copernicus.org/articles/13/789/2020/).\n\n*The figure shows the multi-year global distribution of XCO$_2$ standard error (from 2015 to 2020), derived from the XCO2_OBS4MIPS dataset (version 4.5).*", "text_with_prefix": "EQC Quality Assessment: \"Spatial and temporal completeness and uncertainties of carbon dioxide satellite observations for quantifying atmospheric greenhouse gas variability\"\nDataset: satellite-carbon-dioxide [CDS]\nAspect: uncertainty-quality-flags_q02 | Category: Satellite_ECVs\nSection: Spatial and temporal completeness and uncertainties of carbon dioxide satellite observations for quantifying atmospheric greenhouse gas variability > Analysis and results > 3. Data analysis for the different variables > Global maps of XCO$_2$ uncertainties\n---\nIn this section, we show the multi-year global maps of the XCO$_2$ reported uncertainties, i.e., the \"xco2_stderr\" variable. This parameter is defined as the standard error of the average including single sounding noise and potential seasonal and regional biases [[3]](https://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PUGS_GHG_main_latest.pdf). The yearly averages are calculated by averaging the monthly values provided by the considered dataset. Throughout all the considered years some features emerge, with higher uncertainties over high latitudes, the Himalayas, and the tropical rainforest zone, which are the regions where we observed larger deviations between Obs4MIPs (version 4.5) and the external dataset [[5]](https://doi.org/10.1016/j.envint.2023.108057). This is mainly due to the sparseness in sampling because of frequent cloud cover, as well as large solar zenith angles in high latitudes, which is a challenge for accurate XCO$_2$ retrievals [[6]](https://amt.copernicus.org/articles/13/789/2020/).\n\n*The figure shows the multi-year global distribution of XCO$_2$ standard error (from 2015 to 2020), derived from the XCO2_OBS4MIPS dataset (version 4.5).*"} {"chunk_id": "satellite_satellite-carbon-dioxide_uncertainty-quality-flags_q02__fed4f0bd07b3", "report_id": "satellite_satellite-carbon-dioxide_uncertainty-quality-flags_q02", "dataset_id": "satellite-carbon-dioxide", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty-quality-flags_q02", "aspect_base": "uncertainty-quality-flags", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Spatial and temporal completeness and uncertainties of carbon dioxide satellite observations for quantifying atmospheric greenhouse gas variability > Analysis and results > 3. Data analysis for the different variables > Regional maps of XCO$_2$", "title": "Spatial and temporal completeness and uncertainties of carbon dioxide satellite observations for quantifying atmospheric greenhouse gas variability", "chunk_index": 10, "token_count": 356, "text_raw": "In this section, we show the maps of the XCO$_2$ values for January 2016, both globally and over specific regions. The selected regions (i.e., North America, Europe/Africa, and East Asia) were chosen according to [[5]](https://doi.org/10.1016/j.envint.2023.108057). A direct comparison for January 2015 was not possible, as the Level 3 XCO$_2$ data are missing for that month in the XCO2_OBS4MIPS dataset (version 4.5). The XCO$_2$ patterns are similar and comparable, especially for Europe/Africa and East Asia. However, the Level 3 data cover lesser areas than the product in [[5]](https://doi.org/10.1016/j.envint.2023.108057), with missing data especially at latitudes above 50$^\\circ\\,$N.\n\nTo select specific regions\nfig.suptitle(f\"{variable =} {time = }\")\n\n*The figure shows the spatial distribution of XCO$_2$ values over the globe and selected regions, derived from the XCO2_OBS4MIPS dataset (version 4.5), for January 2016.*", "text_with_prefix": "EQC Quality Assessment: \"Spatial and temporal completeness and uncertainties of carbon dioxide satellite observations for quantifying atmospheric greenhouse gas variability\"\nDataset: satellite-carbon-dioxide [CDS]\nAspect: uncertainty-quality-flags_q02 | Category: Satellite_ECVs\nSection: Spatial and temporal completeness and uncertainties of carbon dioxide satellite observations for quantifying atmospheric greenhouse gas variability > Analysis and results > 3. Data analysis for the different variables > Regional maps of XCO$_2$\n---\nIn this section, we show the maps of the XCO$_2$ values for January 2016, both globally and over specific regions. The selected regions (i.e., North America, Europe/Africa, and East Asia) were chosen according to [[5]](https://doi.org/10.1016/j.envint.2023.108057). A direct comparison for January 2015 was not possible, as the Level 3 XCO$_2$ data are missing for that month in the XCO2_OBS4MIPS dataset (version 4.5). The XCO$_2$ patterns are similar and comparable, especially for Europe/Africa and East Asia. However, the Level 3 data cover lesser areas than the product in [[5]](https://doi.org/10.1016/j.envint.2023.108057), with missing data especially at latitudes above 50$^\\circ\\,$N.\n\nTo select specific regions\nfig.suptitle(f\"{variable =} {time = }\")\n\n*The figure shows the spatial distribution of XCO$_2$ values over the globe and selected regions, derived from the XCO2_OBS4MIPS dataset (version 4.5), for January 2016.*"} {"chunk_id": "satellite_satellite-carbon-dioxide_uncertainty-quality-flags_q02__1f13ac0caa9b", "report_id": "satellite_satellite-carbon-dioxide_uncertainty-quality-flags_q02", "dataset_id": "satellite-carbon-dioxide", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty-quality-flags_q02", "aspect_base": "uncertainty-quality-flags", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Spatial and temporal completeness and uncertainties of carbon dioxide satellite observations for quantifying atmospheric greenhouse gas variability > Analysis and results > 3. Data analysis for the different variables > Time series analysis for the different variables", "title": "Spatial and temporal completeness and uncertainties of carbon dioxide satellite observations for quantifying atmospheric greenhouse gas variability", "chunk_index": 11, "token_count": 928, "text_raw": "In this section, we show the time series for two XCO$_2$ variables (i.e., \"xco2\" and \"xco2_stderr\"), considering both the entire dataset and specific regions of interest. For each variable, each plot displays the monthly spatial average for each region, along with the corresponding monthly standard deviation.\nThis analysis would provide the user with:\n* the time series of XCO$_2$ values (\"xco2\") for each region, highlighting the positive trends;\n* the regional change over time of the uncertainties associated to the XCO$_2$ values (\"xco2_stderr\").\n* the time series of \"xco2_nobs\", which denotes the number of individual XCO$_2$ Level 2 observations used to compute the monthly Level 3 data.\n\nIn particular, an increase in the uncertainties (\"xco2_stderr\") is observed after 2009, likely due to the increase in the number of algorithms used to calculate the median XCO$_2$ data [[7]](http://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_ATBD_GHG_D_latest.pdf). A further increase in the uncertainties is observed from 2020 onwards, and this is probably linked to both the increase in the number of algorithms and the change in the minimum number of algorithms required to calculate the median XCO$_2$ [[7]](http://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_ATBD_GHG_D_latest.pdf). In addition, regions with more difficult retrieval conditions due to frequent cloud cover (e.g., East Asia) show larger uncertainties due to larger inter-algorithm spreads [[6]](https://amt.copernicus.org/articles/13/789/2020/).\n\nThe temporal variability of \"xco2_nobs\" traces the changes over time of the input data availability, and the number of used algorithms to obtain the merged Level 2 data products (XCO2_EMMA), from which the Obs4MIPs data product is derived (please note that the standard deviations have not been reported for \"xco2_nobs\" in order to increase the plot's readability.).\nThe first time period (up to April 2009) is characterised by a significant number of soundings by the SCIAMACHY sensor. The subsequent period (2009-September 2014) reflects the reduced number of single soundings provided by SCIAMACHY with the contribution from GOSAT. The following period, which extends up to 2019, reflects the contributions from both GOSAT and OCO-2. In contrast, the final period of the dataset reflects the contributions from GOSAT, OCO-2, and GOSAT-2. A comprehensive analysis of the number of soundings contained within the EMMA database on a monthly basis, along with the comparative contribution of each individual Level 2 algorithm to the EMMA database, can be found in [[7]](http://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_ATBD_GHG_D_latest.pdf).\n\nTo select specific regions\nIf variaable is xco2_nobs, use log scale for y-axis\nax.set_ylabel(\"n observations\")\nImposta tick logaritmici automatici\n\n*The figure shows the monthly time series of XCO$_2$ (top panels) and their associated uncertainties (middle panels) for the entire dataset, including global and regional analyses. The blue lines represent the monthly spatial average, while the shaded areas indicate $\\pm$1 standard deviation. The bottom panels shows the time series of monthly spatial average of the number of individual XCO$_2$ Level 2 observations used to compute Obs4MIPs (Level 3) data. The main titles indicate the variable shown, while the subtitles specify the corresponding regions.*", "text_with_prefix": "EQC Quality Assessment: \"Spatial and temporal completeness and uncertainties of carbon dioxide satellite observations for quantifying atmospheric greenhouse gas variability\"\nDataset: satellite-carbon-dioxide [CDS]\nAspect: uncertainty-quality-flags_q02 | Category: Satellite_ECVs\nSection: Spatial and temporal completeness and uncertainties of carbon dioxide satellite observations for quantifying atmospheric greenhouse gas variability > Analysis and results > 3. Data analysis for the different variables > Time series analysis for the different variables\n---\nIn this section, we show the time series for two XCO$_2$ variables (i.e., \"xco2\" and \"xco2_stderr\"), considering both the entire dataset and specific regions of interest. For each variable, each plot displays the monthly spatial average for each region, along with the corresponding monthly standard deviation.\nThis analysis would provide the user with:\n* the time series of XCO$_2$ values (\"xco2\") for each region, highlighting the positive trends;\n* the regional change over time of the uncertainties associated to the XCO$_2$ values (\"xco2_stderr\").\n* the time series of \"xco2_nobs\", which denotes the number of individual XCO$_2$ Level 2 observations used to compute the monthly Level 3 data.\n\nIn particular, an increase in the uncertainties (\"xco2_stderr\") is observed after 2009, likely due to the increase in the number of algorithms used to calculate the median XCO$_2$ data [[7]](http://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_ATBD_GHG_D_latest.pdf). A further increase in the uncertainties is observed from 2020 onwards, and this is probably linked to both the increase in the number of algorithms and the change in the minimum number of algorithms required to calculate the median XCO$_2$ [[7]](http://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_ATBD_GHG_D_latest.pdf). In addition, regions with more difficult retrieval conditions due to frequent cloud cover (e.g., East Asia) show larger uncertainties due to larger inter-algorithm spreads [[6]](https://amt.copernicus.org/articles/13/789/2020/).\n\nThe temporal variability of \"xco2_nobs\" traces the changes over time of the input data availability, and the number of used algorithms to obtain the merged Level 2 data products (XCO2_EMMA), from which the Obs4MIPs data product is derived (please note that the standard deviations have not been reported for \"xco2_nobs\" in order to increase the plot's readability.).\nThe first time period (up to April 2009) is characterised by a significant number of soundings by the SCIAMACHY sensor. The subsequent period (2009-September 2014) reflects the reduced number of single soundings provided by SCIAMACHY with the contribution from GOSAT. The following period, which extends up to 2019, reflects the contributions from both GOSAT and OCO-2. In contrast, the final period of the dataset reflects the contributions from GOSAT, OCO-2, and GOSAT-2. A comprehensive analysis of the number of soundings contained within the EMMA database on a monthly basis, along with the comparative contribution of each individual Level 2 algorithm to the EMMA database, can be found in [[7]](http://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_ATBD_GHG_D_latest.pdf).\n\nTo select specific regions\nIf variaable is xco2_nobs, use log scale for y-axis\nax.set_ylabel(\"n observations\")\nImposta tick logaritmici automatici\n\n*The figure shows the monthly time series of XCO$_2$ (top panels) and their associated uncertainties (middle panels) for the entire dataset, including global and regional analyses. The blue lines represent the monthly spatial average, while the shaded areas indicate $\\pm$1 standard deviation. The bottom panels shows the time series of monthly spatial average of the number of individual XCO$_2$ Level 2 observations used to compute Obs4MIPs (Level 3) data. The main titles indicate the variable shown, while the subtitles specify the corresponding regions.*"} {"chunk_id": "satellite_satellite-carbon-dioxide_uncertainty-quality-flags_q02__d1c9b9b4a177", "report_id": "satellite_satellite-carbon-dioxide_uncertainty-quality-flags_q02", "dataset_id": "satellite-carbon-dioxide", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty-quality-flags_q02", "aspect_base": "uncertainty-quality-flags", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Spatial and temporal completeness and uncertainties of carbon dioxide satellite observations for quantifying atmospheric greenhouse gas variability > ℹ️ If you want to know more > Key resources", "title": "Spatial and temporal completeness and uncertainties of carbon dioxide satellite observations for quantifying atmospheric greenhouse gas variability", "chunk_index": 12, "token_count": 272, "text_raw": "The CDS catalogue entries for the data used were:\n* Carbon dioxide data from 2002 to present derived from satellite observations: https://cds.climate.copernicus.eu/datasets/satellite-carbon-dioxide?tab=overview\n\nCode libraries used:\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)\n\nUsers interested in the investigation of CO$_2$ fluxes can consider to use specifically designed Copernicus resources like [CAMS global inversion-optimised greenhouse gas fluxes and concentrations](https://ads.atmosphere.copernicus.eu/datasets/cams-global-greenhouse-gas-inversion?tab=overview).", "text_with_prefix": "EQC Quality Assessment: \"Spatial and temporal completeness and uncertainties of carbon dioxide satellite observations for quantifying atmospheric greenhouse gas variability\"\nDataset: satellite-carbon-dioxide [CDS]\nAspect: uncertainty-quality-flags_q02 | Category: Satellite_ECVs\nSection: Spatial and temporal completeness and uncertainties of carbon dioxide satellite observations for quantifying atmospheric greenhouse gas variability > ℹ️ If you want to know more > Key resources\n---\nThe CDS catalogue entries for the data used were:\n* Carbon dioxide data from 2002 to present derived from satellite observations: https://cds.climate.copernicus.eu/datasets/satellite-carbon-dioxide?tab=overview\n\nCode libraries used:\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)\n\nUsers interested in the investigation of CO$_2$ fluxes can consider to use specifically designed Copernicus resources like [CAMS global inversion-optimised greenhouse gas fluxes and concentrations](https://ads.atmosphere.copernicus.eu/datasets/cams-global-greenhouse-gas-inversion?tab=overview)."} {"chunk_id": "satellite_satellite-carbon-dioxide_uncertainty-quality-flags_q02__9d68468d4277", "report_id": "satellite_satellite-carbon-dioxide_uncertainty-quality-flags_q02", "dataset_id": "satellite-carbon-dioxide", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty-quality-flags_q02", "aspect_base": "uncertainty-quality-flags", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Spatial and temporal completeness and uncertainties of carbon dioxide satellite observations for quantifying atmospheric greenhouse gas variability > ℹ️ If you want to know more > References", "title": "Spatial and temporal completeness and uncertainties of carbon dioxide satellite observations for quantifying atmospheric greenhouse gas variability", "chunk_index": 13, "token_count": 1003, "text_raw": "[[1]](https://library.wmo.int/idurl/4/69057) World Meteorological Organization. (2024). WMO Greenhouse Gas Bulletin, 20, ISSN 2078-0796.\n\n[[2]](https://doi.org/10.5194/essd-17-965-2025) Friedlingstein, P., O'Sullivan, M., Jones, M. W., Andrew, R. M., Hauck, J., Landschützer, P., Le Quéré, C., Li, H., Luijkx, I. T., Olsen, A., Peters, G. P., Peters, W., Pongratz, J., Schwingshackl, C., Sitch, S., Canadell, J. G., Ciais, P., Jackson, R. B., Alin, S. R., Arneth, A., Arora, V., Bates, N. R., Becker, M., Bellouin, N., Berghoff, C. F., Bittig, H. C., Bopp, L., Cadule, P., Campbell, K., Chamberlain, M. A., Chandra, N., Chevallier, F., Chini, L. P., Colligan, T., Decayeux, J., Djeutchouang, L. M., Dou, X., Duran Rojas, C., Enyo, K., Evans, W., Fay, A. R., Feely, R. A., Ford, D. J., Foster, A., Gasser, T., Gehlen, M., Gkritzalis, T., Grassi, G., Gregor, L., Gruber, N., Gürses, Ö., Harris, I., Hefner, M., Heinke, J., Hurtt, G. C., Iida, Y., Ilyina, T., Jacobson, A. R., Jain, A. K., Jarníková, T., Jersild, A., Jiang, F., Jin, Z., Kato, E., Keeling, R. F., Klein Goldewijk, K., Knauer, J., Korsbakken, J. I., Lan, X., Lauvset, S. K., Lefèvre, N., Liu, Z., Liu, J., Ma, L., Maksyutov, S., Marland, G., Mayot, N., McGuire, P. C., Metzl, N., Monacci, N. M., Morgan, E. J., Nakaoka, S.-I., Neill, C., Niwa, Y., Nützel, T., Olivier, L., Ono, T., Palmer, P. I., Pierrot, D., Qin, Z., Resplandy, L., Roobaert, A., Rosan, T. M., Rödenbeck, C., Schwinger, J., Smallman, T. L., Smith, S. M., Sospedra-Alfonso, R., Steinhoff, T., Sun, Q., Sutton, A. J., Séférian, R., Takao, S., Tatebe, H., Tian, H., Tilbrook, B., Torres, O., Tourigny, E., Tsujino, H., Tubiello, F., van der Werf, G., Wanninkhof, R., Wang, X., Yang, D., Yang, X., Yu, Z., Yuan, W., Yue, X., Zaehle, S., Zeng, N., and Zeng, J. (2025). Global Carbon Budget 2024, Earth System Science Data, 17, 965-1039.\n\n[[3]](https://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PUGS_GHG_main_latest.pdf) Buchwitz, M. (2024). Product User Guide and Specification (PUGS) – Main document for Greenhouse Gas (GHG: CO$_2$ & CH$_4$) data set CDR7 (01.2003-12.2022), C3S project 2021/C3S2_312a_Lot2_DLR/SC1, v7.3.", "text_with_prefix": "EQC Quality Assessment: \"Spatial and temporal completeness and uncertainties of carbon dioxide satellite observations for quantifying atmospheric greenhouse gas variability\"\nDataset: satellite-carbon-dioxide [CDS]\nAspect: uncertainty-quality-flags_q02 | Category: Satellite_ECVs\nSection: Spatial and temporal completeness and uncertainties of carbon dioxide satellite observations for quantifying atmospheric greenhouse gas variability > ℹ️ If you want to know more > References\n---\n[[1]](https://library.wmo.int/idurl/4/69057) World Meteorological Organization. (2024). WMO Greenhouse Gas Bulletin, 20, ISSN 2078-0796.\n\n[[2]](https://doi.org/10.5194/essd-17-965-2025) Friedlingstein, P., O'Sullivan, M., Jones, M. W., Andrew, R. M., Hauck, J., Landschützer, P., Le Quéré, C., Li, H., Luijkx, I. T., Olsen, A., Peters, G. P., Peters, W., Pongratz, J., Schwingshackl, C., Sitch, S., Canadell, J. G., Ciais, P., Jackson, R. B., Alin, S. R., Arneth, A., Arora, V., Bates, N. R., Becker, M., Bellouin, N., Berghoff, C. F., Bittig, H. C., Bopp, L., Cadule, P., Campbell, K., Chamberlain, M. A., Chandra, N., Chevallier, F., Chini, L. P., Colligan, T., Decayeux, J., Djeutchouang, L. M., Dou, X., Duran Rojas, C., Enyo, K., Evans, W., Fay, A. R., Feely, R. A., Ford, D. J., Foster, A., Gasser, T., Gehlen, M., Gkritzalis, T., Grassi, G., Gregor, L., Gruber, N., Gürses, Ö., Harris, I., Hefner, M., Heinke, J., Hurtt, G. C., Iida, Y., Ilyina, T., Jacobson, A. R., Jain, A. K., Jarníková, T., Jersild, A., Jiang, F., Jin, Z., Kato, E., Keeling, R. F., Klein Goldewijk, K., Knauer, J., Korsbakken, J. I., Lan, X., Lauvset, S. K., Lefèvre, N., Liu, Z., Liu, J., Ma, L., Maksyutov, S., Marland, G., Mayot, N., McGuire, P. C., Metzl, N., Monacci, N. M., Morgan, E. J., Nakaoka, S.-I., Neill, C., Niwa, Y., Nützel, T., Olivier, L., Ono, T., Palmer, P. I., Pierrot, D., Qin, Z., Resplandy, L., Roobaert, A., Rosan, T. M., Rödenbeck, C., Schwinger, J., Smallman, T. L., Smith, S. M., Sospedra-Alfonso, R., Steinhoff, T., Sun, Q., Sutton, A. J., Séférian, R., Takao, S., Tatebe, H., Tian, H., Tilbrook, B., Torres, O., Tourigny, E., Tsujino, H., Tubiello, F., van der Werf, G., Wanninkhof, R., Wang, X., Yang, D., Yang, X., Yu, Z., Yuan, W., Yue, X., Zaehle, S., Zeng, N., and Zeng, J. (2025). Global Carbon Budget 2024, Earth System Science Data, 17, 965-1039.\n\n[[3]](https://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PUGS_GHG_main_latest.pdf) Buchwitz, M. (2024). Product User Guide and Specification (PUGS) – Main document for Greenhouse Gas (GHG: CO$_2$ & CH$_4$) data set CDR7 (01.2003-12.2022), C3S project 2021/C3S2_312a_Lot2_DLR/SC1, v7.3."} {"chunk_id": "satellite_satellite-carbon-dioxide_uncertainty-quality-flags_q02__855b2cf978bb", "report_id": "satellite_satellite-carbon-dioxide_uncertainty-quality-flags_q02", "dataset_id": "satellite-carbon-dioxide", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty-quality-flags_q02", "aspect_base": "uncertainty-quality-flags", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Spatial and temporal completeness and uncertainties of carbon dioxide satellite observations for quantifying atmospheric greenhouse gas variability > ℹ️ If you want to know more > References", "title": "Spatial and temporal completeness and uncertainties of carbon dioxide satellite observations for quantifying atmospheric greenhouse gas variability", "chunk_index": 14, "token_count": 773, "text_raw": ") data set CDR7 (01.2003-12.2022), C3S project 2021/C3S2_312a_Lot2_DLR/SC1, v7.3.\n\n[[4]](https://doi.org/10.5194/bg-17-6115-2020) Gier, B. K., Buchwitz, M., Reuter, M., Cox, P. M., Friedlingstein, P., and Eyring, V. (2020). Spatially resolved evaluation of Earth system models with satellite column-averaged CO$_2$, Biogeosciences, 17, 6115–6144.\n\n[[5]](https://doi.org/10.1016/j.envint.2023.108057) Zhang, L., Li, T., Wu, J., and Yang, H. (2023). Global estimates of gap-free and fine-scale CO$_2$ concentrations during 2014-2020 from satellite and reanalysis data, Environment International, 178, 108057.\n\n[[6]](https://amt.copernicus.org/articles/13/789/2020/) Reuter, M., Buchwitz, M., Schneising, O., Noël, S., Bovensmann, H., Burrows, J. P., Boesch, H., Di Noia, A., Anand, J., Parker, R. J., Somkuti, P., Wu, L., Hasekamp, O. P., Aben, I., Kuze, A., Suto, H., Shiomi, K., Yoshida, Y., Morino, I., Crisp, D., O'Dell, C. W., Notholt, J., Petri, C., Warneke, T., Velazco, V. A., Deutscher, N. M., Griffith, D. W. T., Kivi, R., Pollard, D. F., Hase, F., Sussmann, R., Té, Y. V., Strong, K., Roche, S., Sha, M. K., De Mazière, M., Feist, D. G., Iraci, L. T., Roehl, C. M., Retscher, C., and Schepers, D. (2020). Ensemble-based satellite-derived carbon dioxide and methane column-averaged dry-air mole fraction data sets (2003–2018) for carbon and climate applications, Atmospheric Measurement Techniques, 13, 789–819.\n\n[[7]](http://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_ATBD_GHG_D_latest.pdf) Reuter, M., and Buchwitz, M. (2024). Algorithm Theoretical Basis Document (ATBD) – ANNEX D for products XCO2_EMMA, XCH4_EMMA, XCO2_OBS4MIPS, XCH4_OBS4MIPS (v4.5, CDR7, 2003-2022), C3S project 2021/C3S2_312a_Lot2_DLR/SC1, v7.1b.", "text_with_prefix": "EQC Quality Assessment: \"Spatial and temporal completeness and uncertainties of carbon dioxide satellite observations for quantifying atmospheric greenhouse gas variability\"\nDataset: satellite-carbon-dioxide [CDS]\nAspect: uncertainty-quality-flags_q02 | Category: Satellite_ECVs\nSection: Spatial and temporal completeness and uncertainties of carbon dioxide satellite observations for quantifying atmospheric greenhouse gas variability > ℹ️ If you want to know more > References\n---\n) data set CDR7 (01.2003-12.2022), C3S project 2021/C3S2_312a_Lot2_DLR/SC1, v7.3.\n\n[[4]](https://doi.org/10.5194/bg-17-6115-2020) Gier, B. K., Buchwitz, M., Reuter, M., Cox, P. M., Friedlingstein, P., and Eyring, V. (2020). Spatially resolved evaluation of Earth system models with satellite column-averaged CO$_2$, Biogeosciences, 17, 6115–6144.\n\n[[5]](https://doi.org/10.1016/j.envint.2023.108057) Zhang, L., Li, T., Wu, J., and Yang, H. (2023). Global estimates of gap-free and fine-scale CO$_2$ concentrations during 2014-2020 from satellite and reanalysis data, Environment International, 178, 108057.\n\n[[6]](https://amt.copernicus.org/articles/13/789/2020/) Reuter, M., Buchwitz, M., Schneising, O., Noël, S., Bovensmann, H., Burrows, J. P., Boesch, H., Di Noia, A., Anand, J., Parker, R. J., Somkuti, P., Wu, L., Hasekamp, O. P., Aben, I., Kuze, A., Suto, H., Shiomi, K., Yoshida, Y., Morino, I., Crisp, D., O'Dell, C. W., Notholt, J., Petri, C., Warneke, T., Velazco, V. A., Deutscher, N. M., Griffith, D. W. T., Kivi, R., Pollard, D. F., Hase, F., Sussmann, R., Té, Y. V., Strong, K., Roche, S., Sha, M. K., De Mazière, M., Feist, D. G., Iraci, L. T., Roehl, C. M., Retscher, C., and Schepers, D. (2020). Ensemble-based satellite-derived carbon dioxide and methane column-averaged dry-air mole fraction data sets (2003–2018) for carbon and climate applications, Atmospheric Measurement Techniques, 13, 789–819.\n\n[[7]](http://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_ATBD_GHG_D_latest.pdf) Reuter, M., and Buchwitz, M. (2024). Algorithm Theoretical Basis Document (ATBD) – ANNEX D for products XCO2_EMMA, XCH4_EMMA, XCO2_OBS4MIPS, XCH4_OBS4MIPS (v4.5, CDR7, 2003-2022), C3S project 2021/C3S2_312a_Lot2_DLR/SC1, v7.1b."} {"chunk_id": "satellite_satellite-methane_completeness_q03__926745e54c55", "report_id": "satellite_satellite-methane_completeness_q03", "dataset_id": "satellite-methane", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q03", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Methane satellite observations uncertainty assessment for greenhouse gas monitoring", "title": "Methane satellite observations uncertainty assessment for greenhouse gas monitoring", "chunk_index": 0, "token_count": 89, "text_raw": "Production date: 21-11-2025\n\nProduced by: Davide Putero (CNR), Paolo Cristofanelli (CNR)", "text_with_prefix": "EQC Quality Assessment: \"Methane satellite observations uncertainty assessment for greenhouse gas monitoring\"\nDataset: satellite-methane [CDS]\nAspect: completeness_q03 | Category: Satellite_ECVs\nSection: Methane satellite observations uncertainty assessment for greenhouse gas monitoring\n---\nProduction date: 21-11-2025\n\nProduced by: Davide Putero (CNR), Paolo Cristofanelli (CNR)"} {"chunk_id": "satellite_satellite-methane_completeness_q03__e70878e31b9d", "report_id": "satellite_satellite-methane_completeness_q03", "dataset_id": "satellite-methane", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q03", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Methane satellite observations uncertainty assessment for greenhouse gas monitoring > Quality assessment questions", "title": "Methane satellite observations uncertainty assessment for greenhouse gas monitoring", "chunk_index": 1, "token_count": 745, "text_raw": "* **How can the variability of spatial and temporal data coverage affect the quantification of the long-term atmospheric methane trends by satellite measurements (XCH$_4$ Level 3 gridded product)?**\n* **What are the uncertainties related to the XCH$_4$ satellite observations?**\n\nMethane (CH$_4$) is the second most important anthropogenic greenhouse gas after carbon dioxide (CO$_2$), representing about 19% of the total radiative forcing by long-lived greenhouse gases [[1]](https://library.wmo.int/idurl/4/68532). Atmospheric CH$_4$ also adversely affects human health as a precursor of tropospheric ozone [[2]](https://doi.org/10.1073/pnas.0600201103). Monitoring the long-term CH$_4$ variability is therefore crucial for monitoring the emission reductions [[3]](https://doi.org/10.5194/essd-17-1873-2025). In this assessment, atmospheric CH$_4$ spatial seasonal means and trends are analysed using the XCH$_4$ Level 3 gridded product (Obs4MIPs, version 4.5 [[4]](https://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PUGS_GHG_main_latest.pdf)), by adopting an approach similar to [[5]](https://doi.org/10.5194/amt-13-789-2020). Furthermore, we also evaluate the uncertainty associated with this dataset.\n\nThe Obs4MIPs XCH$_4$ product is generated by spatial (5°x5°) and temporal (monthly) gridding of the corresponding EMMA Level 2 data product. For more details see [[6]](https://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_ATBD_GHG_D_latest.pdf). By analysing an ensemble of different Level 2 datasets from various retrieval algorithms, the EMMA algorithm generates a dataset containing XCH$_4$ from individual retrievals. In particular, for each month and 10°×10° grid box, the algorithm with the grid box mean closest to the median is selected. The list of the considered retrieval algorithms is available from [[6]](https://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_ATBD_GHG_D_latest.pdf). The various retrieval algorithms are optimised for different instruments (SCIAMACHY, GOSAT, GOSAT-2) which measure backscattered solar radiation in the near-infrared for O$_2$ as well as the absorption bands of CO$_2$ or CH$_4$.\n\nCode is included for transparency but also learning purposes and gives users the chance to adapt the code used for the assesment as they wish. Users should always check product documentation and associated peer-reviewed papers for a more complete reporting of the issues discussed here (e.g., [[7]](https://doi.org/10.24381/14j9-s541), [[8]](https://climate.copernicus.eu/global-climate-highlights-2024)).", "text_with_prefix": "EQC Quality Assessment: \"Methane satellite observations uncertainty assessment for greenhouse gas monitoring\"\nDataset: satellite-methane [CDS]\nAspect: completeness_q03 | Category: Satellite_ECVs\nSection: Methane satellite observations uncertainty assessment for greenhouse gas monitoring > Quality assessment questions\n---\n* **How can the variability of spatial and temporal data coverage affect the quantification of the long-term atmospheric methane trends by satellite measurements (XCH$_4$ Level 3 gridded product)?**\n* **What are the uncertainties related to the XCH$_4$ satellite observations?**\n\nMethane (CH$_4$) is the second most important anthropogenic greenhouse gas after carbon dioxide (CO$_2$), representing about 19% of the total radiative forcing by long-lived greenhouse gases [[1]](https://library.wmo.int/idurl/4/68532). Atmospheric CH$_4$ also adversely affects human health as a precursor of tropospheric ozone [[2]](https://doi.org/10.1073/pnas.0600201103). Monitoring the long-term CH$_4$ variability is therefore crucial for monitoring the emission reductions [[3]](https://doi.org/10.5194/essd-17-1873-2025). In this assessment, atmospheric CH$_4$ spatial seasonal means and trends are analysed using the XCH$_4$ Level 3 gridded product (Obs4MIPs, version 4.5 [[4]](https://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PUGS_GHG_main_latest.pdf)), by adopting an approach similar to [[5]](https://doi.org/10.5194/amt-13-789-2020). Furthermore, we also evaluate the uncertainty associated with this dataset.\n\nThe Obs4MIPs XCH$_4$ product is generated by spatial (5°x5°) and temporal (monthly) gridding of the corresponding EMMA Level 2 data product. For more details see [[6]](https://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_ATBD_GHG_D_latest.pdf). By analysing an ensemble of different Level 2 datasets from various retrieval algorithms, the EMMA algorithm generates a dataset containing XCH$_4$ from individual retrievals. In particular, for each month and 10°×10° grid box, the algorithm with the grid box mean closest to the median is selected. The list of the considered retrieval algorithms is available from [[6]](https://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_ATBD_GHG_D_latest.pdf). The various retrieval algorithms are optimised for different instruments (SCIAMACHY, GOSAT, GOSAT-2) which measure backscattered solar radiation in the near-infrared for O$_2$ as well as the absorption bands of CO$_2$ or CH$_4$.\n\nCode is included for transparency but also learning purposes and gives users the chance to adapt the code used for the assesment as they wish. Users should always check product documentation and associated peer-reviewed papers for a more complete reporting of the issues discussed here (e.g., [[7]](https://doi.org/10.24381/14j9-s541), [[8]](https://climate.copernicus.eu/global-climate-highlights-2024))."} {"chunk_id": "satellite_satellite-methane_completeness_q03__2190ea62a407", "report_id": "satellite_satellite-methane_completeness_q03", "dataset_id": "satellite-methane", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q03", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Methane satellite observations uncertainty assessment for greenhouse gas monitoring > Quality assessment statements", "title": "Methane satellite observations uncertainty assessment for greenhouse gas monitoring", "chunk_index": 2, "token_count": 286, "text_raw": "These are the key outcomes of this assessment\n\n* The dataset \"Methane data from 2002 to present derived from satellite observations\" can be used to evaluate CH$_4$ mean values, climatology and growth rate over the globe, hemispheres or large regions.\n* Caution should be exercised in certain regions (high latitudes, regions with frequent cloud cover, oceans) where data availability varies along the temporal coverage of the dataset: this must be carefully considered when evaluating global or hemispheric information.\n* For data in high latitude regions or in regions with frequent cloud cover, users should consult uncertainty and quality flags as appropriate for their applications. In addition, for period November 2005 - March 2009 only values over land are available, which has to be taken into account for possible applications of this dataset.\n* As the uncertainties associated with the dataset are not constant over space and time, users are advised to consider the evolution of their values over the spatial regions and time periods of interest. The highest uncertainty values were found over high latitudes, the Himalayas and the tropical rainforest zone.\n```", "text_with_prefix": "EQC Quality Assessment: \"Methane satellite observations uncertainty assessment for greenhouse gas monitoring\"\nDataset: satellite-methane [CDS]\nAspect: completeness_q03 | Category: Satellite_ECVs\nSection: Methane satellite observations uncertainty assessment for greenhouse gas monitoring > Quality assessment statements\n---\nThese are the key outcomes of this assessment\n\n* The dataset \"Methane data from 2002 to present derived from satellite observations\" can be used to evaluate CH$_4$ mean values, climatology and growth rate over the globe, hemispheres or large regions.\n* Caution should be exercised in certain regions (high latitudes, regions with frequent cloud cover, oceans) where data availability varies along the temporal coverage of the dataset: this must be carefully considered when evaluating global or hemispheric information.\n* For data in high latitude regions or in regions with frequent cloud cover, users should consult uncertainty and quality flags as appropriate for their applications. In addition, for period November 2005 - March 2009 only values over land are available, which has to be taken into account for possible applications of this dataset.\n* As the uncertainties associated with the dataset are not constant over space and time, users are advised to consider the evolution of their values over the spatial regions and time periods of interest. The highest uncertainty values were found over high latitudes, the Himalayas and the tropical rainforest zone.\n```"} {"chunk_id": "satellite_satellite-methane_completeness_q03__cbdca5de3857", "report_id": "satellite_satellite-methane_completeness_q03", "dataset_id": "satellite-methane", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q03", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Methane satellite observations uncertainty assessment for greenhouse gas monitoring > Methodology", "title": "Methane satellite observations uncertainty assessment for greenhouse gas monitoring", "chunk_index": 3, "token_count": 770, "text_raw": "Spatial seasonal means and trends are presented and assessed using the XCH$_4$ v4.5 Level 3 gridded product (Obs4MIPs), which has been generated using the Level 2 EMMA products [[4]](https://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PUGS_GHG_main_latest.pdf) as input.\n\nTo show how data coverage varies between years and seasons, we have calculated and plotted the average XCH$_4$ values for the different seasons and years, over the period 2013-2022. This time period has been chosen for the purpose of comparing the calculated trends with the results in [[1]](https://library.wmo.int/idurl/4/68532).\n\nSpatial trends are calculated using a linear model (i.e., Theil-Sen slope estimator) over monthly anomalies (i.e., actual monthly values minus climatological monthly means). This should be treated with caution as the long-term trend of atmospheric CH$_4$ is not strictly linear. The statistical significance of the trends is assessed using the Mann-Kendall test. Similar to [[5]](https://doi.org/10.5194/amt-13-789-2020), only land pixels are considered to avoid artefact related to different data availability over oceans (for period November 2005 - March 2009 only values over land are available).\n\nThe global maps of the product uncertainties (\"xch4_stderr\") for a specific subset of years (2015-2020) are shown, together with the time series of selected XCH$_4$ variables (\"xch4\", \"xch4_stderr\" and \"xch4_nobs\"). Please note that this analysis can be extended to one additional variable, not shown here (\"xch4_stddev\"), or can be customized by the user for specific regions of interest.\n\nThe analysis and results are organised in the following steps, which are detailed in the sections below:\n\n**[](template:section-1)**\n * Import all the relevant packages.\n * Choose the temporal and spatial coverage, land mask, and possible spatial regions for the analysis.\n * Cache needed functions.\n\n**[](template:section-2)**\n * In this section, we define the data request to CDS.\n\n**[](template:section-3)**\n * To show how data coverage varies between years and seasons, we have plotted spatial average XCH$_4$ values for seasons and years. To avoid the production of too many maps, this analysis has been limited to the period 2013-2022. Seasons are defined as December-February (DJF), March-May (MAM), June-August (JJA) and September-November (SON).\n\n**[](template:section-4)**\n * Trends (over 2013-2022) are calculated using a linear model (i.e., Theil-Sen slope estimator) over monthly anomalies (i.e., actual monthly values minus climatological monthly means). The statistical significance of the trends is assessed using the Mann-Kendall test.\n\n**[](template:section-5)**\n * Annual global maps of the XCH$_4$ uncertainties are presented, for a subset of data (2015-2020).\n * Global time series of XCH$_4$, related uncertainties and data availability over the entire data period (2003-2022) are shown.", "text_with_prefix": "EQC Quality Assessment: \"Methane satellite observations uncertainty assessment for greenhouse gas monitoring\"\nDataset: satellite-methane [CDS]\nAspect: completeness_q03 | Category: Satellite_ECVs\nSection: Methane satellite observations uncertainty assessment for greenhouse gas monitoring > Methodology\n---\nSpatial seasonal means and trends are presented and assessed using the XCH$_4$ v4.5 Level 3 gridded product (Obs4MIPs), which has been generated using the Level 2 EMMA products [[4]](https://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PUGS_GHG_main_latest.pdf) as input.\n\nTo show how data coverage varies between years and seasons, we have calculated and plotted the average XCH$_4$ values for the different seasons and years, over the period 2013-2022. This time period has been chosen for the purpose of comparing the calculated trends with the results in [[1]](https://library.wmo.int/idurl/4/68532).\n\nSpatial trends are calculated using a linear model (i.e., Theil-Sen slope estimator) over monthly anomalies (i.e., actual monthly values minus climatological monthly means). This should be treated with caution as the long-term trend of atmospheric CH$_4$ is not strictly linear. The statistical significance of the trends is assessed using the Mann-Kendall test. Similar to [[5]](https://doi.org/10.5194/amt-13-789-2020), only land pixels are considered to avoid artefact related to different data availability over oceans (for period November 2005 - March 2009 only values over land are available).\n\nThe global maps of the product uncertainties (\"xch4_stderr\") for a specific subset of years (2015-2020) are shown, together with the time series of selected XCH$_4$ variables (\"xch4\", \"xch4_stderr\" and \"xch4_nobs\"). Please note that this analysis can be extended to one additional variable, not shown here (\"xch4_stddev\"), or can be customized by the user for specific regions of interest.\n\nThe analysis and results are organised in the following steps, which are detailed in the sections below:\n\n**[](template:section-1)**\n * Import all the relevant packages.\n * Choose the temporal and spatial coverage, land mask, and possible spatial regions for the analysis.\n * Cache needed functions.\n\n**[](template:section-2)**\n * In this section, we define the data request to CDS.\n\n**[](template:section-3)**\n * To show how data coverage varies between years and seasons, we have plotted spatial average XCH$_4$ values for seasons and years. To avoid the production of too many maps, this analysis has been limited to the period 2013-2022. Seasons are defined as December-February (DJF), March-May (MAM), June-August (JJA) and September-November (SON).\n\n**[](template:section-4)**\n * Trends (over 2013-2022) are calculated using a linear model (i.e., Theil-Sen slope estimator) over monthly anomalies (i.e., actual monthly values minus climatological monthly means). The statistical significance of the trends is assessed using the Mann-Kendall test.\n\n**[](template:section-5)**\n * Annual global maps of the XCH$_4$ uncertainties are presented, for a subset of data (2015-2020).\n * Global time series of XCH$_4$, related uncertainties and data availability over the entire data period (2003-2022) are shown."} {"chunk_id": "satellite_satellite-methane_completeness_q03__3e6d5f679ac3", "report_id": "satellite_satellite-methane_completeness_q03", "dataset_id": "satellite-methane", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q03", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Methane satellite observations uncertainty assessment for greenhouse gas monitoring > Analysis and results > 1. Choose the data to use and set-up the code > Choose temporal and spatial coverage, land mask", "title": "Methane satellite observations uncertainty assessment for greenhouse gas monitoring", "chunk_index": 4, "token_count": 291, "text_raw": "In this section, we define the parameters to be ingested by the code (that can be customized by the user), i.e.: \n* the temporal period of analysis;\n* the activation/deactivation of the land masking;\n* the regions selected for the analysis. Please note that, for this notebook, only the global maps and time series are reported.\n\nThe analyses presented in this notebook cover different time periods:\n* with the aim to compare our results with existing literature [[1]](https://library.wmo.int/idurl/4/68532), the global variability and trend analyses are limited to 2013-2022;\n* the global plots of the XCH$_4$ uncertainties refer to the years from 2015 to 2020;\n* the time series of XCH$_4$, related uncertainties and data availability cover the entire period of data coverage.\n\nChoose variable\nChoose a time period (to be used for the global variability and trend analysis)\nMinimum value of land fraction used for masking\nDefine regions for analysis", "text_with_prefix": "EQC Quality Assessment: \"Methane satellite observations uncertainty assessment for greenhouse gas monitoring\"\nDataset: satellite-methane [CDS]\nAspect: completeness_q03 | Category: Satellite_ECVs\nSection: Methane satellite observations uncertainty assessment for greenhouse gas monitoring > Analysis and results > 1. Choose the data to use and set-up the code > Choose temporal and spatial coverage, land mask\n---\nIn this section, we define the parameters to be ingested by the code (that can be customized by the user), i.e.: \n* the temporal period of analysis;\n* the activation/deactivation of the land masking;\n* the regions selected for the analysis. Please note that, for this notebook, only the global maps and time series are reported.\n\nThe analyses presented in this notebook cover different time periods:\n* with the aim to compare our results with existing literature [[1]](https://library.wmo.int/idurl/4/68532), the global variability and trend analyses are limited to 2013-2022;\n* the global plots of the XCH$_4$ uncertainties refer to the years from 2015 to 2020;\n* the time series of XCH$_4$, related uncertainties and data availability cover the entire period of data coverage.\n\nChoose variable\nChoose a time period (to be used for the global variability and trend analysis)\nMinimum value of land fraction used for masking\nDefine regions for analysis"} {"chunk_id": "satellite_satellite-methane_completeness_q03__5e14325ab7ea", "report_id": "satellite_satellite-methane_completeness_q03", "dataset_id": "satellite-methane", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q03", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Methane satellite observations uncertainty assessment for greenhouse gas monitoring > Analysis and results > 1. Choose the data to use and set-up the code > Chache needed functions", "title": "Methane satellite observations uncertainty assessment for greenhouse gas monitoring", "chunk_index": 5, "token_count": 461, "text_raw": "In this section, we cached a list of functions used in the analyses.\n\n- The function `get_da` (`get_da_nomask`) is used to subset the data for the defined time period and spatial region by applying (not appliyng) the land mask (as a function of *min_land_fraction*).\n\n- The function `convert_units` rescales XCH$_4$ mole fraction to parts per billion (ppb).\n\n- The `seasonal_weighted_mean` function extracts the regional means over the selected domains. It uses spatial weighting to account for the latitudinal dependence of the grid size in the lon/lat grids used for the reanalysis and for the forecast models. It is used by the function `compute_seasonal_timeseries_nomask` to provide the seasonal XCH$_4$ average value for each year (Fig. 1).\n\n- The function `compute_anomaly_trends` is used to calculate the trend and the related statistical significance.\n\n- The `compute_monthly_anomalies` function is used to derive the monthly XCH$_4$ anomalies before calculating the trends.\n\n- The `mask_scale_and_regionalise` function extracts the XCH$_4$ data over the selected spatial region. It uses spatial weighting to account for the latitudinal dependence of the grid size in the lon/lat grids used for the reanalysis and for the forecast models. It uses the `convert_units` function for rescaling the values to ppb, and it applies the threshold (if any) on the minimum land fraction.\n\nShift years (shift -1 to get D(year-1)J(year)F(year))\nGet rid of 1st JF and last D, so it become [MAM, JJA, SON, DJF, ..., SON]\nMann-Kendall\nDetrended anomalies\n\n(template:section-2)=", "text_with_prefix": "EQC Quality Assessment: \"Methane satellite observations uncertainty assessment for greenhouse gas monitoring\"\nDataset: satellite-methane [CDS]\nAspect: completeness_q03 | Category: Satellite_ECVs\nSection: Methane satellite observations uncertainty assessment for greenhouse gas monitoring > Analysis and results > 1. Choose the data to use and set-up the code > Chache needed functions\n---\nIn this section, we cached a list of functions used in the analyses.\n\n- The function `get_da` (`get_da_nomask`) is used to subset the data for the defined time period and spatial region by applying (not appliyng) the land mask (as a function of *min_land_fraction*).\n\n- The function `convert_units` rescales XCH$_4$ mole fraction to parts per billion (ppb).\n\n- The `seasonal_weighted_mean` function extracts the regional means over the selected domains. It uses spatial weighting to account for the latitudinal dependence of the grid size in the lon/lat grids used for the reanalysis and for the forecast models. It is used by the function `compute_seasonal_timeseries_nomask` to provide the seasonal XCH$_4$ average value for each year (Fig. 1).\n\n- The function `compute_anomaly_trends` is used to calculate the trend and the related statistical significance.\n\n- The `compute_monthly_anomalies` function is used to derive the monthly XCH$_4$ anomalies before calculating the trends.\n\n- The `mask_scale_and_regionalise` function extracts the XCH$_4$ data over the selected spatial region. It uses spatial weighting to account for the latitudinal dependence of the grid size in the lon/lat grids used for the reanalysis and for the forecast models. It uses the `convert_units` function for rescaling the values to ppb, and it applies the threshold (if any) on the minimum land fraction.\n\nShift years (shift -1 to get D(year-1)J(year)F(year))\nGet rid of 1st JF and last D, so it become [MAM, JJA, SON, DJF, ..., SON]\nMann-Kendall\nDetrended anomalies\n\n(template:section-2)="} {"chunk_id": "satellite_satellite-methane_completeness_q03__27c7d4b3cb7d", "report_id": "satellite_satellite-methane_completeness_q03", "dataset_id": "satellite-methane", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q03", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Methane satellite observations uncertainty assessment for greenhouse gas monitoring > Analysis and results > 2. Retrieve XCH$_4$ data (Obs4MIPs)", "title": "Methane satellite observations uncertainty assessment for greenhouse gas monitoring", "chunk_index": 6, "token_count": 432, "text_raw": "In this section, we define the data request to CDS (data product Obs4MIPs, Level 3, version 4.5, XCH$_4$) and download the dataset.\n\n```text\nregion='global'\n```\n\n```text\n0%| | 0/1 [00:00 Analysis and results > 2. Retrieve XCH$_4$ data (Obs4MIPs)\n---\nIn this section, we define the data request to CDS (data product Obs4MIPs, Level 3, version 4.5, XCH$_4$) and download the dataset.\n\n```text\nregion='global'\n```\n\n```text\n0%| | 0/1 [00:00 Analysis and results > 3. Compute and plot the global variability of seasonal XCH$_4$", "title": "Methane satellite observations uncertainty assessment for greenhouse gas monitoring", "chunk_index": 7, "token_count": 663, "text_raw": "To show how data coverage varies between years and seasons, in this section we plot the spatial average XCH$_4$ values for seasons and years. To limit the number of plots produced, this analysis is limited to the period 2013-2022. Seasons are defined as: December-February (DJF), March-May (MAM), June-August (JJA), and September-November (SON). The seasonality of the availability of measurements is evident, with values available in the Southern Hemisphere (SH) mid and high latitudes from September to March, whereas values for the Northern Hemisphere (NH) mid and high latitudes are mostly available from April to August.\n\nAnalysis of global variability of seasonal XCH4\nTo invalidate the cache you can pass the argument invalidate_cache=True\n\n```text\n0%| | 0/1 [00:00 Analysis and results > 3. Compute and plot the global variability of seasonal XCH$_4$\n---\nTo show how data coverage varies between years and seasons, in this section we plot the spatial average XCH$_4$ values for seasons and years. To limit the number of plots produced, this analysis is limited to the period 2013-2022. Seasons are defined as: December-February (DJF), March-May (MAM), June-August (JJA), and September-November (SON). The seasonality of the availability of measurements is evident, with values available in the Southern Hemisphere (SH) mid and high latitudes from September to March, whereas values for the Northern Hemisphere (NH) mid and high latitudes are mostly available from April to August.\n\nAnalysis of global variability of seasonal XCH4\nTo invalidate the cache you can pass the argument invalidate_cache=True\n\n```text\n0%| | 0/1 [00:00 Analysis and results > 4. Compute global and spatial trends", "title": "Methane satellite observations uncertainty assessment for greenhouse gas monitoring", "chunk_index": 8, "token_count": 465, "text_raw": "In this section, XCH$_4$ spatial trends are calculated and statistical significance is assessed for land pixels (\"land_fraction > 0.5\"). Global statistics for individual pixels are also reported as ppb/month. Please note that the low growth rates observed in the tropics are associated with high statistical errors in trend calculations. This is probably due to the frequent presence of clouds resulting in sparse data availability. Over Greenland, however, high retrieval uncertainty may be a factor due to the high surface albedo.\n\nFor 2013-2022, the mean global XCH$_4$ growth rate is 9.71 $\\pm$ 2.21 ppb/yr ($\\pm$ standard error of the trend calculation). Taking into account the associated uncertainties, the different calculation methods and the different vertical representativeness of satellite observations with respect to near-surface measurements, this value is reasonably consistent with the mean absolute increase of 10.20 ppb/yr over the past 10 years, as reported by the WMO's global in-situ observations [[1]](https://library.wmo.int/idurl/4/68532).\n\nCalculation of global and pixel trends\nTo invalidate the cache you can pass the argument invalidate_cache=True\n\n```text\n100%|██████████| 1/1 [00:03<00:00, 3.63s/it]\n```\n\n*Trends (upper map) and related standard error (bottom map) of XCH$_4$ given in ppb/month calculated by using linear model (Theil-Sen) and Mann-Kendall test for statistical significance. Over each pixel, trends are calculated for monthly anomalies over land (\"land_fraction > 0.5\"). In the upper map, any shaded areas indicate pixels that did not pass the Mann-Kendall significance test. Global statistics for individual pixels are shown on the right of the plot.*\n\n(template:section-5)=", "text_with_prefix": "EQC Quality Assessment: \"Methane satellite observations uncertainty assessment for greenhouse gas monitoring\"\nDataset: satellite-methane [CDS]\nAspect: completeness_q03 | Category: Satellite_ECVs\nSection: Methane satellite observations uncertainty assessment for greenhouse gas monitoring > Analysis and results > 4. Compute global and spatial trends\n---\nIn this section, XCH$_4$ spatial trends are calculated and statistical significance is assessed for land pixels (\"land_fraction > 0.5\"). Global statistics for individual pixels are also reported as ppb/month. Please note that the low growth rates observed in the tropics are associated with high statistical errors in trend calculations. This is probably due to the frequent presence of clouds resulting in sparse data availability. Over Greenland, however, high retrieval uncertainty may be a factor due to the high surface albedo.\n\nFor 2013-2022, the mean global XCH$_4$ growth rate is 9.71 $\\pm$ 2.21 ppb/yr ($\\pm$ standard error of the trend calculation). Taking into account the associated uncertainties, the different calculation methods and the different vertical representativeness of satellite observations with respect to near-surface measurements, this value is reasonably consistent with the mean absolute increase of 10.20 ppb/yr over the past 10 years, as reported by the WMO's global in-situ observations [[1]](https://library.wmo.int/idurl/4/68532).\n\nCalculation of global and pixel trends\nTo invalidate the cache you can pass the argument invalidate_cache=True\n\n```text\n100%|██████████| 1/1 [00:03<00:00, 3.63s/it]\n```\n\n*Trends (upper map) and related standard error (bottom map) of XCH$_4$ given in ppb/month calculated by using linear model (Theil-Sen) and Mann-Kendall test for statistical significance. Over each pixel, trends are calculated for monthly anomalies over land (\"land_fraction > 0.5\"). In the upper map, any shaded areas indicate pixels that did not pass the Mann-Kendall significance test. Global statistics for individual pixels are shown on the right of the plot.*\n\n(template:section-5)="} {"chunk_id": "satellite_satellite-methane_completeness_q03__252961b6f69f", "report_id": "satellite_satellite-methane_completeness_q03", "dataset_id": "satellite-methane", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q03", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Methane satellite observations uncertainty assessment for greenhouse gas monitoring > Analysis and results > 5. Analysis concerning XCH$_4$ uncertainties > Global maps of XCH$_4$ uncertainties", "title": "Methane satellite observations uncertainty assessment for greenhouse gas monitoring", "chunk_index": 9, "token_count": 403, "text_raw": "In this section, we show the multi-year (for the 2015-2020 period) global maps of the XCH$_4$ reported uncertainties, i.e., the \"xch4_stderr\" variable. This parameter is defined as the standard error of the average including single sounding noise and potential seasonal and regional biases [[4]](https://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PUGS_GHG_main_latest.pdf). The yearly averages are calculated by averaging the monthly values provided by the considered dataset. Throughout all the selected years some features emerge, with higher uncertainties mainly over the Himalayas, South Asia, high latitudes, and the tropical rainforest zone. This is mainly due to the sparseness in sampling because of frequent cloud cover, as well as large solar zenith angles in high latitudes, which are a challenge for accurate XCH$_4$ retrievals [[5]](https://doi.org/10.5194/amt-13-789-2020). It is worth noting that regions characterised by large uncertainties were also affected by spatial trends that deviated from the global average (see the previous figure). Users should therefore exercise caution when deriving long-term trends for regions affected by high uncertainty.\n\n*Multi-year global distribution of XCH$_4$ standard error (from 2015 to 2020), derived from the XCH4_OBS4MIPS dataset (version 4.5). The title indicates the spatial region and the selected variable.*", "text_with_prefix": "EQC Quality Assessment: \"Methane satellite observations uncertainty assessment for greenhouse gas monitoring\"\nDataset: satellite-methane [CDS]\nAspect: completeness_q03 | Category: Satellite_ECVs\nSection: Methane satellite observations uncertainty assessment for greenhouse gas monitoring > Analysis and results > 5. Analysis concerning XCH$_4$ uncertainties > Global maps of XCH$_4$ uncertainties\n---\nIn this section, we show the multi-year (for the 2015-2020 period) global maps of the XCH$_4$ reported uncertainties, i.e., the \"xch4_stderr\" variable. This parameter is defined as the standard error of the average including single sounding noise and potential seasonal and regional biases [[4]](https://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PUGS_GHG_main_latest.pdf). The yearly averages are calculated by averaging the monthly values provided by the considered dataset. Throughout all the selected years some features emerge, with higher uncertainties mainly over the Himalayas, South Asia, high latitudes, and the tropical rainforest zone. This is mainly due to the sparseness in sampling because of frequent cloud cover, as well as large solar zenith angles in high latitudes, which are a challenge for accurate XCH$_4$ retrievals [[5]](https://doi.org/10.5194/amt-13-789-2020). It is worth noting that regions characterised by large uncertainties were also affected by spatial trends that deviated from the global average (see the previous figure). Users should therefore exercise caution when deriving long-term trends for regions affected by high uncertainty.\n\n*Multi-year global distribution of XCH$_4$ standard error (from 2015 to 2020), derived from the XCH4_OBS4MIPS dataset (version 4.5). The title indicates the spatial region and the selected variable.*"} {"chunk_id": "satellite_satellite-methane_completeness_q03__cd3507321979", "report_id": "satellite_satellite-methane_completeness_q03", "dataset_id": "satellite-methane", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q03", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Methane satellite observations uncertainty assessment for greenhouse gas monitoring > Analysis and results > 5. Analysis concerning XCH$_4$ uncertainties > Time series analysis for the different variables", "title": "Methane satellite observations uncertainty assessment for greenhouse gas monitoring", "chunk_index": 10, "token_count": 944, "text_raw": "In this section, we show the time series for three XCH$_4$ variables (i.e., \"xch4\", \"xch4_stderr\", and \"xch4_nobs\"), considering the entire dataset. For each variable, each plot displays the monthly global spatial average, along with the corresponding monthly standard deviation (except for \"xch4_nobs\").\nThis analysis would provide the user with:\n* the global time series of XCH$_4$ values (\"xch4\"), highlighting the overall positive trend;\n* the change over time of the uncertainty associated to the XCH$_4$ values (\"xch4_stderr\");\n* the time series of \"xch4_nobs\", which denotes the number of individual XCH$_4$ Level 2 observations used to compute the monthly Level 3 data.\n\nThe XCH$_4$ values have been nearly constant until 2007, then a positive trend has been detected. This is likely due to a combination of increasing natural (e.g., wetlands) and anthropogenic (e.g., fossil-fuel related) emissions, and possible decreasing sinks, although this is currently under investigation (see [[3]](https://doi.org/10.5194/essd-17-1873-2025) and references therein).\n\nBy looking at the uncertainties (\"xch4_stderr\"), the most evident feature is the sharp decrease after 2009, likely due to the introduction of algorithms based on GOSAT observations used to calculate the median XCH$_4$ data [[6]](https://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_ATBD_GHG_D_latest.pdf). A further increase in the uncertainties is observed from 2020 onwards, and this is probably linked to the introduction of GOSAT-2 measurements [[6]](https://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_ATBD_GHG_D_latest.pdf).\n\nThe temporal variability of \"xch4_nobs\" traces the changes over time of the input data availability, and the number of used algorithms to obtain the merged Level 2 data products, from which the Obs4MIPs product is derived (please note that the standard deviations have not been reported for \"xch4_nobs\" to increase the plot readability). The first time period (up to April 2009) is characterized by a significant number of observations by the SCIAMACHY WFMD product only. The following period (up to 2022) reflects the reduced number of soundings provided by SCIAMACHY with the contributions from GOSAT and GOSAT-2 (from January 2019). For details about the different Level 2 products used as input for the generation of the Level 3 XCH$_4$ data, see [[5]](https://doi.org/10.5194/amt-13-789-2020) and [[6]](https://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_ATBD_GHG_D_latest.pdf).\n\nPlease note that this analysis can be customized by the user to limit it to specific regions, or to include one additional variable (not shown here): \"xch4_stddev\", which represents the standard deviation of the XCH$_4$ Level 2 observations within each grid box.\n\nIf variaable is xco2_nobs, use log scale for y-axis\nax.set_ylabel(\"n observations\")\nImposta tick logaritmici automatici\n\n*Global monthly time series of XCH$_4$ (top panel) and their associated uncertainties (middle panel) for the entire dataset. The blue lines represent the monthly spatial average, while the shaded areas indicate $\\pm$1 standard deviation. The bottom panel shows the time series of monthly spatial average of the number of individual XCH$_4$ Level 2 observations used to compute Obs4MIPs (Level 3) data. The main titles indicate the variable shown, while the subtitles specify the corresponding region.*", "text_with_prefix": "EQC Quality Assessment: \"Methane satellite observations uncertainty assessment for greenhouse gas monitoring\"\nDataset: satellite-methane [CDS]\nAspect: completeness_q03 | Category: Satellite_ECVs\nSection: Methane satellite observations uncertainty assessment for greenhouse gas monitoring > Analysis and results > 5. Analysis concerning XCH$_4$ uncertainties > Time series analysis for the different variables\n---\nIn this section, we show the time series for three XCH$_4$ variables (i.e., \"xch4\", \"xch4_stderr\", and \"xch4_nobs\"), considering the entire dataset. For each variable, each plot displays the monthly global spatial average, along with the corresponding monthly standard deviation (except for \"xch4_nobs\").\nThis analysis would provide the user with:\n* the global time series of XCH$_4$ values (\"xch4\"), highlighting the overall positive trend;\n* the change over time of the uncertainty associated to the XCH$_4$ values (\"xch4_stderr\");\n* the time series of \"xch4_nobs\", which denotes the number of individual XCH$_4$ Level 2 observations used to compute the monthly Level 3 data.\n\nThe XCH$_4$ values have been nearly constant until 2007, then a positive trend has been detected. This is likely due to a combination of increasing natural (e.g., wetlands) and anthropogenic (e.g., fossil-fuel related) emissions, and possible decreasing sinks, although this is currently under investigation (see [[3]](https://doi.org/10.5194/essd-17-1873-2025) and references therein).\n\nBy looking at the uncertainties (\"xch4_stderr\"), the most evident feature is the sharp decrease after 2009, likely due to the introduction of algorithms based on GOSAT observations used to calculate the median XCH$_4$ data [[6]](https://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_ATBD_GHG_D_latest.pdf). A further increase in the uncertainties is observed from 2020 onwards, and this is probably linked to the introduction of GOSAT-2 measurements [[6]](https://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_ATBD_GHG_D_latest.pdf).\n\nThe temporal variability of \"xch4_nobs\" traces the changes over time of the input data availability, and the number of used algorithms to obtain the merged Level 2 data products, from which the Obs4MIPs product is derived (please note that the standard deviations have not been reported for \"xch4_nobs\" to increase the plot readability). The first time period (up to April 2009) is characterized by a significant number of observations by the SCIAMACHY WFMD product only. The following period (up to 2022) reflects the reduced number of soundings provided by SCIAMACHY with the contributions from GOSAT and GOSAT-2 (from January 2019). For details about the different Level 2 products used as input for the generation of the Level 3 XCH$_4$ data, see [[5]](https://doi.org/10.5194/amt-13-789-2020) and [[6]](https://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_ATBD_GHG_D_latest.pdf).\n\nPlease note that this analysis can be customized by the user to limit it to specific regions, or to include one additional variable (not shown here): \"xch4_stddev\", which represents the standard deviation of the XCH$_4$ Level 2 observations within each grid box.\n\nIf variaable is xco2_nobs, use log scale for y-axis\nax.set_ylabel(\"n observations\")\nImposta tick logaritmici automatici\n\n*Global monthly time series of XCH$_4$ (top panel) and their associated uncertainties (middle panel) for the entire dataset. The blue lines represent the monthly spatial average, while the shaded areas indicate $\\pm$1 standard deviation. The bottom panel shows the time series of monthly spatial average of the number of individual XCH$_4$ Level 2 observations used to compute Obs4MIPs (Level 3) data. The main titles indicate the variable shown, while the subtitles specify the corresponding region.*"} {"chunk_id": "satellite_satellite-methane_completeness_q03__13ba1ff3aa4f", "report_id": "satellite_satellite-methane_completeness_q03", "dataset_id": "satellite-methane", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q03", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Methane satellite observations uncertainty assessment for greenhouse gas monitoring > ℹ️ If you want to know more > Key resources", "title": "Methane satellite observations uncertainty assessment for greenhouse gas monitoring", "chunk_index": 11, "token_count": 322, "text_raw": "The CDS catalogue entries for the data used were:\n* Methane data from 2002 to present derived from satellite observations: https://cds.climate.copernicus.eu/datasets/satellite-methane?tab=overview\n\nCode libraries used:\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)\n\nUsers interested in obtaining updated figures for methane growth rates and global trends are directed to official C3S sources for precise reporting: [[7]](https://doi.org/10.24381/14j9-s541), [[8]](https://climate.copernicus.eu/global-climate-highlights-2024).\n\nUsers interested in near-real time detection of hot-spot locations for methane emissions can consider to use the CAMS Methane Hotspot Explorer: https://atmosphere.copernicus.eu/ghg-services/cams-methane-hotspot-explorer?utm_source=press&utm_medium=referral&utm_campaign=CH4-app-2025", "text_with_prefix": "EQC Quality Assessment: \"Methane satellite observations uncertainty assessment for greenhouse gas monitoring\"\nDataset: satellite-methane [CDS]\nAspect: completeness_q03 | Category: Satellite_ECVs\nSection: Methane satellite observations uncertainty assessment for greenhouse gas monitoring > ℹ️ If you want to know more > Key resources\n---\nThe CDS catalogue entries for the data used were:\n* Methane data from 2002 to present derived from satellite observations: https://cds.climate.copernicus.eu/datasets/satellite-methane?tab=overview\n\nCode libraries used:\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)\n\nUsers interested in obtaining updated figures for methane growth rates and global trends are directed to official C3S sources for precise reporting: [[7]](https://doi.org/10.24381/14j9-s541), [[8]](https://climate.copernicus.eu/global-climate-highlights-2024).\n\nUsers interested in near-real time detection of hot-spot locations for methane emissions can consider to use the CAMS Methane Hotspot Explorer: https://atmosphere.copernicus.eu/ghg-services/cams-methane-hotspot-explorer?utm_source=press&utm_medium=referral&utm_campaign=CH4-app-2025"} {"chunk_id": "satellite_satellite-methane_completeness_q03__2e9c2065f537", "report_id": "satellite_satellite-methane_completeness_q03", "dataset_id": "satellite-methane", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q03", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Methane satellite observations uncertainty assessment for greenhouse gas monitoring > ℹ️ If you want to know more > References", "title": "Methane satellite observations uncertainty assessment for greenhouse gas monitoring", "chunk_index": 12, "token_count": 796, "text_raw": "[[1]](https://library.wmo.int/idurl/4/68532) World Meteorological Organization (2023). WMO Greenhouse Gas Bulletin, No. 19, ISSN 2078-0796.\n\n[[2]](https://doi.org/10.1073/pnas.0600201103) West, J. J., Fiore, A. M., Horowitz, L. W., and Mauzerall, D. L. (2006). Global health benefits of mitigating ozone pollution with methane emission controls. Proceedings of the National Academy of Sciences USA, 103, 3988–3993.\n\n[[3]](https://doi.org/10.5194/essd-17-1873-2025) Saunois, M., Martinez, A., Poulter, B., Zhang, Z., Raymond, P. A., Regnier, P., Canadell, J. G., Jackson, R. B., Patra, P. K., Bousquet, P., Ciais, P., Dlugokencky, E. J., Lan, X., Allen, G. H., Bastviken, D., Beerling, D. J., Belikov, D. A., Blake, D. R., Castaldi, S., Crippa, M., Deemer, B. R., Dennison, F., Etiope, G., Gedney, N., Höglund-Isaksson, L., Holgerson, M. A., Hopcroft, P. O., Hugelius, G., Ito, A., Jain, A. K., Janardanan, R., Johnson, M. S., Kleinen, T., Krummel, P. B., Lauerwald, R., Li, T., Liu, X., McDonald, K. C., Melton, J. R., Mühle, J., Müller, J., Murguia-Flores, F., Niwa, Y., Noce, S., Pan, S., Parker, R. J., Peng, C., Ramonet, M., Riley, W. J., Rocher-Ros, G., Rosentreter, J. A., Sasakawa, M., Segers, A., Smith, S. J., Stanley, E. H., Thanwerdas, J., Tian, H., Tsuruta, A., Tubiello, F. N., Weber, T. S., van der Werf, G. R., Worthy, D. E. J., Xi, Y., Yoshida, Y., Zhang, W., Zheng, B., Zhu, Q., Zhu, Q., and Zhuang, Q. (2025). Global Methane Budget 2000–2020, Earth System Science Data, 17, 1873–1958.\n\n[[4]](https://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PUGS_GHG_main_latest.pdf) Buchwitz, M. (2024). Product User Guide and Specification (PUGS) – Main document for Greenhouse Gas (GHG: CO$_2$ & CH$_4$) data set CDR7 (01.2003-12.2022), C3S project 2021/C3S2_312a_Lot2_DLR/SC1, v7.3.", "text_with_prefix": "EQC Quality Assessment: \"Methane satellite observations uncertainty assessment for greenhouse gas monitoring\"\nDataset: satellite-methane [CDS]\nAspect: completeness_q03 | Category: Satellite_ECVs\nSection: Methane satellite observations uncertainty assessment for greenhouse gas monitoring > ℹ️ If you want to know more > References\n---\n[[1]](https://library.wmo.int/idurl/4/68532) World Meteorological Organization (2023). WMO Greenhouse Gas Bulletin, No. 19, ISSN 2078-0796.\n\n[[2]](https://doi.org/10.1073/pnas.0600201103) West, J. J., Fiore, A. M., Horowitz, L. W., and Mauzerall, D. L. (2006). Global health benefits of mitigating ozone pollution with methane emission controls. Proceedings of the National Academy of Sciences USA, 103, 3988–3993.\n\n[[3]](https://doi.org/10.5194/essd-17-1873-2025) Saunois, M., Martinez, A., Poulter, B., Zhang, Z., Raymond, P. A., Regnier, P., Canadell, J. G., Jackson, R. B., Patra, P. K., Bousquet, P., Ciais, P., Dlugokencky, E. J., Lan, X., Allen, G. H., Bastviken, D., Beerling, D. J., Belikov, D. A., Blake, D. R., Castaldi, S., Crippa, M., Deemer, B. R., Dennison, F., Etiope, G., Gedney, N., Höglund-Isaksson, L., Holgerson, M. A., Hopcroft, P. O., Hugelius, G., Ito, A., Jain, A. K., Janardanan, R., Johnson, M. S., Kleinen, T., Krummel, P. B., Lauerwald, R., Li, T., Liu, X., McDonald, K. C., Melton, J. R., Mühle, J., Müller, J., Murguia-Flores, F., Niwa, Y., Noce, S., Pan, S., Parker, R. J., Peng, C., Ramonet, M., Riley, W. J., Rocher-Ros, G., Rosentreter, J. A., Sasakawa, M., Segers, A., Smith, S. J., Stanley, E. H., Thanwerdas, J., Tian, H., Tsuruta, A., Tubiello, F. N., Weber, T. S., van der Werf, G. R., Worthy, D. E. J., Xi, Y., Yoshida, Y., Zhang, W., Zheng, B., Zhu, Q., Zhu, Q., and Zhuang, Q. (2025). Global Methane Budget 2000–2020, Earth System Science Data, 17, 1873–1958.\n\n[[4]](https://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PUGS_GHG_main_latest.pdf) Buchwitz, M. (2024). Product User Guide and Specification (PUGS) – Main document for Greenhouse Gas (GHG: CO$_2$ & CH$_4$) data set CDR7 (01.2003-12.2022), C3S project 2021/C3S2_312a_Lot2_DLR/SC1, v7.3."} {"chunk_id": "satellite_satellite-methane_completeness_q03__a4c9c6509dc2", "report_id": "satellite_satellite-methane_completeness_q03", "dataset_id": "satellite-methane", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q03", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Methane satellite observations uncertainty assessment for greenhouse gas monitoring > ℹ️ If you want to know more > References", "title": "Methane satellite observations uncertainty assessment for greenhouse gas monitoring", "chunk_index": 13, "token_count": 680, "text_raw": "2003-12.2022), C3S project 2021/C3S2_312a_Lot2_DLR/SC1, v7.3.\n\n[[5]](https://doi.org/10.5194/amt-13-789-2020) Reuter, M., Buchwitz, M., Schneising, O., Noël, S., Bovensmann, H., Burrows, J. P., Boesch, H., Di Noia, A., Anand, J., Parker, R. J., Somkuti, P., Wu, L., Hasekamp, O. P., Aben, I., Kuze, A., Suto, H., Shiomi, K., Yoshida, Y., Morino, I., Crisp, D., O'Dell, C. W., Notholt, J., Petri, C., Warneke, T., Velazco, V. A., Deutscher, N. M., Griffith, D. W. T., Kivi, R., Pollard, D. F., Hase, F., Sussmann, R., Té, Y. V., Strong, K., Roche, S., Sha, M. K., De Mazière, M., Feist, D. G., Iraci, L. T., Roehl, C. M., Retscher, C., and Schepers, D. (2020). Ensemble-based satellite-derived carbon dioxide and methane column-averaged dry-air mole fraction data sets (2003–2018) for carbon and climate applications, Atmospheric Measurement Techniques, 13, 789–819.\n\n[[6]](https://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_ATBD_GHG_D_latest.pdf) Reuter,M. and Buchwitz, M. (2024). Algorithm Theoretical Basis Document (ATBD) – ANNEX D for products XCO2_EMMA, XCH4_EMMA, XCO2_OBS4MIPS, XCH4_OBS4MIPS (v4.5, CDR7, 2003-2022), C3S project 2021/C3S2_312a_Lot2_DLR/SC1, v7.1b.\n\n[[7]](https://doi.org/10.24381/14j9-s541) Copernicus Climate Change Service (C3S) and World Meteorological Organization (WMO). (2025). European State of the Climate 2024. https://doi.org/10.24381/14j9-s541\n\n[[8]](https://climate.copernicus.eu/global-climate-highlights-2024) Copernicus Climate Change Service (C3S). (2025). Global Climate Highlights 2024.", "text_with_prefix": "EQC Quality Assessment: \"Methane satellite observations uncertainty assessment for greenhouse gas monitoring\"\nDataset: satellite-methane [CDS]\nAspect: completeness_q03 | Category: Satellite_ECVs\nSection: Methane satellite observations uncertainty assessment for greenhouse gas monitoring > ℹ️ If you want to know more > References\n---\n2003-12.2022), C3S project 2021/C3S2_312a_Lot2_DLR/SC1, v7.3.\n\n[[5]](https://doi.org/10.5194/amt-13-789-2020) Reuter, M., Buchwitz, M., Schneising, O., Noël, S., Bovensmann, H., Burrows, J. P., Boesch, H., Di Noia, A., Anand, J., Parker, R. J., Somkuti, P., Wu, L., Hasekamp, O. P., Aben, I., Kuze, A., Suto, H., Shiomi, K., Yoshida, Y., Morino, I., Crisp, D., O'Dell, C. W., Notholt, J., Petri, C., Warneke, T., Velazco, V. A., Deutscher, N. M., Griffith, D. W. T., Kivi, R., Pollard, D. F., Hase, F., Sussmann, R., Té, Y. V., Strong, K., Roche, S., Sha, M. K., De Mazière, M., Feist, D. G., Iraci, L. T., Roehl, C. M., Retscher, C., and Schepers, D. (2020). Ensemble-based satellite-derived carbon dioxide and methane column-averaged dry-air mole fraction data sets (2003–2018) for carbon and climate applications, Atmospheric Measurement Techniques, 13, 789–819.\n\n[[6]](https://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_ATBD_GHG_D_latest.pdf) Reuter,M. and Buchwitz, M. (2024). Algorithm Theoretical Basis Document (ATBD) – ANNEX D for products XCO2_EMMA, XCH4_EMMA, XCO2_OBS4MIPS, XCH4_OBS4MIPS (v4.5, CDR7, 2003-2022), C3S project 2021/C3S2_312a_Lot2_DLR/SC1, v7.1b.\n\n[[7]](https://doi.org/10.24381/14j9-s541) Copernicus Climate Change Service (C3S) and World Meteorological Organization (WMO). (2025). European State of the Climate 2024. https://doi.org/10.24381/14j9-s541\n\n[[8]](https://climate.copernicus.eu/global-climate-highlights-2024) Copernicus Climate Change Service (C3S). (2025). Global Climate Highlights 2024."} {"chunk_id": "satellite_satellite-methane_completeness_q04__b6dda8cc64c6", "report_id": "satellite_satellite-methane_completeness_q04", "dataset_id": "satellite-methane", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q04", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Methane satellite observations completeness assessment for greenhouse gas monitoring", "title": "Methane satellite observations completeness assessment for greenhouse gas monitoring", "chunk_index": 0, "token_count": 867, "text_raw": "Production date: 11-02-2026\n\nProduced by: Davide Putero, Paolo Cristofanelli (CNR)\n\nPlease note that this assessment also refers to the deprecated Level 2 XCH$_4$ data products which won't be extended after 2022.\n\n## 🌍 Use case: Using satellite column-averaged XCH$_4$ observations for investigating global and zonal methane growth rates\n\n## ❓ Quality assessment question\n* **Is the spatial coverage of XCH$_4$ EMMA Level 2 merged dataset suitable to calculate the methane annual global and zonal growth rates?**\n\nMethane (CH$_4$) is the second most important anthropogenic greenhouse gas after carbon dioxide (CO$_2$), representing about 19% of the total radiative forcing by long-lived greenhouse gases [[1]](https://library.wmo.int/idurl/4/68532). Atmospheric CH$_4$ also adversely affects human health as a precursor of tropospheric ozone [[2]](https://doi.org/10.1073/pnas.0600201103). Monitoring the long-term CH$_4$ variability is therefore crucial for tracking the emission reductions [[3]](https://doi.org/10.5194/essd-17-1873-2025). Calculating growth rates is particularly important, as they provide a direct measure of how atmospheric CH$_4$ responds to changing emissions and sinks, and help disentangle interannual variability from long-term trends ([[4]](https://doi.org/10.5194/acp-24-577-2024), [[5]](https://doi.org/10.1038/s41558-023-01629-0), [[6]](https://doi.org/10.5194/amt-13-789-2020)).\n\nIn this assessment, atmospheric CH$_4$ global and zonal growth rates are derived from the XCH$_4$ merged Ensemble Median Algorithm (EMMA) Level 2 dataset (version 4.5), by adopting an approach similar to [[4]](https://doi.org/10.5194/acp-24-577-2024).\n\nThe EMMA Level 2 data product comprise individual soundings retrieved by algorithms that can change from grid box to grid box and from month to month. For more details see [[7]](https://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PUGS_GHG_main_latest.pdf). By analyzing an ensemble of different Level 2 datasets from various retrieval algorithms, the EMMA algorithm generates a dataset containing XCH$_4$ from individual retrievals. In particular, for each month and 10°$\\times$10° grid box, the algorithm with the grid box mean closest to the median is selected. The list of the considered retrieval algorithms is available from [[7]](https://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PUGS_GHG_main_latest.pdf). The various retrieval algorithms are optimized for different instruments (SCIAMACHY, GOSAT, GOSAT-2), which measure backscattered solar radiation in the near-infrared for O$_2$ as well as the absorption bands of CO$_2$ or CH$_4$.\n\nThe code is included for transparency and learning purposes, giving users the chance to adapt it for their own analyses. Users should always refer to the official reports for scientific assessments related to atmospheric CH$_4$ variability (e.g., [[9]](https://doi.org/10.24381/14j9-s541), [[10]](https://climate.copernicus.eu/global-climate-highlights-2024)).\n\n## 📢 Quality assessment statements\n\nThese are the key outcomes of this assessment", "text_with_prefix": "EQC Quality Assessment: \"Methane satellite observations completeness assessment for greenhouse gas monitoring\"\nDataset: satellite-methane [CDS]\nAspect: completeness_q04 | Category: Satellite_ECVs\nSection: Methane satellite observations completeness assessment for greenhouse gas monitoring\n---\nProduction date: 11-02-2026\n\nProduced by: Davide Putero, Paolo Cristofanelli (CNR)\n\nPlease note that this assessment also refers to the deprecated Level 2 XCH$_4$ data products which won't be extended after 2022.\n\n## 🌍 Use case: Using satellite column-averaged XCH$_4$ observations for investigating global and zonal methane growth rates\n\n## ❓ Quality assessment question\n* **Is the spatial coverage of XCH$_4$ EMMA Level 2 merged dataset suitable to calculate the methane annual global and zonal growth rates?**\n\nMethane (CH$_4$) is the second most important anthropogenic greenhouse gas after carbon dioxide (CO$_2$), representing about 19% of the total radiative forcing by long-lived greenhouse gases [[1]](https://library.wmo.int/idurl/4/68532). Atmospheric CH$_4$ also adversely affects human health as a precursor of tropospheric ozone [[2]](https://doi.org/10.1073/pnas.0600201103). Monitoring the long-term CH$_4$ variability is therefore crucial for tracking the emission reductions [[3]](https://doi.org/10.5194/essd-17-1873-2025). Calculating growth rates is particularly important, as they provide a direct measure of how atmospheric CH$_4$ responds to changing emissions and sinks, and help disentangle interannual variability from long-term trends ([[4]](https://doi.org/10.5194/acp-24-577-2024), [[5]](https://doi.org/10.1038/s41558-023-01629-0), [[6]](https://doi.org/10.5194/amt-13-789-2020)).\n\nIn this assessment, atmospheric CH$_4$ global and zonal growth rates are derived from the XCH$_4$ merged Ensemble Median Algorithm (EMMA) Level 2 dataset (version 4.5), by adopting an approach similar to [[4]](https://doi.org/10.5194/acp-24-577-2024).\n\nThe EMMA Level 2 data product comprise individual soundings retrieved by algorithms that can change from grid box to grid box and from month to month. For more details see [[7]](https://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PUGS_GHG_main_latest.pdf). By analyzing an ensemble of different Level 2 datasets from various retrieval algorithms, the EMMA algorithm generates a dataset containing XCH$_4$ from individual retrievals. In particular, for each month and 10°$\\times$10° grid box, the algorithm with the grid box mean closest to the median is selected. The list of the considered retrieval algorithms is available from [[7]](https://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PUGS_GHG_main_latest.pdf). The various retrieval algorithms are optimized for different instruments (SCIAMACHY, GOSAT, GOSAT-2), which measure backscattered solar radiation in the near-infrared for O$_2$ as well as the absorption bands of CO$_2$ or CH$_4$.\n\nThe code is included for transparency and learning purposes, giving users the chance to adapt it for their own analyses. Users should always refer to the official reports for scientific assessments related to atmospheric CH$_4$ variability (e.g., [[9]](https://doi.org/10.24381/14j9-s541), [[10]](https://climate.copernicus.eu/global-climate-highlights-2024)).\n\n## 📢 Quality assessment statements\n\nThese are the key outcomes of this assessment"} {"chunk_id": "satellite_satellite-methane_completeness_q04__0fa1be9f2d0c", "report_id": "satellite_satellite-methane_completeness_q04", "dataset_id": "satellite-methane", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q04", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Methane satellite observations completeness assessment for greenhouse gas monitoring", "title": "Methane satellite observations completeness assessment for greenhouse gas monitoring", "chunk_index": 1, "token_count": 1098, "text_raw": "541), [[10]](https://climate.copernicus.eu/global-climate-highlights-2024)).\n\n## 📢 Quality assessment statements\n\nThese are the key outcomes of this assessment\n\n* The dataset \"Methane data from 2002 to present derived from satellite observations\" (deprecated EMMA Level 2) can be used to evaluate XCH$_4$ global and zonal growth rates.\n* For the high-latitude zonal bands ([70 °N-90 °N] and [90 °S-70 °S]), the few available data prevent a solid quantification of annual growth rates. In addition, for the latitude bands [70 °S-50 °S] and [50 °N-70 °N], the monthly growth rates before 2009 should be treated with extreme caution, as they are affected by non-uniform data coverage and high uncertainties in the data.\n* Before 2009, the global annual growth rates are characterised by higher standard deviations and thus must be considered with caution.\n* The different averaging methods do not particularly affect the growth rate distributions and growth rate anomalies, both globally and zonally.\n* As the EMMA Level 2 XCH$_4$ data product will not be updated in the future, users can use the Level 3 XCH4_OBS4MIPS and MTCH4_OBS4MIPS data products to investigate global and zonal methane growth rates.\n```\n\n## 📋 Methodology\n\nIn this notebook, we investigate whether the merged EMMA Level 2 product is suitable for calculating XCH$_4$ growth rates, over 2003-2022. The input data are the XCH$_4$ merged EMMA Level 2 files (version 4.5, see [[7]](https://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PUGS_GHG_main_latest.pdf)). For the calculations, and following the approach presented in [[4]](https://doi.org/10.5194/acp-24-577-2024), the merged EMMA Level 2 data were first regridded onto a daily regular 2°$\\times$2° grid, using spatially weighted mean. For the years 2003 and 2022, the annual growth rates were calculated over six-month periods (June–December and January–June, respectively), as the calculation method required rolling operations over a 12-month time window.\n\nIn this analysis, to estimate XCH$_4$ global increases and zonal growth rates (in 20° latitude bands), we calculated the monthly growth rates as the centered rolling difference between the XCH$_4$ value of month $i$ and the XCH$_4$ value of month $i$-12, following the approach presented in [[11]](https://doi.org/10.5194/acp-18-17355-2018) and also further used in [[6]](https://doi.org/10.5194/amt-13-789-2020). As done in [[4]](https://doi.org/10.5194/acp-24-577-2024), we also compare two different methods for calculating monthly averages, i.e.: \n* the \"standard averaging\", defined as the area-weighted mean of all grid cells in the region of interest;\n* the \"zonal-first averaging\", which is computed by first evaluating the $m$ averages for all 2° latitude bands, and then averaging these $m$ zonal averages. $m$ can be equal to 90 for global data, or, e.g., 10 for a 20° zonal band.\n\nThe zonal-first average should ensure a consistent weighting of all the latitudes, regardless of their individual coverage ([[4]](https://doi.org/10.5194/acp-24-577-2024)).\n\nFor better investigating the differences between the hemispheres concerning growth rates, we also computed growth rate anomalies, defined as the difference between the zonal and the global growth rates.\n\nThe analysis and results are organized in the following steps, which are detailed in the sections below:\n\n**[](template:section-1)**\n * Import all the relevant packages.\n * Choose the temporal coverage for the analysis.\n * Cache needed functions.\n\n**[](template:section-2)**\n * In this section, we define the data request to CDS.\n\n**[](template:section-3)**\n\nThis section presents several results for the XCH$_4$ growth rates, i.e.:\n * the calculation of global XCH$_4$ increases\n * the calculation of the zonal XCH$_4$ growth rates (into 20° latitude bands), to provide spatial information on the global CH$_4$ increases\n * the computation of the zonal growth rate anomalies, defined as the differences between zonal and global growth rates, for the time periods 2003-2022 and 2018-2022", "text_with_prefix": "EQC Quality Assessment: \"Methane satellite observations completeness assessment for greenhouse gas monitoring\"\nDataset: satellite-methane [CDS]\nAspect: completeness_q04 | Category: Satellite_ECVs\nSection: Methane satellite observations completeness assessment for greenhouse gas monitoring\n---\n541), [[10]](https://climate.copernicus.eu/global-climate-highlights-2024)).\n\n## 📢 Quality assessment statements\n\nThese are the key outcomes of this assessment\n\n* The dataset \"Methane data from 2002 to present derived from satellite observations\" (deprecated EMMA Level 2) can be used to evaluate XCH$_4$ global and zonal growth rates.\n* For the high-latitude zonal bands ([70 °N-90 °N] and [90 °S-70 °S]), the few available data prevent a solid quantification of annual growth rates. In addition, for the latitude bands [70 °S-50 °S] and [50 °N-70 °N], the monthly growth rates before 2009 should be treated with extreme caution, as they are affected by non-uniform data coverage and high uncertainties in the data.\n* Before 2009, the global annual growth rates are characterised by higher standard deviations and thus must be considered with caution.\n* The different averaging methods do not particularly affect the growth rate distributions and growth rate anomalies, both globally and zonally.\n* As the EMMA Level 2 XCH$_4$ data product will not be updated in the future, users can use the Level 3 XCH4_OBS4MIPS and MTCH4_OBS4MIPS data products to investigate global and zonal methane growth rates.\n```\n\n## 📋 Methodology\n\nIn this notebook, we investigate whether the merged EMMA Level 2 product is suitable for calculating XCH$_4$ growth rates, over 2003-2022. The input data are the XCH$_4$ merged EMMA Level 2 files (version 4.5, see [[7]](https://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PUGS_GHG_main_latest.pdf)). For the calculations, and following the approach presented in [[4]](https://doi.org/10.5194/acp-24-577-2024), the merged EMMA Level 2 data were first regridded onto a daily regular 2°$\\times$2° grid, using spatially weighted mean. For the years 2003 and 2022, the annual growth rates were calculated over six-month periods (June–December and January–June, respectively), as the calculation method required rolling operations over a 12-month time window.\n\nIn this analysis, to estimate XCH$_4$ global increases and zonal growth rates (in 20° latitude bands), we calculated the monthly growth rates as the centered rolling difference between the XCH$_4$ value of month $i$ and the XCH$_4$ value of month $i$-12, following the approach presented in [[11]](https://doi.org/10.5194/acp-18-17355-2018) and also further used in [[6]](https://doi.org/10.5194/amt-13-789-2020). As done in [[4]](https://doi.org/10.5194/acp-24-577-2024), we also compare two different methods for calculating monthly averages, i.e.: \n* the \"standard averaging\", defined as the area-weighted mean of all grid cells in the region of interest;\n* the \"zonal-first averaging\", which is computed by first evaluating the $m$ averages for all 2° latitude bands, and then averaging these $m$ zonal averages. $m$ can be equal to 90 for global data, or, e.g., 10 for a 20° zonal band.\n\nThe zonal-first average should ensure a consistent weighting of all the latitudes, regardless of their individual coverage ([[4]](https://doi.org/10.5194/acp-24-577-2024)).\n\nFor better investigating the differences between the hemispheres concerning growth rates, we also computed growth rate anomalies, defined as the difference between the zonal and the global growth rates.\n\nThe analysis and results are organized in the following steps, which are detailed in the sections below:\n\n**[](template:section-1)**\n * Import all the relevant packages.\n * Choose the temporal coverage for the analysis.\n * Cache needed functions.\n\n**[](template:section-2)**\n * In this section, we define the data request to CDS.\n\n**[](template:section-3)**\n\nThis section presents several results for the XCH$_4$ growth rates, i.e.:\n * the calculation of global XCH$_4$ increases\n * the calculation of the zonal XCH$_4$ growth rates (into 20° latitude bands), to provide spatial information on the global CH$_4$ increases\n * the computation of the zonal growth rate anomalies, defined as the differences between zonal and global growth rates, for the time periods 2003-2022 and 2018-2022"} {"chunk_id": "satellite_satellite-methane_completeness_q04__d4ba96b822c8", "report_id": "satellite_satellite-methane_completeness_q04", "dataset_id": "satellite-methane", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q04", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Methane satellite observations completeness assessment for greenhouse gas monitoring", "title": "Methane satellite observations completeness assessment for greenhouse gas monitoring", "chunk_index": 2, "token_count": 1057, "text_raw": "information on the global CH$_4$ increases\n * the computation of the zonal growth rate anomalies, defined as the differences between zonal and global growth rates, for the time periods 2003-2022 and 2018-2022\n\n## 📈 Analysis and results\n\n(template:section-1)=\n### 1. Choose the data to use and set-up the code\n\n#### Import all relevant packages\nIn this section, we import all the relevant packages needed for running the notebook and we define the style for plot appearance (\"seaborn-v0_8-notebook\").\n\n#### Choose the temporal coverage\nIn this section, we define the temporal period of the analysis, that can be customized by the user. For this notebook, we selected the 2003-2022 period.\n\n#### Cache needed functions\nIn this section, we cached a list of functions used in the analyses.\n\n- The functions `spatial_weighted_mean` and `spatial_weighted_std`compute the spatial XCH$_4$ mean and standard deviation, respectively, over the selected domain. Through the function named `weight_dataset`, they use spatial weighting to account for the latitudinal dependence of the grid size in the lon/lat grids. This is a common approach, e.g., in the treatment of reanalysis and forecast model data.\n\n- The function `regrid` regrids the XCH$_4$ Level 2 data onto a regular $m \\times n$ grid, where $m$ and $n$ can be customized by the user. It uses `spatial_weighted_mean` for grouping data onto the regular grid.\n\n- The function `daily_regrid` uses the aforementioned `regrid` function for regridding the XCH$_4$ data onto a regular $m$°$\\times$$n$° grid, with a daily time resolution. Again, $m$ and $n$ can be customized by the user.\n\n- The function `monthly_regrid_in_bands` resamples the input dataset into $n$° latitude bands, by using the `regrid` function and a monthly time resolution. $n$ is a parameter that can be specified by the user within the function. If the option `zonal_first` is selected, then the zonal-first averaging is calculated, following the methodology presented in [[4]](https://doi.org/10.5194/acp-24-577-2024) and described above.\n\n- The function `compute_growth_rate` calculates the monthly resolved annual growth rates of XCH$_4$, computed as the centered rolling difference of the XCH$_4$ value of month $i$ minus the XCH$_4$ value of month $i$-12.\n\n- The function `band_from_central_latitude` is used for explicitly specifying the latitude extremes rather than the central latitude for each band.\n\n(template:section-2)=\n### 2. Retrieve XCH$_4$ data (EMMA, Level 2)\nIn this section, we define the data request to CDS (merged EMMA data product, Level 2, version 4.5, XCH$_4$) and download the dataset.\n\n```text\n100%|██████████| 20/20 [00:19<00:00, 1.01it/s]\n```\n\n(template:section-3)=\n### 3. Data analysis\n\n#### Global annual XCH$_4$ increases\nIn this section, we present the global annual growth rates of XCH$_4$, defined as the average of all monthly growth rates within each calendar year. Please note that the annual growth rate for 2022 is obtained by averaging the available monthly values for January to June, as the data for the first six months of 2023 are not available. For the calculation of growth rates, two different averaging methods were applied to the regridded (onto a daily regular 2°$\\times$2° grid) Level 2 data, following the methodology presented in [[4]](https://doi.org/10.5194/acp-24-577-2024). The \"standard averaging\" is defined as the spatially weighted average of all grid cells in a given region. This was compared to the so-called \"zonal-first averaging\", which is computed as follows: (i) first, the 90 averages for all 2° latitude bands are computed; then (ii) the average of these 90 zonal averages is calculated, yielding each monthly value of the global time series. As shown in [[4]](https://doi.org/10.5194/acp-24-577-2024), the zonal-first averaging has the advantage of taking into account the inhomogeneous sampling at each latitude, which is influenced by both the distribution of land mass and the seasonal coverage of measurements.", "text_with_prefix": "EQC Quality Assessment: \"Methane satellite observations completeness assessment for greenhouse gas monitoring\"\nDataset: satellite-methane [CDS]\nAspect: completeness_q04 | Category: Satellite_ECVs\nSection: Methane satellite observations completeness assessment for greenhouse gas monitoring\n---\ninformation on the global CH$_4$ increases\n * the computation of the zonal growth rate anomalies, defined as the differences between zonal and global growth rates, for the time periods 2003-2022 and 2018-2022\n\n## 📈 Analysis and results\n\n(template:section-1)=\n### 1. Choose the data to use and set-up the code\n\n#### Import all relevant packages\nIn this section, we import all the relevant packages needed for running the notebook and we define the style for plot appearance (\"seaborn-v0_8-notebook\").\n\n#### Choose the temporal coverage\nIn this section, we define the temporal period of the analysis, that can be customized by the user. For this notebook, we selected the 2003-2022 period.\n\n#### Cache needed functions\nIn this section, we cached a list of functions used in the analyses.\n\n- The functions `spatial_weighted_mean` and `spatial_weighted_std`compute the spatial XCH$_4$ mean and standard deviation, respectively, over the selected domain. Through the function named `weight_dataset`, they use spatial weighting to account for the latitudinal dependence of the grid size in the lon/lat grids. This is a common approach, e.g., in the treatment of reanalysis and forecast model data.\n\n- The function `regrid` regrids the XCH$_4$ Level 2 data onto a regular $m \\times n$ grid, where $m$ and $n$ can be customized by the user. It uses `spatial_weighted_mean` for grouping data onto the regular grid.\n\n- The function `daily_regrid` uses the aforementioned `regrid` function for regridding the XCH$_4$ data onto a regular $m$°$\\times$$n$° grid, with a daily time resolution. Again, $m$ and $n$ can be customized by the user.\n\n- The function `monthly_regrid_in_bands` resamples the input dataset into $n$° latitude bands, by using the `regrid` function and a monthly time resolution. $n$ is a parameter that can be specified by the user within the function. If the option `zonal_first` is selected, then the zonal-first averaging is calculated, following the methodology presented in [[4]](https://doi.org/10.5194/acp-24-577-2024) and described above.\n\n- The function `compute_growth_rate` calculates the monthly resolved annual growth rates of XCH$_4$, computed as the centered rolling difference of the XCH$_4$ value of month $i$ minus the XCH$_4$ value of month $i$-12.\n\n- The function `band_from_central_latitude` is used for explicitly specifying the latitude extremes rather than the central latitude for each band.\n\n(template:section-2)=\n### 2. Retrieve XCH$_4$ data (EMMA, Level 2)\nIn this section, we define the data request to CDS (merged EMMA data product, Level 2, version 4.5, XCH$_4$) and download the dataset.\n\n```text\n100%|██████████| 20/20 [00:19<00:00, 1.01it/s]\n```\n\n(template:section-3)=\n### 3. Data analysis\n\n#### Global annual XCH$_4$ increases\nIn this section, we present the global annual growth rates of XCH$_4$, defined as the average of all monthly growth rates within each calendar year. Please note that the annual growth rate for 2022 is obtained by averaging the available monthly values for January to June, as the data for the first six months of 2023 are not available. For the calculation of growth rates, two different averaging methods were applied to the regridded (onto a daily regular 2°$\\times$2° grid) Level 2 data, following the methodology presented in [[4]](https://doi.org/10.5194/acp-24-577-2024). The \"standard averaging\" is defined as the spatially weighted average of all grid cells in a given region. This was compared to the so-called \"zonal-first averaging\", which is computed as follows: (i) first, the 90 averages for all 2° latitude bands are computed; then (ii) the average of these 90 zonal averages is calculated, yielding each monthly value of the global time series. As shown in [[4]](https://doi.org/10.5194/acp-24-577-2024), the zonal-first averaging has the advantage of taking into account the inhomogeneous sampling at each latitude, which is influenced by both the distribution of land mass and the seasonal coverage of measurements."} {"chunk_id": "satellite_satellite-methane_completeness_q04__14bd5dcf04a2", "report_id": "satellite_satellite-methane_completeness_q04", "dataset_id": "satellite-methane", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q04", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Methane satellite observations completeness assessment for greenhouse gas monitoring", "title": "Methane satellite observations completeness assessment for greenhouse gas monitoring", "chunk_index": 3, "token_count": 1071, "text_raw": "p-24-577-2024), the zonal-first averaging has the advantage of taking into account the inhomogeneous sampling at each latitude, which is influenced by both the distribution of land mass and the seasonal coverage of measurements.\n\nIn agreement with recent studies (e.g., [[12]](https://doi.org/10.5194/acp-23-4863-2023)), the behavior of the global XCH$_4$ growth rate has shown a near-zero growth until 2007, followed by a clear acceleration in subsequent years due to increased emissions from wetlands in the tropics and from anthropogenic sources at the mid-latitudes of the Northern Hemisphere ([[1]](https://library.wmo.int/idurl/4/68532)), with maxima observed in the last three years of the time series. The highest global annual XCH$_4$ increases were recorded in 2021, followed by 2022 and 2020. According to [[12]](https://doi.org/10.5194/acp-23-4863-2023), most of the observed increase in atmospheric XCH$_4$ during 2020 and 2021 was due to increased emissions, with a significant contribution from reduced levels of OH. A recent study ([[13]](https://doi.org/10.5194/acp-25-6757-2025)) suggests a prominent role of biogenic emissions in Asian regions for the persistent high growth rate observed during 2020–2022. The results obtained using the two different averaging methods are comparable for all years, indicating that the choice of the averaging method does not significantly affect the estimated growth rates.\n\nThe global growth rates obtained by the EMMA data product are in very good agreement with those provided by the global network of in-situ observations coordinated by the Global Atmosphere Watch (GAW) programme by the World Meteorological Organization (WMO)[[1]](https://library.wmo.int/idurl/4/68532). Here, for comparison, we provided a plot of the WMO/GAW global growth rates as obtained by averaging the monthly values provided by the World Data Center for Greenhouse Gases (WDCGG)[[14]](https://gaw.kishou.go.jp/publications/global_mean_mole_fractions). The general variability of the annual CH$_4$ growth rates is generally consistent between the two datasets. The mean differences are 0.20 ppb/yr and 0.23 ppb/yr for the \"standard\" and \"zonal-first\" averaging methods, respectively. For the \"standard averaging\" method, only four years and for the \"zonal-first averaging\" method, only five years, reported differences exceeding ±2 ppb/yr. We should keep in mind that perfect agreement is not to be expected, as the EMMA product is column-averaged CH$_4$, whereas WMO/GAW used near-surface data.\n\nIn the XCH$_4$ plot, also uncertainties are reported, and these are computed as the annual standard deviation of all monthly growth rates within each calendar year. The most evident feature is the overall decrease in the uncertainties after 2009, likely due to the introduction of algorithms based on GOSAT observations used to calculate the median XCH$_4$ data [[8]](https://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_ATBD_GHG_D_latest.pdf).\n\n*Global annual XCH$_4$ increases over 2003-2022. The colors differentiate between the standard averaging (orange) and the zonal-first averaging (blue) methods. Bars indicate the uncertainties in the annual increases, computed as $\\pm$1 standard deviation. Please note that the annual growth rates in 2003 and 2022 were obtained by averaging only six monthly values (June-December and January-June, respectively).*\n\n![GR_CH4_WDCGG_GLOBAL_REV.png](attachment:110aaca5-28c5-4237-aad0-4e987ef7c0e2.png)\n\n*Global annual CH$_4$ increases over 2003-2022 as obtained by the global in-situ dataset provided by the World Data Center for Greenhouse Gases [[13]](https://doi.org/10.5194/acp-25-6757-2025). Please note that the annual growth rates in 2003 and 2022 were obtained by averaging only six monthly values (June-December and January-June, respectively). The error bars show the standard deviation of the monthly growth rate values.*\n\n#### Zonal growth rates of XCH$_4$ (2003-2022)\nIn this section, we calculate and report the monthly growth rates of XCH$_4$, divided into 20° latitude bands, to provide spatial context for the global growth rates presented above.", "text_with_prefix": "EQC Quality Assessment: \"Methane satellite observations completeness assessment for greenhouse gas monitoring\"\nDataset: satellite-methane [CDS]\nAspect: completeness_q04 | Category: Satellite_ECVs\nSection: Methane satellite observations completeness assessment for greenhouse gas monitoring\n---\np-24-577-2024), the zonal-first averaging has the advantage of taking into account the inhomogeneous sampling at each latitude, which is influenced by both the distribution of land mass and the seasonal coverage of measurements.\n\nIn agreement with recent studies (e.g., [[12]](https://doi.org/10.5194/acp-23-4863-2023)), the behavior of the global XCH$_4$ growth rate has shown a near-zero growth until 2007, followed by a clear acceleration in subsequent years due to increased emissions from wetlands in the tropics and from anthropogenic sources at the mid-latitudes of the Northern Hemisphere ([[1]](https://library.wmo.int/idurl/4/68532)), with maxima observed in the last three years of the time series. The highest global annual XCH$_4$ increases were recorded in 2021, followed by 2022 and 2020. According to [[12]](https://doi.org/10.5194/acp-23-4863-2023), most of the observed increase in atmospheric XCH$_4$ during 2020 and 2021 was due to increased emissions, with a significant contribution from reduced levels of OH. A recent study ([[13]](https://doi.org/10.5194/acp-25-6757-2025)) suggests a prominent role of biogenic emissions in Asian regions for the persistent high growth rate observed during 2020–2022. The results obtained using the two different averaging methods are comparable for all years, indicating that the choice of the averaging method does not significantly affect the estimated growth rates.\n\nThe global growth rates obtained by the EMMA data product are in very good agreement with those provided by the global network of in-situ observations coordinated by the Global Atmosphere Watch (GAW) programme by the World Meteorological Organization (WMO)[[1]](https://library.wmo.int/idurl/4/68532). Here, for comparison, we provided a plot of the WMO/GAW global growth rates as obtained by averaging the monthly values provided by the World Data Center for Greenhouse Gases (WDCGG)[[14]](https://gaw.kishou.go.jp/publications/global_mean_mole_fractions). The general variability of the annual CH$_4$ growth rates is generally consistent between the two datasets. The mean differences are 0.20 ppb/yr and 0.23 ppb/yr for the \"standard\" and \"zonal-first\" averaging methods, respectively. For the \"standard averaging\" method, only four years and for the \"zonal-first averaging\" method, only five years, reported differences exceeding ±2 ppb/yr. We should keep in mind that perfect agreement is not to be expected, as the EMMA product is column-averaged CH$_4$, whereas WMO/GAW used near-surface data.\n\nIn the XCH$_4$ plot, also uncertainties are reported, and these are computed as the annual standard deviation of all monthly growth rates within each calendar year. The most evident feature is the overall decrease in the uncertainties after 2009, likely due to the introduction of algorithms based on GOSAT observations used to calculate the median XCH$_4$ data [[8]](https://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_ATBD_GHG_D_latest.pdf).\n\n*Global annual XCH$_4$ increases over 2003-2022. The colors differentiate between the standard averaging (orange) and the zonal-first averaging (blue) methods. Bars indicate the uncertainties in the annual increases, computed as $\\pm$1 standard deviation. Please note that the annual growth rates in 2003 and 2022 were obtained by averaging only six monthly values (June-December and January-June, respectively).*\n\n![GR_CH4_WDCGG_GLOBAL_REV.png](attachment:110aaca5-28c5-4237-aad0-4e987ef7c0e2.png)\n\n*Global annual CH$_4$ increases over 2003-2022 as obtained by the global in-situ dataset provided by the World Data Center for Greenhouse Gases [[13]](https://doi.org/10.5194/acp-25-6757-2025). Please note that the annual growth rates in 2003 and 2022 were obtained by averaging only six monthly values (June-December and January-June, respectively). The error bars show the standard deviation of the monthly growth rate values.*\n\n#### Zonal growth rates of XCH$_4$ (2003-2022)\nIn this section, we calculate and report the monthly growth rates of XCH$_4$, divided into 20° latitude bands, to provide spatial context for the global growth rates presented above."} {"chunk_id": "satellite_satellite-methane_completeness_q04__3220bc9fd9a7", "report_id": "satellite_satellite-methane_completeness_q04", "dataset_id": "satellite-methane", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q04", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Methane satellite observations completeness assessment for greenhouse gas monitoring", "title": "Methane satellite observations completeness assessment for greenhouse gas monitoring", "chunk_index": 4, "token_count": 976, "text_raw": "$_4$ (2003-2022)\nIn this section, we calculate and report the monthly growth rates of XCH$_4$, divided into 20° latitude bands, to provide spatial context for the global growth rates presented above.\n\nThe figure shows the zonal growth rates for the nine 20° latitude bands, using the two different averaging methods described previously. Please note that the limits on the y-axis can be customized by the user.\n\nFor most latitude bands, the behavior is similar to that observed for the global growth rates. The [70 °N-90 °N] band, however, is characterized by almost no clear tendency throughout the years, and this is because the high-latitude regions are characterized by sparse sampling and higher uncertainties in the data. Data below 60 °S are only available until 2005, and this is the cause for almost completely missing data in the [70 °S-90 °S] latitude band ([[7]](https://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PUGS_GHG_main_latest.pdf)). For the latitude bands [70 °S-50 °S] and [50 °N-70 °N], the monthly growth rates are characterised by a higher temporal variability before March 2009, as data over land is only available from the SCIAMACHY WFMD product in this period ([[7]](https://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PUGS_GHG_main_latest.pdf)).\n\nThe two different averaging methods show better agreement after 2009, with nearly identical results across all zonal bands. Before 2009, the growth rates exhibit higher variability, with larger differences between standard and zonal-first averaging. The improvement of the agreement between the two calculation methods, likely reflects the introduction, after 2009, of algorithms incorporating GOSAT observations to calculate median XCH$_4$ values [[8]](https://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_ATBD_GHG_D_latest.pdf). This highlights that caution should be exercised when using the EMMA data product to derive zonal growth rates prior to 2009 and that results are sensitive to the calculation method adopted.\n\n*Zonal growth rates of XCH$_4$, over 2003-2022. The panels indicate the 20° latitude bands selected for the analysis. The colors differentiate between the standard averaging (orange) and the zonal-first averaging (blue) methods.*\n\n#### Zonal growth rates of XCH$_4$ (2018-2022)\nTo better focus on the more recent years, characterised by the highest CH$_4$ growth rates, in the following we provide a zoom on the monthly growth rates of XCH$_4$ from 2018 to 2022.\nInterestingly, as also reported in [[4]](https://doi.org/10.5194/acp-24-577-2024), the reduction in global XCH$_4$ growth in 2022 with respect to 2021 can be attributed to decreased growth rates in the Northern Hemisphere (particularly evident in the [10 °N-30 °N] and [30 °N-50 °N] latitude bands), while growth rates in the Southern Hemisphere remain high.\n\n*Zonal growth rates of XCH$_4$, over 2018-2022. The panels indicate the 20° latitude bands selected for the analysis. The colors differentiate between the standard averaging (orange) and the zonal-first averaging (blue) methods.*\n\n#### Zonal growth rate anomalies of XCH$_4$\nIn this section, we calculate and report the zonal growth rate anomalies, defined as the difference between the zonal and global growth rates for each of the 20° latitude bands presented above. As also indicated in [[4]](https://doi.org/10.5194/acp-24-577-2024), the zonal growth rate anomalies can help in better visualizing differences between the hemispheres. Moreover, the anomalies, together with the zonal growth rates, are important for better interpreting the changes in the global methane increases.\n\nAs already observed for the zonal growth rates presented above, the choice of the averaging method does not significantly affect the calculated anomalies.", "text_with_prefix": "EQC Quality Assessment: \"Methane satellite observations completeness assessment for greenhouse gas monitoring\"\nDataset: satellite-methane [CDS]\nAspect: completeness_q04 | Category: Satellite_ECVs\nSection: Methane satellite observations completeness assessment for greenhouse gas monitoring\n---\n$_4$ (2003-2022)\nIn this section, we calculate and report the monthly growth rates of XCH$_4$, divided into 20° latitude bands, to provide spatial context for the global growth rates presented above.\n\nThe figure shows the zonal growth rates for the nine 20° latitude bands, using the two different averaging methods described previously. Please note that the limits on the y-axis can be customized by the user.\n\nFor most latitude bands, the behavior is similar to that observed for the global growth rates. The [70 °N-90 °N] band, however, is characterized by almost no clear tendency throughout the years, and this is because the high-latitude regions are characterized by sparse sampling and higher uncertainties in the data. Data below 60 °S are only available until 2005, and this is the cause for almost completely missing data in the [70 °S-90 °S] latitude band ([[7]](https://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PUGS_GHG_main_latest.pdf)). For the latitude bands [70 °S-50 °S] and [50 °N-70 °N], the monthly growth rates are characterised by a higher temporal variability before March 2009, as data over land is only available from the SCIAMACHY WFMD product in this period ([[7]](https://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PUGS_GHG_main_latest.pdf)).\n\nThe two different averaging methods show better agreement after 2009, with nearly identical results across all zonal bands. Before 2009, the growth rates exhibit higher variability, with larger differences between standard and zonal-first averaging. The improvement of the agreement between the two calculation methods, likely reflects the introduction, after 2009, of algorithms incorporating GOSAT observations to calculate median XCH$_4$ values [[8]](https://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_ATBD_GHG_D_latest.pdf). This highlights that caution should be exercised when using the EMMA data product to derive zonal growth rates prior to 2009 and that results are sensitive to the calculation method adopted.\n\n*Zonal growth rates of XCH$_4$, over 2003-2022. The panels indicate the 20° latitude bands selected for the analysis. The colors differentiate between the standard averaging (orange) and the zonal-first averaging (blue) methods.*\n\n#### Zonal growth rates of XCH$_4$ (2018-2022)\nTo better focus on the more recent years, characterised by the highest CH$_4$ growth rates, in the following we provide a zoom on the monthly growth rates of XCH$_4$ from 2018 to 2022.\nInterestingly, as also reported in [[4]](https://doi.org/10.5194/acp-24-577-2024), the reduction in global XCH$_4$ growth in 2022 with respect to 2021 can be attributed to decreased growth rates in the Northern Hemisphere (particularly evident in the [10 °N-30 °N] and [30 °N-50 °N] latitude bands), while growth rates in the Southern Hemisphere remain high.\n\n*Zonal growth rates of XCH$_4$, over 2018-2022. The panels indicate the 20° latitude bands selected for the analysis. The colors differentiate between the standard averaging (orange) and the zonal-first averaging (blue) methods.*\n\n#### Zonal growth rate anomalies of XCH$_4$\nIn this section, we calculate and report the zonal growth rate anomalies, defined as the difference between the zonal and global growth rates for each of the 20° latitude bands presented above. As also indicated in [[4]](https://doi.org/10.5194/acp-24-577-2024), the zonal growth rate anomalies can help in better visualizing differences between the hemispheres. Moreover, the anomalies, together with the zonal growth rates, are important for better interpreting the changes in the global methane increases.\n\nAs already observed for the zonal growth rates presented above, the choice of the averaging method does not significantly affect the calculated anomalies."} {"chunk_id": "satellite_satellite-methane_completeness_q04__311ed388b0ef", "report_id": "satellite_satellite-methane_completeness_q04", "dataset_id": "satellite-methane", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q04", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Methane satellite observations completeness assessment for greenhouse gas monitoring", "title": "Methane satellite observations completeness assessment for greenhouse gas monitoring", "chunk_index": 5, "token_count": 648, "text_raw": "onal growth rates, are important for better interpreting the changes in the global methane increases.\n\nAs already observed for the zonal growth rates presented above, the choice of the averaging method does not significantly affect the calculated anomalies.\n\nFrom the heatmaps of the anomalies, several features are consistent with previous studies (e.g., [[4]](https://doi.org/10.5194/acp-24-577-2024)), indicating that the EMMA Level 2 dataset can be used to evaluate the time evolution of CH$_4$ zonal growth rates. In 2019, a slowdown in growth rates was observed in the Southern Hemisphere, while growth rates in the Northern Hemisphere remained positive. This was followed by predominantly positive anomalies until 2022. In that year, as already noted for the zonal growth rates, the Northern Hemisphere showed a marked slowdown in XCH$_4$ increases, as reflected by the negative anomalies.\n\n*Zonal growth rate anomalies of XCH$_4$, for the 20° latitude bands presented above, over 2003-2022. The anomalies are defined as the differences between zonal and global growth rates. The top panel is for the standard averaging, while the bottom panel is for the zonal-first averaging.*\n\n## ℹ️ If you want to know more\n\n### Key resources\n\nThe CDS catalogue entries for the data used were:\n* Methane data from 2002 to present derived from satellite observations: https://cds.climate.copernicus.eu/datasets/satellite-methane?tab=overview\n\nUsers interested in near-real time detection of hot-spot locations for methane emissions can consider to use the CAMS Methane Hotspot Explorer: https://atmosphere.copernicus.eu/ghg-services/cams-methane-hotspot-explorer?utm_source=press&utm_medium=referral&utm_campaign=CH4-app-2025\n\nCode libraries used:\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)\n\n### References\n[[1]](https://library.wmo.int/idurl/4/68532) World Meteorological Organization (2023). WMO Greenhouse Gas Bulletin, No. 19, ISSN 2078-0796.\n\n[[2]](https://doi.org/10.1073/pnas.0600201103) West, J. J., Fiore, A. M., Horowitz, L. W., and Mauzerall, D. L. (2006). Global health benefits of mitigating ozone pollution with methane emission controls, Proceedings of the National Academy of Sciences USA, 103, 3988–3993.", "text_with_prefix": "EQC Quality Assessment: \"Methane satellite observations completeness assessment for greenhouse gas monitoring\"\nDataset: satellite-methane [CDS]\nAspect: completeness_q04 | Category: Satellite_ECVs\nSection: Methane satellite observations completeness assessment for greenhouse gas monitoring\n---\nonal growth rates, are important for better interpreting the changes in the global methane increases.\n\nAs already observed for the zonal growth rates presented above, the choice of the averaging method does not significantly affect the calculated anomalies.\n\nFrom the heatmaps of the anomalies, several features are consistent with previous studies (e.g., [[4]](https://doi.org/10.5194/acp-24-577-2024)), indicating that the EMMA Level 2 dataset can be used to evaluate the time evolution of CH$_4$ zonal growth rates. In 2019, a slowdown in growth rates was observed in the Southern Hemisphere, while growth rates in the Northern Hemisphere remained positive. This was followed by predominantly positive anomalies until 2022. In that year, as already noted for the zonal growth rates, the Northern Hemisphere showed a marked slowdown in XCH$_4$ increases, as reflected by the negative anomalies.\n\n*Zonal growth rate anomalies of XCH$_4$, for the 20° latitude bands presented above, over 2003-2022. The anomalies are defined as the differences between zonal and global growth rates. The top panel is for the standard averaging, while the bottom panel is for the zonal-first averaging.*\n\n## ℹ️ If you want to know more\n\n### Key resources\n\nThe CDS catalogue entries for the data used were:\n* Methane data from 2002 to present derived from satellite observations: https://cds.climate.copernicus.eu/datasets/satellite-methane?tab=overview\n\nUsers interested in near-real time detection of hot-spot locations for methane emissions can consider to use the CAMS Methane Hotspot Explorer: https://atmosphere.copernicus.eu/ghg-services/cams-methane-hotspot-explorer?utm_source=press&utm_medium=referral&utm_campaign=CH4-app-2025\n\nCode libraries used:\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)\n\n### References\n[[1]](https://library.wmo.int/idurl/4/68532) World Meteorological Organization (2023). WMO Greenhouse Gas Bulletin, No. 19, ISSN 2078-0796.\n\n[[2]](https://doi.org/10.1073/pnas.0600201103) West, J. J., Fiore, A. M., Horowitz, L. W., and Mauzerall, D. L. (2006). Global health benefits of mitigating ozone pollution with methane emission controls, Proceedings of the National Academy of Sciences USA, 103, 3988–3993."} {"chunk_id": "satellite_satellite-methane_completeness_q04__2c81f3dd6868", "report_id": "satellite_satellite-methane_completeness_q04", "dataset_id": "satellite-methane", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q04", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Methane satellite observations completeness assessment for greenhouse gas monitoring", "title": "Methane satellite observations completeness assessment for greenhouse gas monitoring", "chunk_index": 6, "token_count": 1041, "text_raw": "ating ozone pollution with methane emission controls, Proceedings of the National Academy of Sciences USA, 103, 3988–3993.\n\n[[3]](https://doi.org/10.5194/essd-17-1873-2025) Saunois, M., Martinez, A., Poulter, B., Zhang, Z., Raymond, P. A., Regnier, P., Canadell, J. G., Jackson, R. B., Patra, P. K., Bousquet, P., Ciais, P., Dlugokencky, E. J., Lan, X., Allen, G. H., Bastviken, D., Beerling, D. J., Belikov, D. A., Blake, D. R., Castaldi, S., Crippa, M., Deemer, B. R., Dennison, F., Etiope, G., Gedney, N., Höglund-Isaksson, L., Holgerson, M. A., Hopcroft, P. O., Hugelius, G., Ito, A., Jain, A. K., Janardanan, R., Johnson, M. S., Kleinen, T., Krummel, P. B., Lauerwald, R., Li, T., Liu, X., McDonald, K. C., Melton, J. R., Mühle, J., Müller, J., Murguia-Flores, F., Niwa, Y., Noce, S., Pan, S., Parker, R. J., Peng, C., Ramonet, M., Riley, W. J., Rocher-Ros, G., Rosentreter, J. A., Sasakawa, M., Segers, A., Smith, S. J., Stanley, E. H., Thanwerdas, J., Tian, H., Tsuruta, A., Tubiello, F. N., Weber, T. S., van der Werf, G. R., Worthy, D. E. J., Xi, Y., Yoshida, Y., Zhang, W., Zheng, B., Zhu, Q., Zhu, Q., and Zhuang, Q. (2025). Global Methane Budget 2000–2020, Earth System Science Data, 17, 1873-1958.\n\n[[4]](https://doi.org/10.5194/acp-24-577-2024) Hachmeister, J., Schneising, O., Buchwitz, M., Burrows, J. P., Notholt, J., and Buschmann, M. (2024). Zonal variability of methane trends derived from satellite data, Atmospheric Chemistry and Physics, 24, 577–595.\n\n[[5]](https://doi.org/10.1038/s41558-023-01629-0) Zhang, Z., Poulter, B., Feldman, A.F., Ying, Q., Ciais, P., Peng, S., and Li, X. (2023). Recent intensification of wetland methane feedback, Nature Climate Change, 13, 430–433.\n\n[[6]](https://doi.org/10.5194/amt-13-789-2020) Reuter, M., Buchwitz, M., Schneising, O., Noël, S., Bovensmann, H., Burrows, J. P., Boesch, H., Di Noia, A., Anand, J., Parker, R. J., Somkuti, P., Wu, L., Hasekamp, O. P., Aben, I., Kuze, A., Suto, H., Shiomi, K., Yoshida, Y., Morino, I., Crisp, D., O'Dell, C. W., Notholt, J., Petri, C., Warneke, T., Velazco, V. A., Deutscher, N. M., Griffith, D. W. T., Kivi, R., Pollard, D. F., Hase, F., Sussmann, R., Te, Y. V., Strong, K., Roche, S., Sha, M. K., De Maziere, M., Feist, D. G., Iraci, L. T., Roehl, C. M., Retscher, C., and Schepers, D. (2020). Ensemble-based satellite-derived carbon dioxide and methane column-averaged dry-air mole fraction data sets (2003-2018) for carbon and climate applications, Atmospheric Measurement Techniques, 13, 789-819.", "text_with_prefix": "EQC Quality Assessment: \"Methane satellite observations completeness assessment for greenhouse gas monitoring\"\nDataset: satellite-methane [CDS]\nAspect: completeness_q04 | Category: Satellite_ECVs\nSection: Methane satellite observations completeness assessment for greenhouse gas monitoring\n---\nating ozone pollution with methane emission controls, Proceedings of the National Academy of Sciences USA, 103, 3988–3993.\n\n[[3]](https://doi.org/10.5194/essd-17-1873-2025) Saunois, M., Martinez, A., Poulter, B., Zhang, Z., Raymond, P. A., Regnier, P., Canadell, J. G., Jackson, R. B., Patra, P. K., Bousquet, P., Ciais, P., Dlugokencky, E. J., Lan, X., Allen, G. H., Bastviken, D., Beerling, D. J., Belikov, D. A., Blake, D. R., Castaldi, S., Crippa, M., Deemer, B. R., Dennison, F., Etiope, G., Gedney, N., Höglund-Isaksson, L., Holgerson, M. A., Hopcroft, P. O., Hugelius, G., Ito, A., Jain, A. K., Janardanan, R., Johnson, M. S., Kleinen, T., Krummel, P. B., Lauerwald, R., Li, T., Liu, X., McDonald, K. C., Melton, J. R., Mühle, J., Müller, J., Murguia-Flores, F., Niwa, Y., Noce, S., Pan, S., Parker, R. J., Peng, C., Ramonet, M., Riley, W. J., Rocher-Ros, G., Rosentreter, J. A., Sasakawa, M., Segers, A., Smith, S. J., Stanley, E. H., Thanwerdas, J., Tian, H., Tsuruta, A., Tubiello, F. N., Weber, T. S., van der Werf, G. R., Worthy, D. E. J., Xi, Y., Yoshida, Y., Zhang, W., Zheng, B., Zhu, Q., Zhu, Q., and Zhuang, Q. (2025). Global Methane Budget 2000–2020, Earth System Science Data, 17, 1873-1958.\n\n[[4]](https://doi.org/10.5194/acp-24-577-2024) Hachmeister, J., Schneising, O., Buchwitz, M., Burrows, J. P., Notholt, J., and Buschmann, M. (2024). Zonal variability of methane trends derived from satellite data, Atmospheric Chemistry and Physics, 24, 577–595.\n\n[[5]](https://doi.org/10.1038/s41558-023-01629-0) Zhang, Z., Poulter, B., Feldman, A.F., Ying, Q., Ciais, P., Peng, S., and Li, X. (2023). Recent intensification of wetland methane feedback, Nature Climate Change, 13, 430–433.\n\n[[6]](https://doi.org/10.5194/amt-13-789-2020) Reuter, M., Buchwitz, M., Schneising, O., Noël, S., Bovensmann, H., Burrows, J. P., Boesch, H., Di Noia, A., Anand, J., Parker, R. J., Somkuti, P., Wu, L., Hasekamp, O. P., Aben, I., Kuze, A., Suto, H., Shiomi, K., Yoshida, Y., Morino, I., Crisp, D., O'Dell, C. W., Notholt, J., Petri, C., Warneke, T., Velazco, V. A., Deutscher, N. M., Griffith, D. W. T., Kivi, R., Pollard, D. F., Hase, F., Sussmann, R., Te, Y. V., Strong, K., Roche, S., Sha, M. K., De Maziere, M., Feist, D. G., Iraci, L. T., Roehl, C. M., Retscher, C., and Schepers, D. (2020). Ensemble-based satellite-derived carbon dioxide and methane column-averaged dry-air mole fraction data sets (2003-2018) for carbon and climate applications, Atmospheric Measurement Techniques, 13, 789-819."} {"chunk_id": "satellite_satellite-methane_completeness_q04__dad41dbb5a55", "report_id": "satellite_satellite-methane_completeness_q04", "dataset_id": "satellite-methane", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q04", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Methane satellite observations completeness assessment for greenhouse gas monitoring", "title": "Methane satellite observations completeness assessment for greenhouse gas monitoring", "chunk_index": 7, "token_count": 1056, "text_raw": "(2020). Ensemble-based satellite-derived carbon dioxide and methane column-averaged dry-air mole fraction data sets (2003-2018) for carbon and climate applications, Atmospheric Measurement Techniques, 13, 789-819.\n\n[[7]](https://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PUGS_GHG_main_latest.pdf) Buchwitz, M. (2024). Product User Guide and Specification (PUGS) – Main document for Greenhouse Gas (GHG: CO$_2$ & CH$_4$) data set CDR7 (01.2003-12.2022), C3S project 2021/C3S2_312a_Lot2_DLR/SC1, v7.3.\n\n[[8]](https://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_ATBD_GHG_D_latest.pdf) Reuter, M., and Buchwitz, M. (2024). Algorithm Theoretical Basis Document (ATBD) – ANNEX D for products XCO2_EMMA, XCH4_EMMA, XCO2_OBS4MIPS, XCH4_OBS4MIPS (v4.5, CDR7, 2003-2022), C3S project 2021/C3S2_312a_Lot2_DLR/SC1, v7.1b.\n\n[[9]](https://doi.org/10.24381/14j9-s541) Copernicus Climate Change Service (C3S) and World Meteorological Organization (WMO). (2025). European State of the Climate 2024.\n\n[[10]](https://climate.copernicus.eu/global-climate-highlights-2024) Copernicus Climate Change Service (C3S). (2025). Global Climate Highlights 2024.\n\n[[11]](https://doi.org/10.5194/acp-18-17355-2018) Buchwitz, M., Reuter, M., Schneising, O., Noël, S., Gier, B., Bovensmann, H., Burrows, J. P., Boesch, H., Anand, J., Parker, R. J., Somkuti, P., Detmers, R. G., Hasekamp, O. P., Aben, I., Butz, A., Kuze, A., Suto, H., Yoshida, Y., Crisp, D., and O'Dell, C. (2018). Computation and analysis of atmospheric carbon dioxide annual mean growth rates from satellite observations during 2003-2016, Atmospheric Chemistry and Physics, 18, 17355-17370.\n\n[[12]](https://doi.org/10.5194/acp-23-4863-2023) Feng, L., Palmer, P. I., Parker, R. J., Lunt, M. F., and Bösch, H. (2023). Methane emissions are predominantly responsible for record-breaking atmospheric methane growth rates in 2020 and 2021, Atmospheric Chemistry and Physics, 23, 4863-4880.\n\n[[13]](https://doi.org/10.5194/acp-25-6757-2025) Niwa, Y., Tohjima, Y., Terao, Y., Saeki, T., Ito, A., Umezawa, T., Yamada, K., Sasakawa, M., Machida, T., Nakaoka, S.-I., Nara, H., Tanimoto, H., Mukai, H., Yoshida, Y., Morimoto, S., Takatsuji, S., Tsuboi, K., Sawa, Y., Matsueda, H., Ishijima, K., Fujita, R., Goto, D., Lan, X., Schuldt, K., Heliasz, M., Biermann, T., Chmura, L., Necki, J., Xueref-Remy, I., and Sferlazzo, D. (2025). Multi-observational estimation of regional and sectoral emission contributions to the persistent high growth rate of atmospheric CH$_4$ for 2020–2022, Atmospheric Chemistry and Physics, 25, 6757–6785.\n\n[[14]](https://gaw.kishou.go.jp/publications/global_mean_mole_fractions) World Meteorological Organization (2025). WMO Greenhouse Gas Bulletin, No. 21, 2025 (web site, last access: 16 January 2026).", "text_with_prefix": "EQC Quality Assessment: \"Methane satellite observations completeness assessment for greenhouse gas monitoring\"\nDataset: satellite-methane [CDS]\nAspect: completeness_q04 | Category: Satellite_ECVs\nSection: Methane satellite observations completeness assessment for greenhouse gas monitoring\n---\n(2020). Ensemble-based satellite-derived carbon dioxide and methane column-averaged dry-air mole fraction data sets (2003-2018) for carbon and climate applications, Atmospheric Measurement Techniques, 13, 789-819.\n\n[[7]](https://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PUGS_GHG_main_latest.pdf) Buchwitz, M. (2024). Product User Guide and Specification (PUGS) – Main document for Greenhouse Gas (GHG: CO$_2$ & CH$_4$) data set CDR7 (01.2003-12.2022), C3S project 2021/C3S2_312a_Lot2_DLR/SC1, v7.3.\n\n[[8]](https://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_ATBD_GHG_D_latest.pdf) Reuter, M., and Buchwitz, M. (2024). Algorithm Theoretical Basis Document (ATBD) – ANNEX D for products XCO2_EMMA, XCH4_EMMA, XCO2_OBS4MIPS, XCH4_OBS4MIPS (v4.5, CDR7, 2003-2022), C3S project 2021/C3S2_312a_Lot2_DLR/SC1, v7.1b.\n\n[[9]](https://doi.org/10.24381/14j9-s541) Copernicus Climate Change Service (C3S) and World Meteorological Organization (WMO). (2025). European State of the Climate 2024.\n\n[[10]](https://climate.copernicus.eu/global-climate-highlights-2024) Copernicus Climate Change Service (C3S). (2025). Global Climate Highlights 2024.\n\n[[11]](https://doi.org/10.5194/acp-18-17355-2018) Buchwitz, M., Reuter, M., Schneising, O., Noël, S., Gier, B., Bovensmann, H., Burrows, J. P., Boesch, H., Anand, J., Parker, R. J., Somkuti, P., Detmers, R. G., Hasekamp, O. P., Aben, I., Butz, A., Kuze, A., Suto, H., Yoshida, Y., Crisp, D., and O'Dell, C. (2018). Computation and analysis of atmospheric carbon dioxide annual mean growth rates from satellite observations during 2003-2016, Atmospheric Chemistry and Physics, 18, 17355-17370.\n\n[[12]](https://doi.org/10.5194/acp-23-4863-2023) Feng, L., Palmer, P. I., Parker, R. J., Lunt, M. F., and Bösch, H. (2023). Methane emissions are predominantly responsible for record-breaking atmospheric methane growth rates in 2020 and 2021, Atmospheric Chemistry and Physics, 23, 4863-4880.\n\n[[13]](https://doi.org/10.5194/acp-25-6757-2025) Niwa, Y., Tohjima, Y., Terao, Y., Saeki, T., Ito, A., Umezawa, T., Yamada, K., Sasakawa, M., Machida, T., Nakaoka, S.-I., Nara, H., Tanimoto, H., Mukai, H., Yoshida, Y., Morimoto, S., Takatsuji, S., Tsuboi, K., Sawa, Y., Matsueda, H., Ishijima, K., Fujita, R., Goto, D., Lan, X., Schuldt, K., Heliasz, M., Biermann, T., Chmura, L., Necki, J., Xueref-Remy, I., and Sferlazzo, D. (2025). Multi-observational estimation of regional and sectoral emission contributions to the persistent high growth rate of atmospheric CH$_4$ for 2020–2022, Atmospheric Chemistry and Physics, 25, 6757–6785.\n\n[[14]](https://gaw.kishou.go.jp/publications/global_mean_mole_fractions) World Meteorological Organization (2025). WMO Greenhouse Gas Bulletin, No. 21, 2025 (web site, last access: 16 January 2026)."} {"chunk_id": "satellite_satellite-methane_extremes-detection_q02__7ca022fa0659", "report_id": "satellite_satellite-methane_extremes-detection_q02", "dataset_id": "satellite-methane", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extremes-detection_q02", "aspect_base": "extremes-detection", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Spatial and temporal completeness of methane satellite observations for the analysis of extreme events related to large episodic releases > Quality assessment question", "title": "Spatial and temporal completeness of methane satellite observations for the analysis of extreme events related to large episodic releases", "chunk_index": 0, "token_count": 539, "text_raw": "* **Are methane satellite observations suitable for detecting atmospheric signals associated with extreme emission releases?**\n* **How well, in terms of completeness, do Level 2 mid-tropospheric CH$_4$ columns capture the spatial variability of methane associated with large release events?**\n\nMethane (CH$_4$) is the second most important anthropogenic greenhouse gas after carbon dioxide (CO$_2$), representing about 19% of the total radiative forcing by long-lived greenhouse gases [[1]](https://wmo.int/publication-series/wmo-greenhouse-gas-bulletin-no-19). Atmospheric CH$_4$ also adversely affects human health as a precursor of tropospheric ozone [[2]](https://doi.org/10.1073/pnas.0600201103). Thus, sudden large release of CH$_4$ into the atmosphere could have significant consequences in terms of climate change and health.\n\nOn 26 September 2022, multiple gas leaks were detected from the Nord Stream, an offshore submerged pipeline system that carries natural gas from Russian facilities into Western Europe ([[3]](https://doi.org/10.5194/acp-24-10639-2024), [[4]](https://www.unep.org/news-and-stories/story/pipeline-blasts-released-record-shattering-amount-methane-unep-study)).\n\nIn this application, the Level-2 mid-tropospheric CH$_4$ columns from the IASI instruments (dataset version 10.2) is evaluated. A Level 2 dataset corresponds to data from a specific combination of satellite sensor and retrivial algorithm for each individual satellite footprint along the orbit tracks (see [[5]](https://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PUGS_GHG_main_latest.pdf)).\n\nAlthough the CH$_4$ satellite observations provided by the Climate Data Store are not specifically designed to identify sudden large release events, this application explores the possibility that mid-tropospheric satellite IASI CH$_4$ (v10.2) data can be used to identify the atmospheric signal of emissions as those associated with the Nord Stream event.", "text_with_prefix": "EQC Quality Assessment: \"Spatial and temporal completeness of methane satellite observations for the analysis of extreme events related to large episodic releases\"\nDataset: satellite-methane [CDS]\nAspect: extremes-detection_q02 | Category: Satellite_ECVs\nSection: Spatial and temporal completeness of methane satellite observations for the analysis of extreme events related to large episodic releases > Quality assessment question\n---\n* **Are methane satellite observations suitable for detecting atmospheric signals associated with extreme emission releases?**\n* **How well, in terms of completeness, do Level 2 mid-tropospheric CH$_4$ columns capture the spatial variability of methane associated with large release events?**\n\nMethane (CH$_4$) is the second most important anthropogenic greenhouse gas after carbon dioxide (CO$_2$), representing about 19% of the total radiative forcing by long-lived greenhouse gases [[1]](https://wmo.int/publication-series/wmo-greenhouse-gas-bulletin-no-19). Atmospheric CH$_4$ also adversely affects human health as a precursor of tropospheric ozone [[2]](https://doi.org/10.1073/pnas.0600201103). Thus, sudden large release of CH$_4$ into the atmosphere could have significant consequences in terms of climate change and health.\n\nOn 26 September 2022, multiple gas leaks were detected from the Nord Stream, an offshore submerged pipeline system that carries natural gas from Russian facilities into Western Europe ([[3]](https://doi.org/10.5194/acp-24-10639-2024), [[4]](https://www.unep.org/news-and-stories/story/pipeline-blasts-released-record-shattering-amount-methane-unep-study)).\n\nIn this application, the Level-2 mid-tropospheric CH$_4$ columns from the IASI instruments (dataset version 10.2) is evaluated. A Level 2 dataset corresponds to data from a specific combination of satellite sensor and retrivial algorithm for each individual satellite footprint along the orbit tracks (see [[5]](https://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PUGS_GHG_main_latest.pdf)).\n\nAlthough the CH$_4$ satellite observations provided by the Climate Data Store are not specifically designed to identify sudden large release events, this application explores the possibility that mid-tropospheric satellite IASI CH$_4$ (v10.2) data can be used to identify the atmospheric signal of emissions as those associated with the Nord Stream event."} {"chunk_id": "satellite_satellite-methane_extremes-detection_q02__7128316f89c7", "report_id": "satellite_satellite-methane_extremes-detection_q02", "dataset_id": "satellite-methane", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extremes-detection_q02", "aspect_base": "extremes-detection", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Spatial and temporal completeness of methane satellite observations for the analysis of extreme events related to large episodic releases > Quality assessment statement", "title": "Spatial and temporal completeness of methane satellite observations for the analysis of extreme events related to large episodic releases", "chunk_index": 1, "token_count": 551, "text_raw": "These are the key outcomes of this assessment\n\n* The IASI CH$_4$ (v10.2) mid-tropospheric product was only able to partially detect the atmospheric signal of massive CH$_4$ release associated with the Nord Stream leak on September 2022.\n* In terms of completeness, the ability of IASI CH$_4$ (v10.2) mid-tropospheric product to detect the atmospheric signal of the CH$_4$ plume was mostly limited by the cloud coverage and users need to be aware about the strict cloud screening adopted for the production of this dataset.\n* As the vertical sensitivity of CH$_4$ can affect the quantification of the atmospheric CH$_4$ signal associated with surface releases, users must be aware of the vertical sensitivity of the product. They should consult the the Documentation (e.g., [[7]](https://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_ATBD_GHG_E_latest.pdf)) for information about the altitude sensitivity of the retrieved CH$_4$.\n* The use of this data product to investigate CH$_4$ fluxes occurring at the Earth's surface is recommended in conjunction with appropriate (inversion) modelling.\n* Since this dataset is updated on a yearly basis, users interested in near real time applications should consider other Copernicus resources (e.g., [[11]](https://atmosphere.copernicus.eu/ghg-services/cams-methane-hotspot-explorer?utm_source=press&utm_medium=referral&utm_campaign=CH4-app-2025)).\n```\n\n![Figure1.jpg](attachment:c54ae6d5-d20f-46bf-b1c3-0cfa38b58afb.jpg)\n\n*The figure shows mid-tropospheric IASI column average CH$_4$ (ppb) for 26–28 September 2022 as provided by [[3]](https://doi.org/10.5194/acp-24-10639-2024) under CC-BY license. Turquoise boxes show regions characterised by the presence of the methane plume. AM (PM) indicate the morning (evening) satellite overpasses of each day*", "text_with_prefix": "EQC Quality Assessment: \"Spatial and temporal completeness of methane satellite observations for the analysis of extreme events related to large episodic releases\"\nDataset: satellite-methane [CDS]\nAspect: extremes-detection_q02 | Category: Satellite_ECVs\nSection: Spatial and temporal completeness of methane satellite observations for the analysis of extreme events related to large episodic releases > Quality assessment statement\n---\nThese are the key outcomes of this assessment\n\n* The IASI CH$_4$ (v10.2) mid-tropospheric product was only able to partially detect the atmospheric signal of massive CH$_4$ release associated with the Nord Stream leak on September 2022.\n* In terms of completeness, the ability of IASI CH$_4$ (v10.2) mid-tropospheric product to detect the atmospheric signal of the CH$_4$ plume was mostly limited by the cloud coverage and users need to be aware about the strict cloud screening adopted for the production of this dataset.\n* As the vertical sensitivity of CH$_4$ can affect the quantification of the atmospheric CH$_4$ signal associated with surface releases, users must be aware of the vertical sensitivity of the product. They should consult the the Documentation (e.g., [[7]](https://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_ATBD_GHG_E_latest.pdf)) for information about the altitude sensitivity of the retrieved CH$_4$.\n* The use of this data product to investigate CH$_4$ fluxes occurring at the Earth's surface is recommended in conjunction with appropriate (inversion) modelling.\n* Since this dataset is updated on a yearly basis, users interested in near real time applications should consider other Copernicus resources (e.g., [[11]](https://atmosphere.copernicus.eu/ghg-services/cams-methane-hotspot-explorer?utm_source=press&utm_medium=referral&utm_campaign=CH4-app-2025)).\n```\n\n![Figure1.jpg](attachment:c54ae6d5-d20f-46bf-b1c3-0cfa38b58afb.jpg)\n\n*The figure shows mid-tropospheric IASI column average CH$_4$ (ppb) for 26–28 September 2022 as provided by [[3]](https://doi.org/10.5194/acp-24-10639-2024) under CC-BY license. Turquoise boxes show regions characterised by the presence of the methane plume. AM (PM) indicate the morning (evening) satellite overpasses of each day*"} {"chunk_id": "satellite_satellite-methane_extremes-detection_q02__85a5f210c8d5", "report_id": "satellite_satellite-methane_extremes-detection_q02", "dataset_id": "satellite-methane", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extremes-detection_q02", "aspect_base": "extremes-detection", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Spatial and temporal completeness of methane satellite observations for the analysis of extreme events related to large episodic releases > Methodology", "title": "Spatial and temporal completeness of methane satellite observations for the analysis of extreme events related to large episodic releases", "chunk_index": 2, "token_count": 463, "text_raw": "This notebook has three main objectives:\n* To compare the spatial fields of mid-tropospheric column-averaged methane (CH$_4$) mixing ratios from IASI sensors with independent analyses [[3]](https://doi.org/10.5194/acp-24-10639-2024) over the geographical region and time frame (26-28 September 2022) affected by the Nord Stream emissions.\n* To attribute the observed differences as a function of the cloud cover over the observation scene and the adopted averaging kernels (describing the vertical sensitivity of the retrieved products).\n* To guide the user in the correct use of the evaluated data sets.\n\nThe analysis methodology is divided into the following steps:\n\n**[](section-1)**\n * Define the libraries to be used.\n * Define notebook setup parameters and data requests.\n * Define the functions used in the analysis.\n\n**[](section-2)**\n * Download mid-tropospheric CH$_4$ and cloud cover data for the time period and spatial region of interest.\n\n**[](section-3)**\n * Spatial and temporal analysis of mid-tropospheric CH$_4$.\n * Spatial and temporal analysis of cloud cover.\n * Analysis of the altitude sensitivity of the retrieved mid-tropospheric CH$_4$.\n * Temporal variability of mid-tropospheric CH4 over the region.\n\nNote that prior to the analyses, the original Level 2 mid-tropospheric CH$_4$ data were regridded to a regular 1° x 1° grid and the cloud cover data were retrieved from the Climate Data Store dataset [[6]](https://cds.climate.copernicus.eu/datasets/satellite-cloud-properties?tab=overview) \"Cloud properties global gridded monthly and daily data from 1979 to present derived from satellite observations\" (CLARA-A3).", "text_with_prefix": "EQC Quality Assessment: \"Spatial and temporal completeness of methane satellite observations for the analysis of extreme events related to large episodic releases\"\nDataset: satellite-methane [CDS]\nAspect: extremes-detection_q02 | Category: Satellite_ECVs\nSection: Spatial and temporal completeness of methane satellite observations for the analysis of extreme events related to large episodic releases > Methodology\n---\nThis notebook has three main objectives:\n* To compare the spatial fields of mid-tropospheric column-averaged methane (CH$_4$) mixing ratios from IASI sensors with independent analyses [[3]](https://doi.org/10.5194/acp-24-10639-2024) over the geographical region and time frame (26-28 September 2022) affected by the Nord Stream emissions.\n* To attribute the observed differences as a function of the cloud cover over the observation scene and the adopted averaging kernels (describing the vertical sensitivity of the retrieved products).\n* To guide the user in the correct use of the evaluated data sets.\n\nThe analysis methodology is divided into the following steps:\n\n**[](section-1)**\n * Define the libraries to be used.\n * Define notebook setup parameters and data requests.\n * Define the functions used in the analysis.\n\n**[](section-2)**\n * Download mid-tropospheric CH$_4$ and cloud cover data for the time period and spatial region of interest.\n\n**[](section-3)**\n * Spatial and temporal analysis of mid-tropospheric CH$_4$.\n * Spatial and temporal analysis of cloud cover.\n * Analysis of the altitude sensitivity of the retrieved mid-tropospheric CH$_4$.\n * Temporal variability of mid-tropospheric CH4 over the region.\n\nNote that prior to the analyses, the original Level 2 mid-tropospheric CH$_4$ data were regridded to a regular 1° x 1° grid and the cloud cover data were retrieved from the Climate Data Store dataset [[6]](https://cds.climate.copernicus.eu/datasets/satellite-cloud-properties?tab=overview) \"Cloud properties global gridded monthly and daily data from 1979 to present derived from satellite observations\" (CLARA-A3)."} {"chunk_id": "satellite_satellite-methane_extremes-detection_q02__983140ff7419", "report_id": "satellite_satellite-methane_extremes-detection_q02", "dataset_id": "satellite-methane", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extremes-detection_q02", "aspect_base": "extremes-detection", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Spatial and temporal completeness of methane satellite observations for the analysis of extreme events related to large episodic releases > Analysis and results > 1. Choose the data to use and setup the code > Define notebook setting parameters and data request", "title": "Spatial and temporal completeness of methane satellite observations for the analysis of extreme events related to large episodic releases", "chunk_index": 3, "token_count": 205, "text_raw": "In this step, we define:\n* the list of sensors to be used for getting methane data.\n* the region of interest.\n* the variables to be retrieved by the datasets.\n* the time frame of data to be analysed.\n* the range of legend variability to plot CH4 values in the spatial map.\n* the list of Climate Data Store datasets to be downloaded.\n\nRegion to plot\nVariable to plot\nList of days to be considered (format 'YYYY-MM-DD')\nTo define the range of map legend", "text_with_prefix": "EQC Quality Assessment: \"Spatial and temporal completeness of methane satellite observations for the analysis of extreme events related to large episodic releases\"\nDataset: satellite-methane [CDS]\nAspect: extremes-detection_q02 | Category: Satellite_ECVs\nSection: Spatial and temporal completeness of methane satellite observations for the analysis of extreme events related to large episodic releases > Analysis and results > 1. Choose the data to use and setup the code > Define notebook setting parameters and data request\n---\nIn this step, we define:\n* the list of sensors to be used for getting methane data.\n* the region of interest.\n* the variables to be retrieved by the datasets.\n* the time frame of data to be analysed.\n* the range of legend variability to plot CH4 values in the spatial map.\n* the list of Climate Data Store datasets to be downloaded.\n\nRegion to plot\nVariable to plot\nList of days to be considered (format 'YYYY-MM-DD')\nTo define the range of map legend"} {"chunk_id": "satellite_satellite-methane_extremes-detection_q02__f4beab04bf90", "report_id": "satellite_satellite-methane_extremes-detection_q02", "dataset_id": "satellite-methane", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extremes-detection_q02", "aspect_base": "extremes-detection", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Spatial and temporal completeness of methane satellite observations for the analysis of extreme events related to large episodic releases > Analysis and results > 1. Choose the data to use and setup the code > Define functions used in the analysis", "title": "Spatial and temporal completeness of methane satellite observations for the analysis of extreme events related to large episodic releases", "chunk_index": 4, "token_count": 167, "text_raw": "In this step we define the functions to be used in the analysis:\n* Function to regrid the data to a regular spatial grid.\n* Function to extract mean kernel values and associated pressure levels from the mid-tropospheric IASI CH4 (v10.2) dataset and perform spatial aggregation.\n\n(section-2)=", "text_with_prefix": "EQC Quality Assessment: \"Spatial and temporal completeness of methane satellite observations for the analysis of extreme events related to large episodic releases\"\nDataset: satellite-methane [CDS]\nAspect: extremes-detection_q02 | Category: Satellite_ECVs\nSection: Spatial and temporal completeness of methane satellite observations for the analysis of extreme events related to large episodic releases > Analysis and results > 1. Choose the data to use and setup the code > Define functions used in the analysis\n---\nIn this step we define the functions to be used in the analysis:\n* Function to regrid the data to a regular spatial grid.\n* Function to extract mean kernel values and associated pressure levels from the mid-tropospheric IASI CH4 (v10.2) dataset and perform spatial aggregation.\n\n(section-2)="} {"chunk_id": "satellite_satellite-methane_extremes-detection_q02__817e0b0dac25", "report_id": "satellite_satellite-methane_extremes-detection_q02", "dataset_id": "satellite-methane", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extremes-detection_q02", "aspect_base": "extremes-detection", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Spatial and temporal completeness of methane satellite observations for the analysis of extreme events related to large episodic releases > Analysis and results > 2. Data retrieval > Download and transform data", "title": "Spatial and temporal completeness of methane satellite observations for the analysis of extreme events related to large episodic releases", "chunk_index": 5, "token_count": 278, "text_raw": "In this step, mid-tropospheric IASI CH4 (v10.2) and cloud cover data (CLARA-A3) are downloaded from the Climate Data Store; functions are applied to the datasets.\n\n```text\nsensor = 'iasi_metop_b_nlis'\n```\n\n```text\n100%|██████████| 1/1 [00:00<00:00, 1.63it/s]\n```\n\n```text\nsensor = 'iasi_metop_c_nlis'\n```\n\n```text\n100%|██████████| 1/1 [00:00<00:00, 5.96it/s]\n```\n\n```text\ncloud cover\n```\n\n```text\n100%|██████████| 1/1 [00:00<00:00, 1.60it/s]\n```\n\n(section-3)=", "text_with_prefix": "EQC Quality Assessment: \"Spatial and temporal completeness of methane satellite observations for the analysis of extreme events related to large episodic releases\"\nDataset: satellite-methane [CDS]\nAspect: extremes-detection_q02 | Category: Satellite_ECVs\nSection: Spatial and temporal completeness of methane satellite observations for the analysis of extreme events related to large episodic releases > Analysis and results > 2. Data retrieval > Download and transform data\n---\nIn this step, mid-tropospheric IASI CH4 (v10.2) and cloud cover data (CLARA-A3) are downloaded from the Climate Data Store; functions are applied to the datasets.\n\n```text\nsensor = 'iasi_metop_b_nlis'\n```\n\n```text\n100%|██████████| 1/1 [00:00<00:00, 1.63it/s]\n```\n\n```text\nsensor = 'iasi_metop_c_nlis'\n```\n\n```text\n100%|██████████| 1/1 [00:00<00:00, 5.96it/s]\n```\n\n```text\ncloud cover\n```\n\n```text\n100%|██████████| 1/1 [00:00<00:00, 1.60it/s]\n```\n\n(section-3)="} {"chunk_id": "satellite_satellite-methane_extremes-detection_q02__fd581cec0e2e", "report_id": "satellite_satellite-methane_extremes-detection_q02", "dataset_id": "satellite-methane", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extremes-detection_q02", "aspect_base": "extremes-detection", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Spatial and temporal completeness of methane satellite observations for the analysis of extreme events related to large episodic releases > Analysis and results > 3. Analyses and plotting > Spatial and temporal analysis of mid-tropospheric CH4", "title": "Spatial and temporal completeness of methane satellite observations for the analysis of extreme events related to large episodic releases", "chunk_index": 6, "token_count": 530, "text_raw": "Here we have plotted the mid-tropospheric CH$_4$ data from IASI-B and IASI-C in the spatial domain and temporal framework defined by [[3]](https://doi.org/10.5194/acp-24-10639-2024). \nAs described by [[3]](https://doi.org/10.5194/acp-24-10639-2024), the CH$_4$ released from the Nordstream pipeline moved from the Baltic Sea northwards and eastwards from the point of emission during the first day, while during the following two days it was transported westwards across the Scandinavian peninsula and the North Sea.\n\nThe area of high CH$_4$ values identified by [[3]](https://doi.org/10.5194/acp-24-10639-2024) as affected by the Nord Stream release over [0°-5° E, 60° - 65°N] is well visible in the IASI-C data on 28 September 2022 and partially detected by IASI-B over the western Scandinavian coasts. However, despite [[3]](https://doi.org/10.5194/acp-24-10639-2024), the Climate Data Store mid-tropospheric CH$_4$ data from the IASI-B and IASI-C were not able to detect the Nord Stream plume over the Baltic Sea on 26 September 2022 and west of 0°E on 28 September 2022. This is related to the fact that the Climate Data Store mid-tropospheric CH$_4$ columns have a lower spatial occurrence of valid data, due to the different algorithm and quality control screening applied in the dataset production.\n\nTo plot CH4 data for sensors and selected days\nSelect days of interest\n\n*The figure shows mid-tropospheric CH$_4$ (expressed in ppb) for 26-28 September 2022 as provided by the IASI-B (upper plots) and IASI-C (lower plots) datasets (v10.2).*", "text_with_prefix": "EQC Quality Assessment: \"Spatial and temporal completeness of methane satellite observations for the analysis of extreme events related to large episodic releases\"\nDataset: satellite-methane [CDS]\nAspect: extremes-detection_q02 | Category: Satellite_ECVs\nSection: Spatial and temporal completeness of methane satellite observations for the analysis of extreme events related to large episodic releases > Analysis and results > 3. Analyses and plotting > Spatial and temporal analysis of mid-tropospheric CH4\n---\nHere we have plotted the mid-tropospheric CH$_4$ data from IASI-B and IASI-C in the spatial domain and temporal framework defined by [[3]](https://doi.org/10.5194/acp-24-10639-2024). \nAs described by [[3]](https://doi.org/10.5194/acp-24-10639-2024), the CH$_4$ released from the Nordstream pipeline moved from the Baltic Sea northwards and eastwards from the point of emission during the first day, while during the following two days it was transported westwards across the Scandinavian peninsula and the North Sea.\n\nThe area of high CH$_4$ values identified by [[3]](https://doi.org/10.5194/acp-24-10639-2024) as affected by the Nord Stream release over [0°-5° E, 60° - 65°N] is well visible in the IASI-C data on 28 September 2022 and partially detected by IASI-B over the western Scandinavian coasts. However, despite [[3]](https://doi.org/10.5194/acp-24-10639-2024), the Climate Data Store mid-tropospheric CH$_4$ data from the IASI-B and IASI-C were not able to detect the Nord Stream plume over the Baltic Sea on 26 September 2022 and west of 0°E on 28 September 2022. This is related to the fact that the Climate Data Store mid-tropospheric CH$_4$ columns have a lower spatial occurrence of valid data, due to the different algorithm and quality control screening applied in the dataset production.\n\nTo plot CH4 data for sensors and selected days\nSelect days of interest\n\n*The figure shows mid-tropospheric CH$_4$ (expressed in ppb) for 26-28 September 2022 as provided by the IASI-B (upper plots) and IASI-C (lower plots) datasets (v10.2).*"} {"chunk_id": "satellite_satellite-methane_extremes-detection_q02__903eab3a2ae8", "report_id": "satellite_satellite-methane_extremes-detection_q02", "dataset_id": "satellite-methane", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extremes-detection_q02", "aspect_base": "extremes-detection", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Spatial and temporal completeness of methane satellite observations for the analysis of extreme events related to large episodic releases > Analysis and results > 3. Analyses and plotting > Spatial and temporal analysis of cloud cover", "title": "Spatial and temporal completeness of methane satellite observations for the analysis of extreme events related to large episodic releases", "chunk_index": 7, "token_count": 507, "text_raw": "To attribute differences in the data availability between mid-tropospheric CH$_4$ data from Climate Data Store and that reported by [[3]](https://doi.org/10.5194/acp-24-10639-2024), we inspected the daily values of cloud cover fraction provided by the Interim Climate Data Record [[6]](https://cds.climate.copernicus.eu/datasets/satellite-cloud-properties?tab=overview) \"Cloud properties global gridded monthly and daily data from 1979 to present derived from satellite observations\" (product family: CLARA-A3).\n\nIt is clear that the the presence of high values of cloud cover fraction over the most part of the observation scene prevented the detection of the Nord Stream plume on 26 - 27 September 202 by the Climate Data Store IASI mid-tropospheric CH$_4$ dataset. According to the Documentation [[7]](https://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_ATBD_GHG_E_latest.pdf), the Climate Data Store mid-tropospheric satellite IASI CH4 (v10.2) dataset is retrieved when no cloud or aerosol is detected. On the other hand, for the dataset used by [[3]](https://doi.org/10.5194/acp-24-10639-2024), only scenes strongly affected by clouds are not processed [[8]](https://doi.org/10.5281/zenodo.5873645).\n\nFilter data for selected days\nSetting plot view\nPlotting cloud cover data\n\n*The figure shows the total cloud coverage fraction (espressed as %) for 26-28 September 2022 as provided by the Climate Data Store Interim Climate Data Record “Cloud properties global gridded monthly and daily data from 1979 to present derived from satellite observations” (product family: CLARA-A3)*", "text_with_prefix": "EQC Quality Assessment: \"Spatial and temporal completeness of methane satellite observations for the analysis of extreme events related to large episodic releases\"\nDataset: satellite-methane [CDS]\nAspect: extremes-detection_q02 | Category: Satellite_ECVs\nSection: Spatial and temporal completeness of methane satellite observations for the analysis of extreme events related to large episodic releases > Analysis and results > 3. Analyses and plotting > Spatial and temporal analysis of cloud cover\n---\nTo attribute differences in the data availability between mid-tropospheric CH$_4$ data from Climate Data Store and that reported by [[3]](https://doi.org/10.5194/acp-24-10639-2024), we inspected the daily values of cloud cover fraction provided by the Interim Climate Data Record [[6]](https://cds.climate.copernicus.eu/datasets/satellite-cloud-properties?tab=overview) \"Cloud properties global gridded monthly and daily data from 1979 to present derived from satellite observations\" (product family: CLARA-A3).\n\nIt is clear that the the presence of high values of cloud cover fraction over the most part of the observation scene prevented the detection of the Nord Stream plume on 26 - 27 September 202 by the Climate Data Store IASI mid-tropospheric CH$_4$ dataset. According to the Documentation [[7]](https://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_ATBD_GHG_E_latest.pdf), the Climate Data Store mid-tropospheric satellite IASI CH4 (v10.2) dataset is retrieved when no cloud or aerosol is detected. On the other hand, for the dataset used by [[3]](https://doi.org/10.5194/acp-24-10639-2024), only scenes strongly affected by clouds are not processed [[8]](https://doi.org/10.5281/zenodo.5873645).\n\nFilter data for selected days\nSetting plot view\nPlotting cloud cover data\n\n*The figure shows the total cloud coverage fraction (espressed as %) for 26-28 September 2022 as provided by the Climate Data Store Interim Climate Data Record “Cloud properties global gridded monthly and daily data from 1979 to present derived from satellite observations” (product family: CLARA-A3)*"} {"chunk_id": "satellite_satellite-methane_extremes-detection_q02__4c4b8f2bad1a", "report_id": "satellite_satellite-methane_extremes-detection_q02", "dataset_id": "satellite-methane", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extremes-detection_q02", "aspect_base": "extremes-detection", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Spatial and temporal completeness of methane satellite observations for the analysis of extreme events related to large episodic releases > Analysis and results > 3. Analyses and plotting > Analysis of the altitude sensitivity of the retrieved mid-tropospheric CH4", "title": "Spatial and temporal completeness of methane satellite observations for the analysis of extreme events related to large episodic releases", "chunk_index": 8, "token_count": 527, "text_raw": "The vertical sensitivity of the IASI CH$_4$ mid-tropospheric product may affect the identification and the quantification of the plume from the Nord Stream pipelines. The averaging kernel (AK), which is provided as variable in the Climate Data Store data, accounts for the vertical sensitivity of the derived CH$_4$ values from the IASI retrievals, as described in [[7]](https://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_ATBD_GHG_E_latest.pdf).\n\nFor the Climate Data Store mid-tropospheric CH$_4$ product, the normalised column AK values averaged over the spatial domain showed a maximum sensitivity in the mid-upper troposphere (from 100 to 600 hPa) during the time period considered. According to [[3]] (https://doi.org/10.5194/acp-24-10639-2024), the southern part of the Nord Stream plume [59-63° N; 0-7° E], extending from the west coast of Scandinavia, was diffuse above 500 hPa, favouring the plume detection by the Climate Data Store mid-tropospheric CH$_4$ IASI data products.\n\nUsers interested in further information on the use of the AKs (as an example for comparison with model simulations) are encouraged to consult the dataset documentation [[9]](https://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PUGS_GHG_E_latest.pdf) and [[5]](https://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PUGS_GHG_main_latest.pdf).\n\nSetting plot view\n\n*IASI-B (top plot) and IASI-C (bottom plot) for CH$_4$ normalised column averaging kernel computed over the analysed region on 28 September 2022.*", "text_with_prefix": "EQC Quality Assessment: \"Spatial and temporal completeness of methane satellite observations for the analysis of extreme events related to large episodic releases\"\nDataset: satellite-methane [CDS]\nAspect: extremes-detection_q02 | Category: Satellite_ECVs\nSection: Spatial and temporal completeness of methane satellite observations for the analysis of extreme events related to large episodic releases > Analysis and results > 3. Analyses and plotting > Analysis of the altitude sensitivity of the retrieved mid-tropospheric CH4\n---\nThe vertical sensitivity of the IASI CH$_4$ mid-tropospheric product may affect the identification and the quantification of the plume from the Nord Stream pipelines. The averaging kernel (AK), which is provided as variable in the Climate Data Store data, accounts for the vertical sensitivity of the derived CH$_4$ values from the IASI retrievals, as described in [[7]](https://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_ATBD_GHG_E_latest.pdf).\n\nFor the Climate Data Store mid-tropospheric CH$_4$ product, the normalised column AK values averaged over the spatial domain showed a maximum sensitivity in the mid-upper troposphere (from 100 to 600 hPa) during the time period considered. According to [[3]] (https://doi.org/10.5194/acp-24-10639-2024), the southern part of the Nord Stream plume [59-63° N; 0-7° E], extending from the west coast of Scandinavia, was diffuse above 500 hPa, favouring the plume detection by the Climate Data Store mid-tropospheric CH$_4$ IASI data products.\n\nUsers interested in further information on the use of the AKs (as an example for comparison with model simulations) are encouraged to consult the dataset documentation [[9]](https://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PUGS_GHG_E_latest.pdf) and [[5]](https://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PUGS_GHG_main_latest.pdf).\n\nSetting plot view\n\n*IASI-B (top plot) and IASI-C (bottom plot) for CH$_4$ normalised column averaging kernel computed over the analysed region on 28 September 2022.*"} {"chunk_id": "satellite_satellite-methane_extremes-detection_q02__e020aa92b189", "report_id": "satellite_satellite-methane_extremes-detection_q02", "dataset_id": "satellite-methane", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extremes-detection_q02", "aspect_base": "extremes-detection", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Spatial and temporal completeness of methane satellite observations for the analysis of extreme events related to large episodic releases > Analysis and results > 3. Analyses and plotting > Temporal variability of mid-tropospheric CH$_4$ over the region", "title": "Spatial and temporal completeness of methane satellite observations for the analysis of extreme events related to large episodic releases", "chunk_index": 9, "token_count": 509, "text_raw": "As clearly stated in the dataset Documentation [[5]](https://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PUGS_GHG_main_latest.pdf), the dataset \"Methane data from 2002 to present derived from satellite observations\" should be used them in combination with appropriate modelling to obtain information on surface CH$_4$ fluxes.\n\nTo illustrate that the use of this dataset without appropriate modelling can lead to misleading conclusions, we analysed the temporal variability of the mid-tropospheric CH$_4$ column over the study region. In particular, we calculated the distribution of CH$_4$ values for each single day during September 2022 for both IASI-B and IASI-C. The box and wiskers plot shows the percentiles of the data distribution, the median (green line) as well as the outliers. \nIt can be seen that despite the occurrence of the Nord Stream leak and the effective detection of the plume signal by IASI-C on 29 September 2022, the CH$_4$ values did not deviate from the monthly variability. This shows that even if the atmospheric data provided by the mid-tropospheric IASI CH$_4$ (v10.2) dataset contain the signal of the CH$_4$ release from the Nord Stream pipelines, it can hardly be detected without the use of appropriate modelling, as adopted by [[3]](https://doi.org/10.5194/acp-24-10639-2024) or [[10]](https://climate.copernicus.eu/esotc/2023/greenhouse-gas-fluxes).\n\n*Daily percentiles and outliers for mid-tropospheric CH4 (expressed in ppb) during September 2022 as provided by the IASI-B (upper plots) and IASI-C (lower plots) datasets (v10.2).*", "text_with_prefix": "EQC Quality Assessment: \"Spatial and temporal completeness of methane satellite observations for the analysis of extreme events related to large episodic releases\"\nDataset: satellite-methane [CDS]\nAspect: extremes-detection_q02 | Category: Satellite_ECVs\nSection: Spatial and temporal completeness of methane satellite observations for the analysis of extreme events related to large episodic releases > Analysis and results > 3. Analyses and plotting > Temporal variability of mid-tropospheric CH$_4$ over the region\n---\nAs clearly stated in the dataset Documentation [[5]](https://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PUGS_GHG_main_latest.pdf), the dataset \"Methane data from 2002 to present derived from satellite observations\" should be used them in combination with appropriate modelling to obtain information on surface CH$_4$ fluxes.\n\nTo illustrate that the use of this dataset without appropriate modelling can lead to misleading conclusions, we analysed the temporal variability of the mid-tropospheric CH$_4$ column over the study region. In particular, we calculated the distribution of CH$_4$ values for each single day during September 2022 for both IASI-B and IASI-C. The box and wiskers plot shows the percentiles of the data distribution, the median (green line) as well as the outliers. \nIt can be seen that despite the occurrence of the Nord Stream leak and the effective detection of the plume signal by IASI-C on 29 September 2022, the CH$_4$ values did not deviate from the monthly variability. This shows that even if the atmospheric data provided by the mid-tropospheric IASI CH$_4$ (v10.2) dataset contain the signal of the CH$_4$ release from the Nord Stream pipelines, it can hardly be detected without the use of appropriate modelling, as adopted by [[3]](https://doi.org/10.5194/acp-24-10639-2024) or [[10]](https://climate.copernicus.eu/esotc/2023/greenhouse-gas-fluxes).\n\n*Daily percentiles and outliers for mid-tropospheric CH4 (expressed in ppb) during September 2022 as provided by the IASI-B (upper plots) and IASI-C (lower plots) datasets (v10.2).*"} {"chunk_id": "satellite_satellite-methane_extremes-detection_q02__b6218c912377", "report_id": "satellite_satellite-methane_extremes-detection_q02", "dataset_id": "satellite-methane", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extremes-detection_q02", "aspect_base": "extremes-detection", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Spatial and temporal completeness of methane satellite observations for the analysis of extreme events related to large episodic releases > ℹ️ If you want to know more > Key resources", "title": "Spatial and temporal completeness of methane satellite observations for the analysis of extreme events related to large episodic releases", "chunk_index": 10, "token_count": 404, "text_raw": "The CDS catalogue entries for the data used were:\n* Methane data from 2002 to present derived from satellite observations: https://cds.climate.copernicus.eu/datasets/satellite-methane?tab=overview\n* Cloud properties global gridded monthly and daily data from 1979 to present derived from satellite observations: https://cds.climate.copernicus.eu/datasets/satellite-cloud-properties?tab=overview\n\nMore information on the Nord Stream leak event can be found in [[4]](https://www.unep.org/news-and-stories/story/pipeline-blasts-released-record-shattering-amount-methane-unep-study).\n\nUsers interested in near-ral time detection of hot-spot locations for methane emissions can consider to use the CAMS Methane Hotspot Explorer [[11]](https://atmosphere.copernicus.eu/ghg-services/cams-methane-hotspot-explorer?utm_source=press&utm_medium=referral&utm_campaign=CH4-app-2025).\n\nCode libraries used:\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)\n\nhttps://www.unep.org/news-and-stories/story/pipeline-blasts-released-record-shattering-amount-methane-unep-study", "text_with_prefix": "EQC Quality Assessment: \"Spatial and temporal completeness of methane satellite observations for the analysis of extreme events related to large episodic releases\"\nDataset: satellite-methane [CDS]\nAspect: extremes-detection_q02 | Category: Satellite_ECVs\nSection: Spatial and temporal completeness of methane satellite observations for the analysis of extreme events related to large episodic releases > ℹ️ If you want to know more > Key resources\n---\nThe CDS catalogue entries for the data used were:\n* Methane data from 2002 to present derived from satellite observations: https://cds.climate.copernicus.eu/datasets/satellite-methane?tab=overview\n* Cloud properties global gridded monthly and daily data from 1979 to present derived from satellite observations: https://cds.climate.copernicus.eu/datasets/satellite-cloud-properties?tab=overview\n\nMore information on the Nord Stream leak event can be found in [[4]](https://www.unep.org/news-and-stories/story/pipeline-blasts-released-record-shattering-amount-methane-unep-study).\n\nUsers interested in near-ral time detection of hot-spot locations for methane emissions can consider to use the CAMS Methane Hotspot Explorer [[11]](https://atmosphere.copernicus.eu/ghg-services/cams-methane-hotspot-explorer?utm_source=press&utm_medium=referral&utm_campaign=CH4-app-2025).\n\nCode libraries used:\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)\n\nhttps://www.unep.org/news-and-stories/story/pipeline-blasts-released-record-shattering-amount-methane-unep-study"} {"chunk_id": "satellite_satellite-methane_extremes-detection_q02__248d3d9fa7fb", "report_id": "satellite_satellite-methane_extremes-detection_q02", "dataset_id": "satellite-methane", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extremes-detection_q02", "aspect_base": "extremes-detection", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Spatial and temporal completeness of methane satellite observations for the analysis of extreme events related to large episodic releases > ℹ️ If you want to know more > References", "title": "Spatial and temporal completeness of methane satellite observations for the analysis of extreme events related to large episodic releases", "chunk_index": 11, "token_count": 1043, "text_raw": "[[1]](https://wmo.int/publication-series/wmo-greenhouse-gas-bulletin-no-19) World Meteorological Organization (2023). WMO Greenhouse Gas Bulletin, No. 19, ISSN 2078-0796.\n\n[[2]](https://doi.org/10.1073/pnas.0600201103) West, J. J., Fiore, A. M., Horowitz, L. W., and Mauzerall, D. L. (2006). Global health benefits of mitigating ozone pollution with methane emission controls. P. Natl. Acad. Sci. USA, 103, 3988–3993.\n\n[[3]](https://doi.org/10.5194/acp-24-10639-2024) Wilson, C., Kerridge, B. J., Siddans, R., Moore, D. P., Ventress, L. J., Dowd, E., Feng, W., Chipperfield, M. P., and Remedios, J. J. (2024). Quantifying large methane emissions from the Nord Stream pipeline gas leak of September 2022 using IASI satellite observations and inverse modelling. Atmos. Chem. Phys., 24, 10639–10653.\n\n[[4]](https://www.unep.org/news-and-stories/story/pipeline-blasts-released-record-shattering-amount-methane-unep-study) United Nations Environmental Programme (UNEP), (2025). Pipeline blasts released record-shattering amount of methane: UNEP study. https://www.unep.org/news-and-stories/story/pipeline-blasts-released-record-shattering-amount-methane-unep-study\n\n[[5]](https://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PUGS_GHG_main_latest.pdf) Buchwitz, M. (2024). Product User Guide and Specification (PUGS) – Main document for Greenhouse Gas (GHG: CO2 & CH4) data set CDR7 (01.2003-12.2022).\n\n[[6]](https://cds.climate.copernicus.eu/datasets/satellite-cloud-properties?tab=overview) Copernicus Climate Change Service (C3S), Climate Data Store (CDS). (2022). Cloud properties global gridded monthly and daily data from 1982 to present derived from satellite observations. Copernicus Climate Change Service (C3S) Climate Data Store (CDS). DOI: 10.24381/cds.68653055 (Accessed on 19-Mar-2025).\n\n[[7]](https://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_ATBD_GHG_E_latest.pdf) Crevoisier, C. (2024). Algorithm Theoretical Basis Document (ATBD) – ANNEX E for IASI CO2 (v10.1) and CH4 (v10.2) and AIRS CO2 mid-tropospheric products.\n\n[[8]](https://doi.org/10.5281/zenodo.5873645). Ventress, L., Siddans, R., Knappett, D. (2021). RAL IASI Methane Retrieval (ATDB).\n\n[[9]](https://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PUGS_GHG_E_latest.pdf) Crevoisier, C. (2023). Product User Guide and Specification (PUGS) – ANNEX E for IASI CO2 (10.1) and CH4 (v10.2) and AIRS CO2 mid-tropospheric products.\n\n[[10]](https://climate.copernicus.eu/esotc/2023/greenhouse-gas-fluxes) Copernicus Climate Change Service (C3S). (2024). European State of the Climate 2023, Web site.\n\n[[11]](https://atmosphere.copernicus.eu/ghg-services/cams-methane-hotspot-explorer?utm_source=press&utm_medium=referral&utm_campaign=CH4-app-2025) Copernicus Atmosphere Monitoring Service (CAMS). (2025). Copernicus: Tool to routinely identify and track methane emissions and leaks goes operational, Web site.", "text_with_prefix": "EQC Quality Assessment: \"Spatial and temporal completeness of methane satellite observations for the analysis of extreme events related to large episodic releases\"\nDataset: satellite-methane [CDS]\nAspect: extremes-detection_q02 | Category: Satellite_ECVs\nSection: Spatial and temporal completeness of methane satellite observations for the analysis of extreme events related to large episodic releases > ℹ️ If you want to know more > References\n---\n[[1]](https://wmo.int/publication-series/wmo-greenhouse-gas-bulletin-no-19) World Meteorological Organization (2023). WMO Greenhouse Gas Bulletin, No. 19, ISSN 2078-0796.\n\n[[2]](https://doi.org/10.1073/pnas.0600201103) West, J. J., Fiore, A. M., Horowitz, L. W., and Mauzerall, D. L. (2006). Global health benefits of mitigating ozone pollution with methane emission controls. P. Natl. Acad. Sci. USA, 103, 3988–3993.\n\n[[3]](https://doi.org/10.5194/acp-24-10639-2024) Wilson, C., Kerridge, B. J., Siddans, R., Moore, D. P., Ventress, L. J., Dowd, E., Feng, W., Chipperfield, M. P., and Remedios, J. J. (2024). Quantifying large methane emissions from the Nord Stream pipeline gas leak of September 2022 using IASI satellite observations and inverse modelling. Atmos. Chem. Phys., 24, 10639–10653.\n\n[[4]](https://www.unep.org/news-and-stories/story/pipeline-blasts-released-record-shattering-amount-methane-unep-study) United Nations Environmental Programme (UNEP), (2025). Pipeline blasts released record-shattering amount of methane: UNEP study. https://www.unep.org/news-and-stories/story/pipeline-blasts-released-record-shattering-amount-methane-unep-study\n\n[[5]](https://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PUGS_GHG_main_latest.pdf) Buchwitz, M. (2024). Product User Guide and Specification (PUGS) – Main document for Greenhouse Gas (GHG: CO2 & CH4) data set CDR7 (01.2003-12.2022).\n\n[[6]](https://cds.climate.copernicus.eu/datasets/satellite-cloud-properties?tab=overview) Copernicus Climate Change Service (C3S), Climate Data Store (CDS). (2022). Cloud properties global gridded monthly and daily data from 1982 to present derived from satellite observations. Copernicus Climate Change Service (C3S) Climate Data Store (CDS). DOI: 10.24381/cds.68653055 (Accessed on 19-Mar-2025).\n\n[[7]](https://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_ATBD_GHG_E_latest.pdf) Crevoisier, C. (2024). Algorithm Theoretical Basis Document (ATBD) – ANNEX E for IASI CO2 (v10.1) and CH4 (v10.2) and AIRS CO2 mid-tropospheric products.\n\n[[8]](https://doi.org/10.5281/zenodo.5873645). Ventress, L., Siddans, R., Knappett, D. (2021). RAL IASI Methane Retrieval (ATDB).\n\n[[9]](https://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PUGS_GHG_E_latest.pdf) Crevoisier, C. (2023). Product User Guide and Specification (PUGS) – ANNEX E for IASI CO2 (10.1) and CH4 (v10.2) and AIRS CO2 mid-tropospheric products.\n\n[[10]](https://climate.copernicus.eu/esotc/2023/greenhouse-gas-fluxes) Copernicus Climate Change Service (C3S). (2024). European State of the Climate 2023, Web site.\n\n[[11]](https://atmosphere.copernicus.eu/ghg-services/cams-methane-hotspot-explorer?utm_source=press&utm_medium=referral&utm_campaign=CH4-app-2025) Copernicus Atmosphere Monitoring Service (CAMS). (2025). Copernicus: Tool to routinely identify and track methane emissions and leaks goes operational, Web site."} {"chunk_id": "satellite_satellite-methane_uncertainty_q01__b4ef2359879c", "report_id": "satellite_satellite-methane_uncertainty_q01", "dataset_id": "satellite-methane", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty_q01", "aspect_base": "uncertainty", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Methane satellite observations uncertainty and completeness assessment for carbon cycle", "title": "Methane satellite observations uncertainty and completeness assessment for carbon cycle", "chunk_index": 0, "token_count": 95, "text_raw": "Production date: 02-09-2024\n\nProduced by: Consiglio Nazionale delle Ricerche ([CNR](https://www.cnr.it/en))", "text_with_prefix": "EQC Quality Assessment: \"Methane satellite observations uncertainty and completeness assessment for carbon cycle\"\nDataset: satellite-methane [CDS]\nAspect: uncertainty_q01 | Category: Satellite_ECVs\nSection: Methane satellite observations uncertainty and completeness assessment for carbon cycle\n---\nProduction date: 02-09-2024\n\nProduced by: Consiglio Nazionale delle Ricerche ([CNR](https://www.cnr.it/en))"} {"chunk_id": "satellite_satellite-methane_uncertainty_q01__f9eb243ae01c", "report_id": "satellite_satellite-methane_uncertainty_q01", "dataset_id": "satellite-methane", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty_q01", "aspect_base": "uncertainty", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Methane satellite observations uncertainty and completeness assessment for carbon cycle > Quality assessment questions", "title": "Methane satellite observations uncertainty and completeness assessment for carbon cycle", "chunk_index": 1, "token_count": 743, "text_raw": "* **Is the uncertainty of satellite-based observations sufficiently low to assess year-to-year anomalies in methane emissions of wetlands?**\n* **Is the spatial and temporal resolution/coverage of satellite-based observations sufficient to assess year-to-year anomalies in methane emissions of wetlands?**\n\nMethane (CH4) is the second most important anthropogenic greenhouse gas, representing about 19% of the total radiative forcing by long living greenhouse gases [[1]](https://library.wmo.int/viewer/68532/?offset=#page=4&viewer=picture&o=bookmarks&n=0&q=). Natural wetlands account for up to 30% of global methane (CH4) emissions, but a large uncertainty (up to 65%) still affects tropical wetland CH4 emission estimates [[2]](https://doi.org/10.5194/essd-12-1561-2020). The positive response of wetland CH4 emissions to climate change is an important feedback that can amplify atmospheric CH4 values. This response can be related to the effect of rising temperatures on microbial activities (e.g., methanogenesis) and to the expansion of wetlands with increased total precipitations. Intensified wetland CH4 emissions have been reported during 2000–2021 ([[3]](https://doi.org/10.1038/s41558-023-01629-0)), highlighting the need for sustained monitoring and observations of global wetland CH4 fluxes. Moreover, future projections of wetland CH4 emissions suggested sustained emission increase under different climate change scenarios ([[3]](https://doi.org/10.1038/s41558-023-01629-0) and [[4]](https://doi.org/10.1126/sciadv.aay4444)). Satellite observations can represent a powerful tool for contributing to enhance the knowledge about CH4 sources and sinks. While these data are commonly used in inverse modelling system (e.g., [[5]](https://ads.atmosphere.copernicus.eu/datasets/cams-global-greenhouse-gas-inversion?tab=overview)), this assessment explores the possibility to directly use satellite CH4 data for investigate year-to-year anomalies in CH4 emissions of tropical wetlands.\n\nattachment:2b185ba0-efe5-4f85-9b1f-19651082b442.png\n---\nheight: 500px\n---\nThe figure shows temporal trends and variations in wetland CH4 emissions during 2000–2021 relative to the baseline of 2000–2006 level in comparison to future projections. Black lines report variations over 2000-2021 as derived from simulations based on two climate forcing datasets: a ground-based dataset from the CRU at the University of East Anglia (continuous line) and a reanalysis-based MERRA2 (dashed line). Coloured lines represent the ensemble mean of each future projection RCP scenario, while shaded areas represent the 1σ range of the future estimates. **The estimated increase in emissions over 2000-2021 appears to be higher than under any of the RCP scenarios.** Image reproduced from [[3]](https://doi.org/10.1038/s41558-023-01629-0) under CC-BY license.\n```", "text_with_prefix": "EQC Quality Assessment: \"Methane satellite observations uncertainty and completeness assessment for carbon cycle\"\nDataset: satellite-methane [CDS]\nAspect: uncertainty_q01 | Category: Satellite_ECVs\nSection: Methane satellite observations uncertainty and completeness assessment for carbon cycle > Quality assessment questions\n---\n* **Is the uncertainty of satellite-based observations sufficiently low to assess year-to-year anomalies in methane emissions of wetlands?**\n* **Is the spatial and temporal resolution/coverage of satellite-based observations sufficient to assess year-to-year anomalies in methane emissions of wetlands?**\n\nMethane (CH4) is the second most important anthropogenic greenhouse gas, representing about 19% of the total radiative forcing by long living greenhouse gases [[1]](https://library.wmo.int/viewer/68532/?offset=#page=4&viewer=picture&o=bookmarks&n=0&q=). Natural wetlands account for up to 30% of global methane (CH4) emissions, but a large uncertainty (up to 65%) still affects tropical wetland CH4 emission estimates [[2]](https://doi.org/10.5194/essd-12-1561-2020). The positive response of wetland CH4 emissions to climate change is an important feedback that can amplify atmospheric CH4 values. This response can be related to the effect of rising temperatures on microbial activities (e.g., methanogenesis) and to the expansion of wetlands with increased total precipitations. Intensified wetland CH4 emissions have been reported during 2000–2021 ([[3]](https://doi.org/10.1038/s41558-023-01629-0)), highlighting the need for sustained monitoring and observations of global wetland CH4 fluxes. Moreover, future projections of wetland CH4 emissions suggested sustained emission increase under different climate change scenarios ([[3]](https://doi.org/10.1038/s41558-023-01629-0) and [[4]](https://doi.org/10.1126/sciadv.aay4444)). Satellite observations can represent a powerful tool for contributing to enhance the knowledge about CH4 sources and sinks. While these data are commonly used in inverse modelling system (e.g., [[5]](https://ads.atmosphere.copernicus.eu/datasets/cams-global-greenhouse-gas-inversion?tab=overview)), this assessment explores the possibility to directly use satellite CH4 data for investigate year-to-year anomalies in CH4 emissions of tropical wetlands.\n\nattachment:2b185ba0-efe5-4f85-9b1f-19651082b442.png\n---\nheight: 500px\n---\nThe figure shows temporal trends and variations in wetland CH4 emissions during 2000–2021 relative to the baseline of 2000–2006 level in comparison to future projections. Black lines report variations over 2000-2021 as derived from simulations based on two climate forcing datasets: a ground-based dataset from the CRU at the University of East Anglia (continuous line) and a reanalysis-based MERRA2 (dashed line). Coloured lines represent the ensemble mean of each future projection RCP scenario, while shaded areas represent the 1σ range of the future estimates. **The estimated increase in emissions over 2000-2021 appears to be higher than under any of the RCP scenarios.** Image reproduced from [[3]](https://doi.org/10.1038/s41558-023-01629-0) under CC-BY license.\n```"} {"chunk_id": "satellite_satellite-methane_uncertainty_q01__29344afdc1d9", "report_id": "satellite_satellite-methane_uncertainty_q01", "dataset_id": "satellite-methane", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty_q01", "aspect_base": "uncertainty", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Methane satellite observations uncertainty and completeness assessment for carbon cycle > Quality assessment statement", "title": "Methane satellite observations uncertainty and completeness assessment for carbon cycle", "chunk_index": 2, "token_count": 1063, "text_raw": "These are the key outcomes of this assessment\n\n* In terms of uncertainty, Level 2 XCH4 data products appear appropriate for investigating inter-annual XCH4 variability associated with anomalies in tropical wetland emissions\n* In terms of spatial/temporal resolution, Level 2 XCH4 data products appear appropriate for investigating inter-annual XCH4 variability associated with anomalies in tropical wetland emissions\n* In terms of spatial/temporal coverage, users would use data products derived by algorithms (such as CH4_GOS_OCPR, CH4_GO2_SRPR or XCH4_EMMA) that optimise data availability over regions frequently affected by cloud cover or high aerosol loading. \n* Users should consult the documentation as well as the uncertainty and quality flags in the data files as appropriate for their applications. It should also be emphasised that the use of the XCH4 data provided by satellite is not trivial, and interpretation of these products typically requires appropriate modelling [[6]](http://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PUGS_GHG_main_latest.pdf).\n* Users will need to consult product user guides ([[6]](http://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PUGS_GHG_main_latest.pdf)) for guidance on how to use the averaging kernels when using satellite CH4 data, especially for comparison with model outputs\n\n## 📋 Methodology\n\nTo assess if the uncertainty, spatial/temporal resolution and coverage of satellite-based observations are sufficient to investigate year-to-year anomalies in CH4 emissions of tropical wetlands, we inspected the documentation of the CDS dataset “Methane data from 2002 to present derived from satellite observations” ([[7]](https://cds.climate.copernicus.eu/datasets/satellite-methane?tab=overview), [[6]](http://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PUGS_GHG_main_latest.pdf), [[8]](http://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PQAR_GHG_main_latest.pdf)) as well as a case study available from scientific literature ([[9]](https://doi.org/10.1016/j.rse.2018.02.011)). Please note that to analyse annual anomalies in wetland methane emissions, [[9]](https://doi.org/10.1016/j.rse.2018.02.011) used an earlier version (v7) of Level 2 data from the CH4_GOS_OCPR data product.\n\nattachment:88fcf479-6fcf-4d60-834f-18e8ca2c2bb4.png\n---\nheight: 500px\n---\nMap showing the location of the individual wetland regions identified using the Sustainable Wetlands Adaptation and Mitigation Program (SWAMP) data from the Center for International Forestry Research (CIFOR) and timeseries of detrended XCH4 seasonal cycle over the Pantanal region, including the CH4_GOS_OCFP dataset (“GOSAT”) and model simulations utilising the different wetland emission. Image reproduced from [[9]](https://doi.org/10.1016/j.rse.2018.02.011) under CC-BY license.\n```\n\nThe analysis and results are organised in the following steps, which are detailed in the sections below:\n\n**[](template:section-1)**\n * We inspected the uncertainty related to the data product used by [[9]](https://doi.org/10.1016/j.rse.2018.02.011) and we compared with that reported by the product quality assessment report related to \"Methane data from 2002 to present derived from satellite observations\" ([[8]](http://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PQAR_GHG_main_latest.pdf)).\n\n**[](template:section-2)**\n * We inspected the spatial and temporal resolution of the data product used by [[9]](https://doi.org/10.1016/j.rse.2018.02.011) and we compared with that reported by the Level 2 of the dataset \"Methane data from 2002 to present derived from satellite observations\" ([[6]](http://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PUGS_GHG_main_latest.pdf)).", "text_with_prefix": "EQC Quality Assessment: \"Methane satellite observations uncertainty and completeness assessment for carbon cycle\"\nDataset: satellite-methane [CDS]\nAspect: uncertainty_q01 | Category: Satellite_ECVs\nSection: Methane satellite observations uncertainty and completeness assessment for carbon cycle > Quality assessment statement\n---\nThese are the key outcomes of this assessment\n\n* In terms of uncertainty, Level 2 XCH4 data products appear appropriate for investigating inter-annual XCH4 variability associated with anomalies in tropical wetland emissions\n* In terms of spatial/temporal resolution, Level 2 XCH4 data products appear appropriate for investigating inter-annual XCH4 variability associated with anomalies in tropical wetland emissions\n* In terms of spatial/temporal coverage, users would use data products derived by algorithms (such as CH4_GOS_OCPR, CH4_GO2_SRPR or XCH4_EMMA) that optimise data availability over regions frequently affected by cloud cover or high aerosol loading. \n* Users should consult the documentation as well as the uncertainty and quality flags in the data files as appropriate for their applications. It should also be emphasised that the use of the XCH4 data provided by satellite is not trivial, and interpretation of these products typically requires appropriate modelling [[6]](http://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PUGS_GHG_main_latest.pdf).\n* Users will need to consult product user guides ([[6]](http://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PUGS_GHG_main_latest.pdf)) for guidance on how to use the averaging kernels when using satellite CH4 data, especially for comparison with model outputs\n\n## 📋 Methodology\n\nTo assess if the uncertainty, spatial/temporal resolution and coverage of satellite-based observations are sufficient to investigate year-to-year anomalies in CH4 emissions of tropical wetlands, we inspected the documentation of the CDS dataset “Methane data from 2002 to present derived from satellite observations” ([[7]](https://cds.climate.copernicus.eu/datasets/satellite-methane?tab=overview), [[6]](http://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PUGS_GHG_main_latest.pdf), [[8]](http://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PQAR_GHG_main_latest.pdf)) as well as a case study available from scientific literature ([[9]](https://doi.org/10.1016/j.rse.2018.02.011)). Please note that to analyse annual anomalies in wetland methane emissions, [[9]](https://doi.org/10.1016/j.rse.2018.02.011) used an earlier version (v7) of Level 2 data from the CH4_GOS_OCPR data product.\n\nattachment:88fcf479-6fcf-4d60-834f-18e8ca2c2bb4.png\n---\nheight: 500px\n---\nMap showing the location of the individual wetland regions identified using the Sustainable Wetlands Adaptation and Mitigation Program (SWAMP) data from the Center for International Forestry Research (CIFOR) and timeseries of detrended XCH4 seasonal cycle over the Pantanal region, including the CH4_GOS_OCFP dataset (“GOSAT”) and model simulations utilising the different wetland emission. Image reproduced from [[9]](https://doi.org/10.1016/j.rse.2018.02.011) under CC-BY license.\n```\n\nThe analysis and results are organised in the following steps, which are detailed in the sections below:\n\n**[](template:section-1)**\n * We inspected the uncertainty related to the data product used by [[9]](https://doi.org/10.1016/j.rse.2018.02.011) and we compared with that reported by the product quality assessment report related to \"Methane data from 2002 to present derived from satellite observations\" ([[8]](http://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PQAR_GHG_main_latest.pdf)).\n\n**[](template:section-2)**\n * We inspected the spatial and temporal resolution of the data product used by [[9]](https://doi.org/10.1016/j.rse.2018.02.011) and we compared with that reported by the Level 2 of the dataset \"Methane data from 2002 to present derived from satellite observations\" ([[6]](http://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PUGS_GHG_main_latest.pdf))."} {"chunk_id": "satellite_satellite-methane_uncertainty_q01__d808fab09a52", "report_id": "satellite_satellite-methane_uncertainty_q01", "dataset_id": "satellite-methane", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty_q01", "aspect_base": "uncertainty", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Methane satellite observations uncertainty and completeness assessment for carbon cycle > Quality assessment statement", "title": "Methane satellite observations uncertainty and completeness assessment for carbon cycle", "chunk_index": 3, "token_count": 1118, "text_raw": "to present derived from satellite observations\" ([[6]](http://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PUGS_GHG_main_latest.pdf)).\n\n**[](template:section-3)**\n * [[9]](https://doi.org/10.1016/j.rse.2018.02.011) used a 'proxy (PR)' data product (CH4_GOS_OCPR) to take advantage of the higher data availability compared to the 'full physics (FP)' product (CH4_GOS_OCFP) in regions such as the tropics, which are characterised by the prevalence of clouds and high aerosol loading. PR products typically contain more data points and better coverage compared to FP: PR products suffer less from potential biases and therefore require less strict quality filtering resulting in more data points with quality flag “good” in the final product files ([[6]](http://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PUGS_GHG_main_latest.pdf)).\n\n## 📈 Analysis and results\n\n(template:section-1)=\n### 1. Uncertainty assessment\n\nThe data product used by [[9]](https://doi.org/10.1016/j.rse.2018.02.011) was characterised by a random error of 13 ppb and a relative spatial bias of 2 ppb with respect to TCCON co-located observations ([[10]](https://doi.org/10.5194/acp-21-4339-2021)). According with [[8]](http://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PQAR_GHG_main_latest.pdf), the **Level 2 XCH4 products** of the dataset \"Methane data from 2002 to present derived from satellite observations\" have **uncertainties comparable to those of the dataset used by [[9]](https://doi.org/10.1016/j.rse.2018.02.011)**: individual measurement accuracies are typically less than 20 ppb (13 ppb for the latest version of CH4_GOS_OCPR), while the relative spatial bias is less than 6.2 ppb (4 ppb for CH4_GOS_OCPR). Moreover, **except for the SCIAMACHY XCH4 products**, the single measurements random errors are close to the \"breakthrough\" requirement (< 17 ppb) defined by the Climate Research Group (CRG) of the ESA GHG-CCI project for the usability of satellite measurements to derive information on regional scale sources and sink of methane [[11]](https://www.iup.uni-bremen.de/carbon_ghg/docs/GHG-CCIplus/URD/URDv3.0_GHG-CCIp_Final.pdf).\n\nThus, **the Level 2 XCH4 products of the dataset \"Methane data from 2002 to present derived from satellite observations\" have uncertainties sufficiently low to assess year-to-year anomalies in XCH4 related to wetland emissions in tropical regions.**\n\n(template:section-2)=\n### 2. Spatial resolution assessment\n\nThe Level 2 XCH4 products of the dataset \"Methane data from 2002 to present derived from satellite observations\" are an updated version of the dataset used by [[10]](https://doi.org/10.5194/acp-21-4339-2021) and **they have a spatial resolution characterised by the native satellite footprint** (i.e., with a 10 km diameter and single observations typically on the order of 100 km apart for GOSAT and GOSAT-2).\n\nThus, **the Level 2 XCH4 products of the dataset \"Methane data from 2002 to present derived from satellite observations\" have spatial resolution sufficiently high to assess year-to-year anomalies in XCH4 related to wetland emissions in tropical regions.**\n\n(template:section-3)=\n### 3. Spatial and temporal coverage\n\nThe product user guide related to \"Methane data from 2002 to present derived from satellite observations\" ([[6]](http://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PUGS_GHG_main_latest.pdf)) was inspected. In terms of spatial/temporal coverage users would use data products derived by algorithms (such as CH4_GOS_OCPR, CH4_GO2_SRPR or XCH4_EMMA) that **optimise data availability over regions frequently affected by cloud cover or high aerosol loading**. The temporal coverage of algorithms using GOSAT and GOSAT-2 observations (as done by [[9]](https://doi.org/10.1016/j.rse.2018.02.011)) ranges from 2009 to 2022.", "text_with_prefix": "EQC Quality Assessment: \"Methane satellite observations uncertainty and completeness assessment for carbon cycle\"\nDataset: satellite-methane [CDS]\nAspect: uncertainty_q01 | Category: Satellite_ECVs\nSection: Methane satellite observations uncertainty and completeness assessment for carbon cycle > Quality assessment statement\n---\nto present derived from satellite observations\" ([[6]](http://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PUGS_GHG_main_latest.pdf)).\n\n**[](template:section-3)**\n * [[9]](https://doi.org/10.1016/j.rse.2018.02.011) used a 'proxy (PR)' data product (CH4_GOS_OCPR) to take advantage of the higher data availability compared to the 'full physics (FP)' product (CH4_GOS_OCFP) in regions such as the tropics, which are characterised by the prevalence of clouds and high aerosol loading. PR products typically contain more data points and better coverage compared to FP: PR products suffer less from potential biases and therefore require less strict quality filtering resulting in more data points with quality flag “good” in the final product files ([[6]](http://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PUGS_GHG_main_latest.pdf)).\n\n## 📈 Analysis and results\n\n(template:section-1)=\n### 1. Uncertainty assessment\n\nThe data product used by [[9]](https://doi.org/10.1016/j.rse.2018.02.011) was characterised by a random error of 13 ppb and a relative spatial bias of 2 ppb with respect to TCCON co-located observations ([[10]](https://doi.org/10.5194/acp-21-4339-2021)). According with [[8]](http://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PQAR_GHG_main_latest.pdf), the **Level 2 XCH4 products** of the dataset \"Methane data from 2002 to present derived from satellite observations\" have **uncertainties comparable to those of the dataset used by [[9]](https://doi.org/10.1016/j.rse.2018.02.011)**: individual measurement accuracies are typically less than 20 ppb (13 ppb for the latest version of CH4_GOS_OCPR), while the relative spatial bias is less than 6.2 ppb (4 ppb for CH4_GOS_OCPR). Moreover, **except for the SCIAMACHY XCH4 products**, the single measurements random errors are close to the \"breakthrough\" requirement (< 17 ppb) defined by the Climate Research Group (CRG) of the ESA GHG-CCI project for the usability of satellite measurements to derive information on regional scale sources and sink of methane [[11]](https://www.iup.uni-bremen.de/carbon_ghg/docs/GHG-CCIplus/URD/URDv3.0_GHG-CCIp_Final.pdf).\n\nThus, **the Level 2 XCH4 products of the dataset \"Methane data from 2002 to present derived from satellite observations\" have uncertainties sufficiently low to assess year-to-year anomalies in XCH4 related to wetland emissions in tropical regions.**\n\n(template:section-2)=\n### 2. Spatial resolution assessment\n\nThe Level 2 XCH4 products of the dataset \"Methane data from 2002 to present derived from satellite observations\" are an updated version of the dataset used by [[10]](https://doi.org/10.5194/acp-21-4339-2021) and **they have a spatial resolution characterised by the native satellite footprint** (i.e., with a 10 km diameter and single observations typically on the order of 100 km apart for GOSAT and GOSAT-2).\n\nThus, **the Level 2 XCH4 products of the dataset \"Methane data from 2002 to present derived from satellite observations\" have spatial resolution sufficiently high to assess year-to-year anomalies in XCH4 related to wetland emissions in tropical regions.**\n\n(template:section-3)=\n### 3. Spatial and temporal coverage\n\nThe product user guide related to \"Methane data from 2002 to present derived from satellite observations\" ([[6]](http://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PUGS_GHG_main_latest.pdf)) was inspected. In terms of spatial/temporal coverage users would use data products derived by algorithms (such as CH4_GOS_OCPR, CH4_GO2_SRPR or XCH4_EMMA) that **optimise data availability over regions frequently affected by cloud cover or high aerosol loading**. The temporal coverage of algorithms using GOSAT and GOSAT-2 observations (as done by [[9]](https://doi.org/10.1016/j.rse.2018.02.011)) ranges from 2009 to 2022."} {"chunk_id": "satellite_satellite-methane_uncertainty_q01__03c28c324306", "report_id": "satellite_satellite-methane_uncertainty_q01", "dataset_id": "satellite-methane", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty_q01", "aspect_base": "uncertainty", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Methane satellite observations uncertainty and completeness assessment for carbon cycle > Quality assessment statement", "title": "Methane satellite observations uncertainty and completeness assessment for carbon cycle", "chunk_index": 4, "token_count": 1114, "text_raw": "algorithms using GOSAT and GOSAT-2 observations (as done by [[9]](https://doi.org/10.1016/j.rse.2018.02.011)) ranges from 2009 to 2022.\n\nThus, **the level 2 XCH4 products of the dataset \"Methane data from 2002 to present derived from satellite observations\" have sufficient spatial and temporal coverage to assess year-to-year anomalies in XCH4 related to wetland emissions in tropical regions.** In any case, users should consult the documentation as well as the uncertainty and quality flags in the data files as appropriate for their applications.\n\nattachment:d0fa8c88-b432-41af-8f69-37e786fcb898.png\n---\nheight: 500px\n---\nFigure showing a comparison between FP and PR Level 2 XCH4 products from the University of Leicester GOSAT products from the CDS for January to June 2021 (top) and July to December 2021 (bottom). Right: “CH4_GOS_OCFP algorithm”. Left: “CH4_GOS_OCPR algorithm”. Image adapted from [[6]](http://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PUGS_GHG_main_latest.pdf) under CC-BY license.\n```\n\n## ℹ️ If you want to know more\n\n### Key resources\n\nThe CDS catalogue entries for the data used were:\n* Methane data from 2002 to present derived from satellite observations: https://cds.climate.copernicus.eu/datasets/satellite-methane?tab=overview\n\nMore about wetlands and C3S products that assist wetland conservation: \n* https://climate.copernicus.eu/working-our-wetlands\n\nMore about greenhouse gases by the World Meteorological Organization (WMO): \n* https://wmo.int/publication-series/greenhouse-gas-bulletin\n\n### References\n\n[[1]](https://library.wmo.int/viewer/68532/?offset=#page=4&viewer=picture&o=bookmarks&n=0&q=) World Meteorological Organization (2023). WMO Greenhouse Gas Bulletin, No. 19, ISSN 2078-0796.\n\n[[2]](https://doi.org/10.5194/essd-12-1561-2020) Saunois, M., Stavert, A. R., Poulter, B., Bousquet, P., Canadell, J. G., et al. (2020). The Global Methane Budget 2000–2017. Earth System Science Data, 12, 1561–1623.\n\n[[3]](https://doi.org/10.1038/s41558-023-01629-0) Zhang, Z., Poulter, B., Feldman, A.F. et al. (2023). Recent intensification of wetland methane feedback. Nature Climate Change, 13, 430–433.\n\n[[4]](https://doi.org/10.1126/sciadv.aay4444) Ernest N. Koffi et al. (2020). An observation-constrained assessment of the climate sensitivity and future trajectories of wetland methane emissions. Science Advances, 6,eaay4444.\n\n[[5]](https://ads.atmosphere.copernicus.eu/datasets/cams-global-greenhouse-gas-inversion?tab=overview) Copernicus Atmosphere Monitoring Service (2020). CAMS global inversion-optimised greenhouse gas fluxes and concentrations.\n\n[[6]](http://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PUGS_GHG_main_latest.pdf) Buchwitz, M. (2023). Product User Guide and Specification (PUGS) – Main document for Greenhouse Gas (GHG: CO2 & CH4) data set CDR6 (01.2003-12.2021).\n\n[[7]](https://cds.climate.copernicus.eu/datasets/satellite-methane?tab=overview) Copernicus Climate Change Service, Climate Data Store, (2018): Methane data from 2002 to present derived from satellite observations. Copernicus Climate Change Service (C3S) Climate Data Store (CDS). DOI: 10.24381/cds.b25419f8\n\n[[8]](http://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PQAR_GHG_main_latest.pdf) Buchwitz, M. (2023). Product Quality Assessment Report (PQAR) – Main document for Greenhouse Gas (GHG: CO2 & CH4) data set CDR6 (01.2003-12.2021).", "text_with_prefix": "EQC Quality Assessment: \"Methane satellite observations uncertainty and completeness assessment for carbon cycle\"\nDataset: satellite-methane [CDS]\nAspect: uncertainty_q01 | Category: Satellite_ECVs\nSection: Methane satellite observations uncertainty and completeness assessment for carbon cycle > Quality assessment statement\n---\nalgorithms using GOSAT and GOSAT-2 observations (as done by [[9]](https://doi.org/10.1016/j.rse.2018.02.011)) ranges from 2009 to 2022.\n\nThus, **the level 2 XCH4 products of the dataset \"Methane data from 2002 to present derived from satellite observations\" have sufficient spatial and temporal coverage to assess year-to-year anomalies in XCH4 related to wetland emissions in tropical regions.** In any case, users should consult the documentation as well as the uncertainty and quality flags in the data files as appropriate for their applications.\n\nattachment:d0fa8c88-b432-41af-8f69-37e786fcb898.png\n---\nheight: 500px\n---\nFigure showing a comparison between FP and PR Level 2 XCH4 products from the University of Leicester GOSAT products from the CDS for January to June 2021 (top) and July to December 2021 (bottom). Right: “CH4_GOS_OCFP algorithm”. Left: “CH4_GOS_OCPR algorithm”. Image adapted from [[6]](http://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PUGS_GHG_main_latest.pdf) under CC-BY license.\n```\n\n## ℹ️ If you want to know more\n\n### Key resources\n\nThe CDS catalogue entries for the data used were:\n* Methane data from 2002 to present derived from satellite observations: https://cds.climate.copernicus.eu/datasets/satellite-methane?tab=overview\n\nMore about wetlands and C3S products that assist wetland conservation: \n* https://climate.copernicus.eu/working-our-wetlands\n\nMore about greenhouse gases by the World Meteorological Organization (WMO): \n* https://wmo.int/publication-series/greenhouse-gas-bulletin\n\n### References\n\n[[1]](https://library.wmo.int/viewer/68532/?offset=#page=4&viewer=picture&o=bookmarks&n=0&q=) World Meteorological Organization (2023). WMO Greenhouse Gas Bulletin, No. 19, ISSN 2078-0796.\n\n[[2]](https://doi.org/10.5194/essd-12-1561-2020) Saunois, M., Stavert, A. R., Poulter, B., Bousquet, P., Canadell, J. G., et al. (2020). The Global Methane Budget 2000–2017. Earth System Science Data, 12, 1561–1623.\n\n[[3]](https://doi.org/10.1038/s41558-023-01629-0) Zhang, Z., Poulter, B., Feldman, A.F. et al. (2023). Recent intensification of wetland methane feedback. Nature Climate Change, 13, 430–433.\n\n[[4]](https://doi.org/10.1126/sciadv.aay4444) Ernest N. Koffi et al. (2020). An observation-constrained assessment of the climate sensitivity and future trajectories of wetland methane emissions. Science Advances, 6,eaay4444.\n\n[[5]](https://ads.atmosphere.copernicus.eu/datasets/cams-global-greenhouse-gas-inversion?tab=overview) Copernicus Atmosphere Monitoring Service (2020). CAMS global inversion-optimised greenhouse gas fluxes and concentrations.\n\n[[6]](http://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PUGS_GHG_main_latest.pdf) Buchwitz, M. (2023). Product User Guide and Specification (PUGS) – Main document for Greenhouse Gas (GHG: CO2 & CH4) data set CDR6 (01.2003-12.2021).\n\n[[7]](https://cds.climate.copernicus.eu/datasets/satellite-methane?tab=overview) Copernicus Climate Change Service, Climate Data Store, (2018): Methane data from 2002 to present derived from satellite observations. Copernicus Climate Change Service (C3S) Climate Data Store (CDS). DOI: 10.24381/cds.b25419f8\n\n[[8]](http://wdc.dlr.de/C3S_312b_Lot2/Documentation/GHG/C3S2_312a_Lot2_PQAR_GHG_main_latest.pdf) Buchwitz, M. (2023). Product Quality Assessment Report (PQAR) – Main document for Greenhouse Gas (GHG: CO2 & CH4) data set CDR6 (01.2003-12.2021)."} {"chunk_id": "satellite_satellite-methane_uncertainty_q01__bbeab590cc29", "report_id": "satellite_satellite-methane_uncertainty_q01", "dataset_id": "satellite-methane", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty_q01", "aspect_base": "uncertainty", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Methane satellite observations uncertainty and completeness assessment for carbon cycle > Quality assessment statement", "title": "Methane satellite observations uncertainty and completeness assessment for carbon cycle", "chunk_index": 5, "token_count": 458, "text_raw": "itz, M. (2023). Product Quality Assessment Report (PQAR) – Main document for Greenhouse Gas (GHG: CO2 & CH4) data set CDR6 (01.2003-12.2021).\n\n[[9]](https://doi.org/10.1016/j.rse.2018.02.011) Parker, R.J., Boesch, H., McNorton, J., Comyn-Platt, E., Gloor, M., Wilson, C., Chipperfield, M.P., Hayman, G.D., Bloom, A.A. (2018). Evaluating year-to-year anomalies in tropical wetland methane emissions using satellite CH4 observations. Remote Sensing of Environment, 211, 261-275.\n\n[[10]](https://doi.org/10.5194/acp-21-4339-2021) Maasakkers, J. D., Jacob, D. J., Sulprizio, M. P., Scarpelli, T. R., Nesser, H., Sheng, J., Zhang, Y., Lu, X., Bloom, A. A., Bowman, K. W., Worden, J. R., and Parker, R. J. (2021). 2010–2015 North American methane emissions, sectoral contributions, and trends: a high-resolution inversion of GOSAT observations of atmospheric methane. Atmospheric Chemistry and Physics, 21, 4339–4356.\n\n[[11]](https://www.iup.uni-bremen.de/carbon_ghg/docs/GHG-CCIplus/URD/URDv3.0_GHG-CCIp_Final.pdf) Chevallier, F. et al. (2020). User Requirements Document (URD), ESA Climate Change Initiative (CCI) GHG-CCI project, Version 3.0, pp.42.", "text_with_prefix": "EQC Quality Assessment: \"Methane satellite observations uncertainty and completeness assessment for carbon cycle\"\nDataset: satellite-methane [CDS]\nAspect: uncertainty_q01 | Category: Satellite_ECVs\nSection: Methane satellite observations uncertainty and completeness assessment for carbon cycle > Quality assessment statement\n---\nitz, M. (2023). Product Quality Assessment Report (PQAR) – Main document for Greenhouse Gas (GHG: CO2 & CH4) data set CDR6 (01.2003-12.2021).\n\n[[9]](https://doi.org/10.1016/j.rse.2018.02.011) Parker, R.J., Boesch, H., McNorton, J., Comyn-Platt, E., Gloor, M., Wilson, C., Chipperfield, M.P., Hayman, G.D., Bloom, A.A. (2018). Evaluating year-to-year anomalies in tropical wetland methane emissions using satellite CH4 observations. Remote Sensing of Environment, 211, 261-275.\n\n[[10]](https://doi.org/10.5194/acp-21-4339-2021) Maasakkers, J. D., Jacob, D. J., Sulprizio, M. P., Scarpelli, T. R., Nesser, H., Sheng, J., Zhang, Y., Lu, X., Bloom, A. A., Bowman, K. W., Worden, J. R., and Parker, R. J. (2021). 2010–2015 North American methane emissions, sectoral contributions, and trends: a high-resolution inversion of GOSAT observations of atmospheric methane. Atmospheric Chemistry and Physics, 21, 4339–4356.\n\n[[11]](https://www.iup.uni-bremen.de/carbon_ghg/docs/GHG-CCIplus/URD/URDv3.0_GHG-CCIp_Final.pdf) Chevallier, F. et al. (2020). User Requirements Document (URD), ESA Climate Change Initiative (CCI) GHG-CCI project, Version 3.0, pp.42."} {"chunk_id": "satellite_satellite-ozone-v1_climate-monitoring_q01__4e6dee82c6c1", "report_id": "satellite_satellite-ozone-v1_climate-monitoring_q01", "dataset_id": "satellite-ozone-v1", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-monitoring_q01", "aspect_base": "climate-monitoring", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Temporal and spatial completeness of the satellite-derived Ozone for policy making and variability studies", "title": "Temporal and spatial completeness of the satellite-derived Ozone for policy making and variability studies", "chunk_index": 0, "token_count": 112, "text_raw": "Production date: 06-06-2024\n\nProduced by: Sandro Calmanti (ENEA), Chiara Volta (ENEA), Irene Cionni (ENEA)", "text_with_prefix": "EQC Quality Assessment: \"Temporal and spatial completeness of the satellite-derived Ozone for policy making and variability studies\"\nDataset: satellite-ozone-v1 [CDS]\nAspect: climate-monitoring_q01 | Category: Satellite_ECVs\nSection: Temporal and spatial completeness of the satellite-derived Ozone for policy making and variability studies\n---\nProduction date: 06-06-2024\n\nProduced by: Sandro Calmanti (ENEA), Chiara Volta (ENEA), Irene Cionni (ENEA)"} {"chunk_id": "satellite_satellite-ozone-v1_climate-monitoring_q01__c51f8a6ed0bf", "report_id": "satellite_satellite-ozone-v1_climate-monitoring_q01", "dataset_id": "satellite-ozone-v1", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-monitoring_q01", "aspect_base": "climate-monitoring", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Temporal and spatial completeness of the satellite-derived Ozone for policy making and variability studies > Quality assessment questions", "title": "Temporal and spatial completeness of the satellite-derived Ozone for policy making and variability studies", "chunk_index": 1, "token_count": 430, "text_raw": "* **How well, in terms of completeness, does the longest Level-3 atmospheric ozone content data from merged satellite missions represent the variability of the ozone layer over time?**\n* **How well data from merged satellite missions capture extreme events in ozone concentration?**\n\nAssessing the return of the ozone layer to historical levels and the full recovery from ozone-depleting substances (ODS) under the 1987 Montreal Protocol is a critical task. The Montreal Protocol required the international scientific community to provide governments with regular updates on the latest scientific knowledge about the ozone layer. Since 1987, ten reports called \"Scientific Assessment of Ozone Depletion\" have been produced. These regular assessments have guided policymakers in strengthening the provisions of the Montreal Protocol.\n\nSatellite data undergoes several levels of processing before it can be used to provide decision-relevant information.\n\nIn this application, the Level-3 MERGED-UV dataset version v2000 is evaluated. A Level 3 dataset corresponds to data from a specific satellite sensor that has been quality tagged, converted to the relevant environmental variable, and aggregated on a uniform time-space grid. Level 3 datasets therefore represent a relatively low level of processing to produce usable data for decision making.\n\nThe MERGED-UV dataset combines total column ozone measurements at a 1°x1° spatial resolution from 5 UV nadir satellite sensors (i.e., GOME, SCIAMACHY, GOME-2A/B and OMI) [[1]](https://dast.copernicus-climate.eu/documents/satellite-ozone/C3S2_312a_Lot2_PUGS_O3_latest.pdf) , resulting in the longest L3 satellite product currently available (1995 to present).", "text_with_prefix": "EQC Quality Assessment: \"Temporal and spatial completeness of the satellite-derived Ozone for policy making and variability studies\"\nDataset: satellite-ozone-v1 [CDS]\nAspect: climate-monitoring_q01 | Category: Satellite_ECVs\nSection: Temporal and spatial completeness of the satellite-derived Ozone for policy making and variability studies > Quality assessment questions\n---\n* **How well, in terms of completeness, does the longest Level-3 atmospheric ozone content data from merged satellite missions represent the variability of the ozone layer over time?**\n* **How well data from merged satellite missions capture extreme events in ozone concentration?**\n\nAssessing the return of the ozone layer to historical levels and the full recovery from ozone-depleting substances (ODS) under the 1987 Montreal Protocol is a critical task. The Montreal Protocol required the international scientific community to provide governments with regular updates on the latest scientific knowledge about the ozone layer. Since 1987, ten reports called \"Scientific Assessment of Ozone Depletion\" have been produced. These regular assessments have guided policymakers in strengthening the provisions of the Montreal Protocol.\n\nSatellite data undergoes several levels of processing before it can be used to provide decision-relevant information.\n\nIn this application, the Level-3 MERGED-UV dataset version v2000 is evaluated. A Level 3 dataset corresponds to data from a specific satellite sensor that has been quality tagged, converted to the relevant environmental variable, and aggregated on a uniform time-space grid. Level 3 datasets therefore represent a relatively low level of processing to produce usable data for decision making.\n\nThe MERGED-UV dataset combines total column ozone measurements at a 1°x1° spatial resolution from 5 UV nadir satellite sensors (i.e., GOME, SCIAMACHY, GOME-2A/B and OMI) [[1]](https://dast.copernicus-climate.eu/documents/satellite-ozone/C3S2_312a_Lot2_PUGS_O3_latest.pdf) , resulting in the longest L3 satellite product currently available (1995 to present)."} {"chunk_id": "satellite_satellite-ozone-v1_climate-monitoring_q01__98fab4e6490a", "report_id": "satellite_satellite-ozone-v1_climate-monitoring_q01", "dataset_id": "satellite-ozone-v1", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-monitoring_q01", "aspect_base": "climate-monitoring", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Temporal and spatial completeness of the satellite-derived Ozone for policy making and variability studies > Quality assessment statements", "title": "Temporal and spatial completeness of the satellite-derived Ozone for policy making and variability studies", "chunk_index": 2, "token_count": 365, "text_raw": "These are the key outcomes of this assessment\n\n* The MERGED-UV v2000 provides a long-term, consistent record of total column ozone concentration and is therefore valuable for providing insight into the evolution of the ozone layer, although data prior to 2004 should be carefully evaluated.\n* The MERGED-UV v2000 does not include measurements during the polar night at latitudes higher than 57.5° in both hemispheres and therefore cannot be used to describe the ozone climatology and trends over these regions.\n* The lack of data in high-latitude regions during the polar night hinders the use of the MERGED-UV v2000 dataset to monitor the evolution of alarming ozone depletion at the poles.\n* Data gaps in the polar region are a direct consequence of the decision to produce a long-term consistent ozone data set suitable for trend analysis. If data coverage is to be preferred to consistency, a better choice would be to use the ozone multi-sensor reanalysis (MSR) [[2]](https://doi.org/10.5194/amt-8-3021-2015) available from 1970 to present.\n* Despite the data gaps, the MERGED-UV v2000 dataset reproduces the major ozone variations over time described in the literature.\n* MERGED-UV v2000 captures episodes of extremely low regional total column ozone associated with the impact of volcanic eruptions.\n```", "text_with_prefix": "EQC Quality Assessment: \"Temporal and spatial completeness of the satellite-derived Ozone for policy making and variability studies\"\nDataset: satellite-ozone-v1 [CDS]\nAspect: climate-monitoring_q01 | Category: Satellite_ECVs\nSection: Temporal and spatial completeness of the satellite-derived Ozone for policy making and variability studies > Quality assessment statements\n---\nThese are the key outcomes of this assessment\n\n* The MERGED-UV v2000 provides a long-term, consistent record of total column ozone concentration and is therefore valuable for providing insight into the evolution of the ozone layer, although data prior to 2004 should be carefully evaluated.\n* The MERGED-UV v2000 does not include measurements during the polar night at latitudes higher than 57.5° in both hemispheres and therefore cannot be used to describe the ozone climatology and trends over these regions.\n* The lack of data in high-latitude regions during the polar night hinders the use of the MERGED-UV v2000 dataset to monitor the evolution of alarming ozone depletion at the poles.\n* Data gaps in the polar region are a direct consequence of the decision to produce a long-term consistent ozone data set suitable for trend analysis. If data coverage is to be preferred to consistency, a better choice would be to use the ozone multi-sensor reanalysis (MSR) [[2]](https://doi.org/10.5194/amt-8-3021-2015) available from 1970 to present.\n* Despite the data gaps, the MERGED-UV v2000 dataset reproduces the major ozone variations over time described in the literature.\n* MERGED-UV v2000 captures episodes of extremely low regional total column ozone associated with the impact of volcanic eruptions.\n```"} {"chunk_id": "satellite_satellite-ozone-v1_climate-monitoring_q01__2274cbabda26", "report_id": "satellite_satellite-ozone-v1_climate-monitoring_q01", "dataset_id": "satellite-ozone-v1", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-monitoring_q01", "aspect_base": "climate-monitoring", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Temporal and spatial completeness of the satellite-derived Ozone for policy making and variability studies > Methodology", "title": "Temporal and spatial completeness of the satellite-derived Ozone for policy making and variability studies", "chunk_index": 3, "token_count": 210, "text_raw": "This notebook has two main goals:\n\n- to show how the basic data coverage has changed over time and to highlight the potential impact on the accuracy of the final product;\n - to illustrate how the satellite orbit affects the seasonal data coverage and the resulting impact on the monitoring of total ozone over critical regions.\n\nThe methodology adopted for the analysis is split into the following steps:\n\n[](sec1)\n* Define data request and other parameters.\n* Define required functions\n\n[](sec2)\n* Unit conversion\n* Computation of time series\n\n[](sec3)\n* Discussion of data coverage\n* Analysis of ozone distribution and its seasonal cycle\n* Time series of total column ozone (TCO)", "text_with_prefix": "EQC Quality Assessment: \"Temporal and spatial completeness of the satellite-derived Ozone for policy making and variability studies\"\nDataset: satellite-ozone-v1 [CDS]\nAspect: climate-monitoring_q01 | Category: Satellite_ECVs\nSection: Temporal and spatial completeness of the satellite-derived Ozone for policy making and variability studies > Methodology\n---\nThis notebook has two main goals:\n\n- to show how the basic data coverage has changed over time and to highlight the potential impact on the accuracy of the final product;\n - to illustrate how the satellite orbit affects the seasonal data coverage and the resulting impact on the monitoring of total ozone over critical regions.\n\nThe methodology adopted for the analysis is split into the following steps:\n\n[](sec1)\n* Define data request and other parameters.\n* Define required functions\n\n[](sec2)\n* Unit conversion\n* Computation of time series\n\n[](sec3)\n* Discussion of data coverage\n* Analysis of ozone distribution and its seasonal cycle\n* Time series of total column ozone (TCO)"} {"chunk_id": "satellite_satellite-ozone-v1_climate-monitoring_q01__8d9667ea31d3", "report_id": "satellite_satellite-ozone-v1_climate-monitoring_q01", "dataset_id": "satellite-ozone-v1", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-monitoring_q01", "aspect_base": "climate-monitoring", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Temporal and spatial completeness of the satellite-derived Ozone for policy making and variability studies > Analysis and results > 1. Choose the data to use and setup the code > Define data request and other parameters", "title": "Temporal and spatial completeness of the satellite-derived Ozone for policy making and variability studies", "chunk_index": 4, "token_count": 224, "text_raw": "The analysis performed in this notebook focuses on the time series of the MERGED-UV total column ozone (TCO) over a period of 26 years (January 1996 - December 2022).\n\nAdditionally, spatially weighted means and climatological monthly averages are computed at the global scale and over 5 latitudinal bands: polar (latitudes > 60°) and mid-latitude (30° Analysis and results > 1. Choose the data to use and setup the code > Define data request and other parameters\n---\nThe analysis performed in this notebook focuses on the time series of the MERGED-UV total column ozone (TCO) over a period of 26 years (January 1996 - December 2022).\n\nAdditionally, spatially weighted means and climatological monthly averages are computed at the global scale and over 5 latitudinal bands: polar (latitudes > 60°) and mid-latitude (30° Analysis and results > 1. Choose the data to use and setup the code > Define required functions", "title": "Temporal and spatial completeness of the satellite-derived Ozone for policy making and variability studies", "chunk_index": 5, "token_count": 141, "text_raw": "For this dataset it is necessary to handle the time dimension carefully with a simple dedicated function like `add_time_dim`. \nThe function `spatial_weighted_mean` is used to calculate ozone averages over the regions of interest.\n\n(sec2)=", "text_with_prefix": "EQC Quality Assessment: \"Temporal and spatial completeness of the satellite-derived Ozone for policy making and variability studies\"\nDataset: satellite-ozone-v1 [CDS]\nAspect: climate-monitoring_q01 | Category: Satellite_ECVs\nSection: Temporal and spatial completeness of the satellite-derived Ozone for policy making and variability studies > Analysis and results > 1. Choose the data to use and setup the code > Define required functions\n---\nFor this dataset it is necessary to handle the time dimension carefully with a simple dedicated function like `add_time_dim`. \nThe function `spatial_weighted_mean` is used to calculate ozone averages over the regions of interest.\n\n(sec2)="} {"chunk_id": "satellite_satellite-ozone-v1_climate-monitoring_q01__82fbca1be76b", "report_id": "satellite_satellite-ozone-v1_climate-monitoring_q01", "dataset_id": "satellite-ozone-v1", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-monitoring_q01", "aspect_base": "climate-monitoring", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Temporal and spatial completeness of the satellite-derived Ozone for policy making and variability studies > Analysis and results > 2. Data retrieval", "title": "Temporal and spatial completeness of the satellite-derived Ozone for policy making and variability studies", "chunk_index": 6, "token_count": 251, "text_raw": "The full dataset is retrieved, including information on the number of observations used to compute the monthly TCO at each grid point. All TCO is then converted in Dobson Unit (DU) by multiplying ozone data (in mol m$^{-2}$) by the conversion coefficient provided in the dataset as a metadata attribute.\n\nThe computation of the time series of TCO over the regions of interest is performed by applying a weighted mean to the gridded data.\n\nFull dataset\n\n```text\n100%|██████████| 27/27 [00:00<00:00, 59.02it/s]\n```\n\n```text\nregion='global'\nregion='tropics'\nregion='NH mid-latitudes'\nregion='SH mid-latitudes'\nregion='NH polar'\nregion='SH polar'\n```\n\n(sec3)=", "text_with_prefix": "EQC Quality Assessment: \"Temporal and spatial completeness of the satellite-derived Ozone for policy making and variability studies\"\nDataset: satellite-ozone-v1 [CDS]\nAspect: climate-monitoring_q01 | Category: Satellite_ECVs\nSection: Temporal and spatial completeness of the satellite-derived Ozone for policy making and variability studies > Analysis and results > 2. Data retrieval\n---\nThe full dataset is retrieved, including information on the number of observations used to compute the monthly TCO at each grid point. All TCO is then converted in Dobson Unit (DU) by multiplying ozone data (in mol m$^{-2}$) by the conversion coefficient provided in the dataset as a metadata attribute.\n\nThe computation of the time series of TCO over the regions of interest is performed by applying a weighted mean to the gridded data.\n\nFull dataset\n\n```text\n100%|██████████| 27/27 [00:00<00:00, 59.02it/s]\n```\n\n```text\nregion='global'\nregion='tropics'\nregion='NH mid-latitudes'\nregion='SH mid-latitudes'\nregion='NH polar'\nregion='SH polar'\n```\n\n(sec3)="} {"chunk_id": "satellite_satellite-ozone-v1_climate-monitoring_q01__b49db26c329c", "report_id": "satellite_satellite-ozone-v1_climate-monitoring_q01", "dataset_id": "satellite-ozone-v1", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-monitoring_q01", "aspect_base": "climate-monitoring", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Temporal and spatial completeness of the satellite-derived Ozone for policy making and variability studies > Analysis and results > 3. Plot and describe the results > Data Coverage", "title": "Temporal and spatial completeness of the satellite-derived Ozone for policy making and variability studies", "chunk_index": 7, "token_count": 552, "text_raw": "The global map of data coverage shows that the number of observations decreases from 57.5° poleward in both hemispheres. As shown in the latitudinal time series, this reduction is due to the fact that the dataset does not include measurements during the polar night (i.e., winter and summer months in the northern and the southern hemisphere, respectively). The time series also shows reduced observations between 1996 and 2002, a period during which GOME was the only active sensor. Since 2002, the number of observations slightly increases due to the activation of SCIAMACHY. In 2004, the activation of a third sensor (OMI) strongly enhanced the data coverage. Since then, the data coverage has been guaranteed over time by using at least three sensors simultaneously [[1]](https://dast.copernicus-climate.eu/documents/satellite-ozone/C3S2_312a_Lot2_PUGS_O3_latest.pdf). Small, non-periodic reductions in the number of observations are visible after 2004 and are likely to depend on differences in the sensors and satellites used over time, such as their orbital period and swath [[3]](https://dast.copernicus-climate.eu/documents/satellite-ozone/C3S2_312a_Lot2_ATBD_O3_latest.pdf). Small, but periodic reductions are due to differences in the length of the months (e.g., February).\n\nA significant impact of data coverage on the accuracy of the total ozone content can be derived as a by-product of the data assimilation process for the elaboration of the multi-sensor reanalysis [[2]](https://doi.org/10.5194/amt-8-3021-2015), which is also available on the CDS as the MSR dataset. Increasing the number of measurements used to derive the mean total ozone column by about an order of magnitude (e.g., between 2000 and 2020) corresponds to reducing the absolute error from about 20 DU to about 7 DU, which corresponds to reducing the expected error from about 6% to about 2% [[2]](https://doi.org/10.5194/amt-8-3021-2015).", "text_with_prefix": "EQC Quality Assessment: \"Temporal and spatial completeness of the satellite-derived Ozone for policy making and variability studies\"\nDataset: satellite-ozone-v1 [CDS]\nAspect: climate-monitoring_q01 | Category: Satellite_ECVs\nSection: Temporal and spatial completeness of the satellite-derived Ozone for policy making and variability studies > Analysis and results > 3. Plot and describe the results > Data Coverage\n---\nThe global map of data coverage shows that the number of observations decreases from 57.5° poleward in both hemispheres. As shown in the latitudinal time series, this reduction is due to the fact that the dataset does not include measurements during the polar night (i.e., winter and summer months in the northern and the southern hemisphere, respectively). The time series also shows reduced observations between 1996 and 2002, a period during which GOME was the only active sensor. Since 2002, the number of observations slightly increases due to the activation of SCIAMACHY. In 2004, the activation of a third sensor (OMI) strongly enhanced the data coverage. Since then, the data coverage has been guaranteed over time by using at least three sensors simultaneously [[1]](https://dast.copernicus-climate.eu/documents/satellite-ozone/C3S2_312a_Lot2_PUGS_O3_latest.pdf). Small, non-periodic reductions in the number of observations are visible after 2004 and are likely to depend on differences in the sensors and satellites used over time, such as their orbital period and swath [[3]](https://dast.copernicus-climate.eu/documents/satellite-ozone/C3S2_312a_Lot2_ATBD_O3_latest.pdf). Small, but periodic reductions are due to differences in the length of the months (e.g., February).\n\nA significant impact of data coverage on the accuracy of the total ozone content can be derived as a by-product of the data assimilation process for the elaboration of the multi-sensor reanalysis [[2]](https://doi.org/10.5194/amt-8-3021-2015), which is also available on the CDS as the MSR dataset. Increasing the number of measurements used to derive the mean total ozone column by about an order of magnitude (e.g., between 2000 and 2020) corresponds to reducing the absolute error from about 20 DU to about 7 DU, which corresponds to reducing the expected error from about 6% to about 2% [[2]](https://doi.org/10.5194/amt-8-3021-2015)."} {"chunk_id": "satellite_satellite-ozone-v1_climate-monitoring_q01__07d552b3ea9b", "report_id": "satellite_satellite-ozone-v1_climate-monitoring_q01", "dataset_id": "satellite-ozone-v1", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-monitoring_q01", "aspect_base": "climate-monitoring", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Temporal and spatial completeness of the satellite-derived Ozone for policy making and variability studies > Analysis and results > 3. Plot and describe the results > Total column ozone distribution and seasonal cycle", "title": "Temporal and spatial completeness of the satellite-derived Ozone for policy making and variability studies", "chunk_index": 8, "token_count": 518, "text_raw": "The global map of total column ozone (TCO) shows that TCO varies strongly with latitude. The highest TCO values are observed at latitudes around 60°S and above 30°N due to the poleward transport of tropical ozone by stratospheric air circulation [[4]](https://doi.org/10.1002/2013RG000448). As a result, tropical regions become ozone-depleted. Low TCO values are also observed below 60°S, where the Antarctic ozone hole is located [[5]](https://doi.org/10.1038/315207a0) [[6]](https://doi.org/10.1126/science.aae0061).\n\nThe Hövmoller plot of the climatological monthly TCO means shows that the seasonal variability of ozone levels is small in the tropics and larger above 35° in both hemispheres. The dataset MERGED-UV is able to reproduce the TCO climatology in the tropics and mid-latitudes, with an expected maximum over eastern Asia at about 60°N and the largest TCO values over the northern polar region during winter with TCO above 410 DU [[7]](https://doi.org/10.5194/acp-11-9237-2011). However it fails to reproduce the full climatology in polar regions due to the unavailability of data during the polar night from 57.5° poleward in both hemispheres. In particular, the dataset partially captures the TCO depletion typically observed in early autumn and during the austral spring above and below 60° N and S, respectively, but cannot reproduce the low TCO levels typically observed in early austral autumn [[8]](https://doi.org/10.1029/1999RG900008) [[9]](https://ozone.unep.org/sites/default/files/2023-05/Final_20Qs%202022%20full%20document_26April2023_digital%20version-reduced_0.pdf).", "text_with_prefix": "EQC Quality Assessment: \"Temporal and spatial completeness of the satellite-derived Ozone for policy making and variability studies\"\nDataset: satellite-ozone-v1 [CDS]\nAspect: climate-monitoring_q01 | Category: Satellite_ECVs\nSection: Temporal and spatial completeness of the satellite-derived Ozone for policy making and variability studies > Analysis and results > 3. Plot and describe the results > Total column ozone distribution and seasonal cycle\n---\nThe global map of total column ozone (TCO) shows that TCO varies strongly with latitude. The highest TCO values are observed at latitudes around 60°S and above 30°N due to the poleward transport of tropical ozone by stratospheric air circulation [[4]](https://doi.org/10.1002/2013RG000448). As a result, tropical regions become ozone-depleted. Low TCO values are also observed below 60°S, where the Antarctic ozone hole is located [[5]](https://doi.org/10.1038/315207a0) [[6]](https://doi.org/10.1126/science.aae0061).\n\nThe Hövmoller plot of the climatological monthly TCO means shows that the seasonal variability of ozone levels is small in the tropics and larger above 35° in both hemispheres. The dataset MERGED-UV is able to reproduce the TCO climatology in the tropics and mid-latitudes, with an expected maximum over eastern Asia at about 60°N and the largest TCO values over the northern polar region during winter with TCO above 410 DU [[7]](https://doi.org/10.5194/acp-11-9237-2011). However it fails to reproduce the full climatology in polar regions due to the unavailability of data during the polar night from 57.5° poleward in both hemispheres. In particular, the dataset partially captures the TCO depletion typically observed in early autumn and during the austral spring above and below 60° N and S, respectively, but cannot reproduce the low TCO levels typically observed in early austral autumn [[8]](https://doi.org/10.1029/1999RG900008) [[9]](https://ozone.unep.org/sites/default/files/2023-05/Final_20Qs%202022%20full%20document_26April2023_digital%20version-reduced_0.pdf)."} {"chunk_id": "satellite_satellite-ozone-v1_climate-monitoring_q01__6c941bb7d7e0", "report_id": "satellite_satellite-ozone-v1_climate-monitoring_q01", "dataset_id": "satellite-ozone-v1", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-monitoring_q01", "aspect_base": "climate-monitoring", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Temporal and spatial completeness of the satellite-derived Ozone for policy making and variability studies > Analysis and results > 3. Plot and describe the results > Time series", "title": "Temporal and spatial completeness of the satellite-derived Ozone for policy making and variability studies", "chunk_index": 9, "token_count": 485, "text_raw": "The deseasonalized monthly time series show the limited inter-annual variability at the global scale and over the tropics and mid-latitudes, compared to the larger variability observed over the polar regions. The time series show that the dataset is able to detect documented large variations, such as the large TCO depletion in the Arctic in 2011 and 2020, and the re-emergence of the large, long-lived ozone hole over Antarctica since 2020 due to anomalous circulation patterns [[10]](https://doi.org/10.1038/s41467-023-42637-0) [[11]](https://doi.org/10.1029/2022JD037581). The unique split in the ozone hole that occurred in the second half of 2002 is clearly visible, with anomalies above 50 DU compared to the expected seasonal cycle, consistent with more detailed analysis presented in dedicated studies [[12]](https://doi.org/10.1175/JAS-3338.1). Extremely low TCO levels, triggered by the Calbuco eruption, are also observed below 30°S in 2015 [[6]](https://doi.org/10.1126/science.aae0061). In 2010, high TCO levels are observed at mid-latitudes in the Northern Hemisphere due to simultaneous negative phases of the Arctic and North Atlantic Oscillation [[13]](https://doi.org/10.1029/2010GL046634). High TCO levels, induced by the stratospheric warming over Antarctica, are also observed in the southernmost region in 2019 [[14]](https://doi.org/10.1029/2020GL087810). Furthermore, the time series show that the dataset is able to reproduce the quasi-biennial oscillation in tropical ozone [[15]](https://doi.org/10.1029/2002JD002170).", "text_with_prefix": "EQC Quality Assessment: \"Temporal and spatial completeness of the satellite-derived Ozone for policy making and variability studies\"\nDataset: satellite-ozone-v1 [CDS]\nAspect: climate-monitoring_q01 | Category: Satellite_ECVs\nSection: Temporal and spatial completeness of the satellite-derived Ozone for policy making and variability studies > Analysis and results > 3. Plot and describe the results > Time series\n---\nThe deseasonalized monthly time series show the limited inter-annual variability at the global scale and over the tropics and mid-latitudes, compared to the larger variability observed over the polar regions. The time series show that the dataset is able to detect documented large variations, such as the large TCO depletion in the Arctic in 2011 and 2020, and the re-emergence of the large, long-lived ozone hole over Antarctica since 2020 due to anomalous circulation patterns [[10]](https://doi.org/10.1038/s41467-023-42637-0) [[11]](https://doi.org/10.1029/2022JD037581). The unique split in the ozone hole that occurred in the second half of 2002 is clearly visible, with anomalies above 50 DU compared to the expected seasonal cycle, consistent with more detailed analysis presented in dedicated studies [[12]](https://doi.org/10.1175/JAS-3338.1). Extremely low TCO levels, triggered by the Calbuco eruption, are also observed below 30°S in 2015 [[6]](https://doi.org/10.1126/science.aae0061). In 2010, high TCO levels are observed at mid-latitudes in the Northern Hemisphere due to simultaneous negative phases of the Arctic and North Atlantic Oscillation [[13]](https://doi.org/10.1029/2010GL046634). High TCO levels, induced by the stratospheric warming over Antarctica, are also observed in the southernmost region in 2019 [[14]](https://doi.org/10.1029/2020GL087810). Furthermore, the time series show that the dataset is able to reproduce the quasi-biennial oscillation in tropical ozone [[15]](https://doi.org/10.1029/2002JD002170)."} {"chunk_id": "satellite_satellite-ozone-v1_climate-monitoring_q01__3babd88d4435", "report_id": "satellite_satellite-ozone-v1_climate-monitoring_q01", "dataset_id": "satellite-ozone-v1", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-monitoring_q01", "aspect_base": "climate-monitoring", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Temporal and spatial completeness of the satellite-derived Ozone for policy making and variability studies > ℹ️ If you want to know more > Key resources", "title": "Temporal and spatial completeness of the satellite-derived Ozone for policy making and variability studies", "chunk_index": 10, "token_count": 290, "text_raw": "* CDS catalog entry used in this notebook is the [Ozone monthly gridded data from 1970 to present derived from satellite observations](https://cds.climate.copernicus.eu/datasets/satellite-ozone-v1?tab=overview)\n* Data download from CDS is available [here](https://cds.climate.copernicus.eu/datasets/satellite-ozone-v1?tab=download)\n* A Product User Guide and Specification is available on the [Satellite Ozone Catalogue Entry of the CDS](https://cds.climate.copernicus.eu/datasets/satellite-ozone-v1?tab=documentation).\n\nCode libraries used:\n* [C3S EQC custom function](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)", "text_with_prefix": "EQC Quality Assessment: \"Temporal and spatial completeness of the satellite-derived Ozone for policy making and variability studies\"\nDataset: satellite-ozone-v1 [CDS]\nAspect: climate-monitoring_q01 | Category: Satellite_ECVs\nSection: Temporal and spatial completeness of the satellite-derived Ozone for policy making and variability studies > ℹ️ If you want to know more > Key resources\n---\n* CDS catalog entry used in this notebook is the [Ozone monthly gridded data from 1970 to present derived from satellite observations](https://cds.climate.copernicus.eu/datasets/satellite-ozone-v1?tab=overview)\n* Data download from CDS is available [here](https://cds.climate.copernicus.eu/datasets/satellite-ozone-v1?tab=download)\n* A Product User Guide and Specification is available on the [Satellite Ozone Catalogue Entry of the CDS](https://cds.climate.copernicus.eu/datasets/satellite-ozone-v1?tab=documentation).\n\nCode libraries used:\n* [C3S EQC custom function](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)"} {"chunk_id": "satellite_satellite-ozone-v1_climate-monitoring_q01__e0d016af577a", "report_id": "satellite_satellite-ozone-v1_climate-monitoring_q01", "dataset_id": "satellite-ozone-v1", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-monitoring_q01", "aspect_base": "climate-monitoring", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Temporal and spatial completeness of the satellite-derived Ozone for policy making and variability studies > ℹ️ If you want to know more > References", "title": "Temporal and spatial completeness of the satellite-derived Ozone for policy making and variability studies", "chunk_index": 11, "token_count": 1040, "text_raw": "[[1]](https://dast.copernicus-climate.eu/documents/satellite-ozone/C3S2_312a_Lot2_PUGS_O3_latest.pdf) Van Roozendael, M., et al. (2021a). Ozone Product User Guide and Specification (PUGS). Last access: March 27, 2024\n\n[[2]](https://doi.org/10.5194/amt-8-3021-2015) van der A, R.J., et al. (2015). Extended and refined multi sensor reanalysis of total ozone for the period 1970-2012. Atmospheric Measurement Techniques, 8, 3021-3035.\n\n[[3]](https://dast.copernicus-climate.eu/documents/satellite-ozone/C3S2_312a_Lot2_ATBD_O3_latest.pdf) Van Roozendael, M., et al. (2021b). Ozone Algorithm Theoretical Basis Document (ATBD). Last access: March 28, 2024\n\n[[4]](https://doi.org/10.1002/2013RG000448) Butchart, N. (2014). The Brewer-Dobson circulation. Review of Geophysics, 52(2), 157-184.\n\n[[5]](https://doi.org/10.1038/315207a0) Farman, J.C., et al. (1985). Large losses of total ozone in Antarctica reveal seasonal ClO$_x$/NO$_x$ interaction. Nature, 315, 207-210.\n\n[[6]](https://doi.org/10.1126/science.aae0061) Solomon, S., Ivy, D. J., Kinnison, D., Mills, M. J., Neely III, R. R., & Schmidt, A. (2016). Emergence of healing in the Antarctic ozone layer. Science, 353(6296), 269-274.\n\n[[7]](https://doi.org/10.5194/acp-11-9237-2011) Ziemke, J. R., Chandra, S., Labow, G. J., Bhartia, P. K., Froidevaux, L., & Witte, J. C. (2011). A global climatology of tropospheric and stratospheric ozone derived from Aura OMI and MLS measurements. Atmospheric Chemistry and Physics, 11(17), 9237-9251.\n\n[[8]](https://doi.org/10.1029/1999RG900008) Solomon, S. (1999). Stratospheric ozone depletion: A review of concepts and history. Review of Geophysics, 37(3), 275-316.\n\n[[9]](https://ozone.unep.org/sites/default/files/2023-05/Final_20Qs%202022%20full%20document_26April2023_digital%20version-reduced_0.pdf) Ross, J.S., et al. (2023). Twenty Questions and Answers About the Ozone Layer: 2022 Update, Scientific Assessment of Ozone Depletion: 2022, 75 pp., WMO, Geneva, Switzerland.\n\n[[10]](https://doi.org/10.1038/s41467-023-42637-0) Kessenich, H.E., et al. (2023). Potential drivers of the recent large Antarctic ozone holes. Nature Communications, 14, 7259.\n\n[[11]](https://doi.org/10.1029/2022JD037581) Petkov, B.H., et al. (2023. Un Unprecedented Arctic Ozone Depletion Evnet During Spring 2020 and Its Impacts Across Europe. Journal of Geophysical Research: Atmospheres, 128, e2022JD037581.\n\n[[12]](https://doi.org/10.1175/JAS-3338.1) Stolarski, R. S., McPeters, R. D., & Newman, P. A. (2005). The ozone hole of 2002 as measured by TOMS. Journal of the atmospheric sciences, 62(3), 716-720.\n\n[[13]](https://doi.org/10.1029/2010GL046634) Steinbrecht, W., et al. (2011). Very high ozone columns at northern mid-latitudes in 2010. Geophysical research Letters, 38, L06803.", "text_with_prefix": "EQC Quality Assessment: \"Temporal and spatial completeness of the satellite-derived Ozone for policy making and variability studies\"\nDataset: satellite-ozone-v1 [CDS]\nAspect: climate-monitoring_q01 | Category: Satellite_ECVs\nSection: Temporal and spatial completeness of the satellite-derived Ozone for policy making and variability studies > ℹ️ If you want to know more > References\n---\n[[1]](https://dast.copernicus-climate.eu/documents/satellite-ozone/C3S2_312a_Lot2_PUGS_O3_latest.pdf) Van Roozendael, M., et al. (2021a). Ozone Product User Guide and Specification (PUGS). Last access: March 27, 2024\n\n[[2]](https://doi.org/10.5194/amt-8-3021-2015) van der A, R.J., et al. (2015). Extended and refined multi sensor reanalysis of total ozone for the period 1970-2012. Atmospheric Measurement Techniques, 8, 3021-3035.\n\n[[3]](https://dast.copernicus-climate.eu/documents/satellite-ozone/C3S2_312a_Lot2_ATBD_O3_latest.pdf) Van Roozendael, M., et al. (2021b). Ozone Algorithm Theoretical Basis Document (ATBD). Last access: March 28, 2024\n\n[[4]](https://doi.org/10.1002/2013RG000448) Butchart, N. (2014). The Brewer-Dobson circulation. Review of Geophysics, 52(2), 157-184.\n\n[[5]](https://doi.org/10.1038/315207a0) Farman, J.C., et al. (1985). Large losses of total ozone in Antarctica reveal seasonal ClO$_x$/NO$_x$ interaction. Nature, 315, 207-210.\n\n[[6]](https://doi.org/10.1126/science.aae0061) Solomon, S., Ivy, D. J., Kinnison, D., Mills, M. J., Neely III, R. R., & Schmidt, A. (2016). Emergence of healing in the Antarctic ozone layer. Science, 353(6296), 269-274.\n\n[[7]](https://doi.org/10.5194/acp-11-9237-2011) Ziemke, J. R., Chandra, S., Labow, G. J., Bhartia, P. K., Froidevaux, L., & Witte, J. C. (2011). A global climatology of tropospheric and stratospheric ozone derived from Aura OMI and MLS measurements. Atmospheric Chemistry and Physics, 11(17), 9237-9251.\n\n[[8]](https://doi.org/10.1029/1999RG900008) Solomon, S. (1999). Stratospheric ozone depletion: A review of concepts and history. Review of Geophysics, 37(3), 275-316.\n\n[[9]](https://ozone.unep.org/sites/default/files/2023-05/Final_20Qs%202022%20full%20document_26April2023_digital%20version-reduced_0.pdf) Ross, J.S., et al. (2023). Twenty Questions and Answers About the Ozone Layer: 2022 Update, Scientific Assessment of Ozone Depletion: 2022, 75 pp., WMO, Geneva, Switzerland.\n\n[[10]](https://doi.org/10.1038/s41467-023-42637-0) Kessenich, H.E., et al. (2023). Potential drivers of the recent large Antarctic ozone holes. Nature Communications, 14, 7259.\n\n[[11]](https://doi.org/10.1029/2022JD037581) Petkov, B.H., et al. (2023. Un Unprecedented Arctic Ozone Depletion Evnet During Spring 2020 and Its Impacts Across Europe. Journal of Geophysical Research: Atmospheres, 128, e2022JD037581.\n\n[[12]](https://doi.org/10.1175/JAS-3338.1) Stolarski, R. S., McPeters, R. D., & Newman, P. A. (2005). The ozone hole of 2002 as measured by TOMS. Journal of the atmospheric sciences, 62(3), 716-720.\n\n[[13]](https://doi.org/10.1029/2010GL046634) Steinbrecht, W., et al. (2011). Very high ozone columns at northern mid-latitudes in 2010. Geophysical research Letters, 38, L06803."} {"chunk_id": "satellite_satellite-ozone-v1_climate-monitoring_q01__b0cbe90be47d", "report_id": "satellite_satellite-ozone-v1_climate-monitoring_q01", "dataset_id": "satellite-ozone-v1", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-monitoring_q01", "aspect_base": "climate-monitoring", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Temporal and spatial completeness of the satellite-derived Ozone for policy making and variability studies > ℹ️ If you want to know more > References", "title": "Temporal and spatial completeness of the satellite-derived Ozone for policy making and variability studies", "chunk_index": 12, "token_count": 259, "text_raw": "9/2010GL046634) Steinbrecht, W., et al. (2011). Very high ozone columns at northern mid-latitudes in 2010. Geophysical research Letters, 38, L06803.\n\n[[14]](https://doi.org/10.1029/2020GL087810) Safieddine, S., et al. (2020). Antarctic ozone enhancement during the 2019 sudden stratospheric warming event. Geophysical Research Letters,47,e2020GL087810.\n\n[[15]](https://doi.org/10.1029/2002JD002170) Logan, J.A., et al. (2003). Quasibiennial oscillation in tropical ozone as revealed by ozonesonde and satellite date. Journal of Geophysical Research, 108, 4244(D8).", "text_with_prefix": "EQC Quality Assessment: \"Temporal and spatial completeness of the satellite-derived Ozone for policy making and variability studies\"\nDataset: satellite-ozone-v1 [CDS]\nAspect: climate-monitoring_q01 | Category: Satellite_ECVs\nSection: Temporal and spatial completeness of the satellite-derived Ozone for policy making and variability studies > ℹ️ If you want to know more > References\n---\n9/2010GL046634) Steinbrecht, W., et al. (2011). Very high ozone columns at northern mid-latitudes in 2010. Geophysical research Letters, 38, L06803.\n\n[[14]](https://doi.org/10.1029/2020GL087810) Safieddine, S., et al. (2020). Antarctic ozone enhancement during the 2019 sudden stratospheric warming event. Geophysical Research Letters,47,e2020GL087810.\n\n[[15]](https://doi.org/10.1029/2002JD002170) Logan, J.A., et al. (2003). Quasibiennial oscillation in tropical ozone as revealed by ozonesonde and satellite date. Journal of Geophysical Research, 108, 4244(D8)."} {"chunk_id": "satellite_derived-gridded-glacier-mass-change_data-completeness_q01__4d4c4377f2cc", "report_id": "satellite_derived-gridded-glacier-mass-change_data-completeness_q01", "dataset_id": "derived-gridded-glacier-mass-change", "store": "CDS", "doc_type": "EQC_QA", "aspect": "data-completeness_q01", "aspect_base": "data-completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Glacier mass change data from satellite and in-situ observations: uncertainty analysis for glaciological and climate change monitoring", "title": "Glacier mass change data from satellite and in-situ observations: uncertainty analysis for glaciological and climate change monitoring", "chunk_index": 0, "token_count": 139, "text_raw": "Production date: 31-05-2025\n\nDataset version: WGMS-FOG-2023-09\n\nProduced by: Yoni Verhaegen and Philippe Huybrechts (Vrije Universiteit Brussel)", "text_with_prefix": "EQC Quality Assessment: \"Glacier mass change data from satellite and in-situ observations: uncertainty analysis for glaciological and climate change monitoring\"\nDataset: derived-gridded-glacier-mass-change [CDS]\nAspect: data-completeness_q01 | Category: Satellite_ECVs\nSection: Glacier mass change data from satellite and in-situ observations: uncertainty analysis for glaciological and climate change monitoring\n---\nProduction date: 31-05-2025\n\nDataset version: WGMS-FOG-2023-09\n\nProduced by: Yoni Verhaegen and Philippe Huybrechts (Vrije Universiteit Brussel)"} {"chunk_id": "satellite_derived-gridded-glacier-mass-change_data-completeness_q01__54eceb88a783", "report_id": "satellite_derived-gridded-glacier-mass-change_data-completeness_q01", "dataset_id": "derived-gridded-glacier-mass-change", "store": "CDS", "doc_type": "EQC_QA", "aspect": "data-completeness_q01", "aspect_base": "data-completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Glacier mass change data from satellite and in-situ observations: uncertainty analysis for glaciological and climate change monitoring > Quality assessment question", "title": "Glacier mass change data from satellite and in-situ observations: uncertainty analysis for glaciological and climate change monitoring", "chunk_index": 1, "token_count": 490, "text_raw": "* **\"Is the glacier mass change dataset sufficiently adequate in terms of its uncertainty to effectively monitor and evaluate global (cumulative) glacier mass changes over time, including their associated impacts on global sea level rise?\"**\n\nGlaciers significantly impact global sea-level rise, freshwater resources, natural hazards, hydro-power generation, recreation and tourism. Assessing glacier mass changes due to climate warming is therefore crucial for addressing these issues. The \"[Glacier mass change gridded data from 1976 to present derived from the Fluctuations of Glaciers Database](https://cds.climate.copernicus.eu/datasets/derived-gridded-glacier-mass-change?tab=overview)\" (version WGMS-FOG-2023-09) on the Climate Data Store (CDS), offers a global coverage of glacier mass changes by integrating in-situ, aerial, and satellite data [[1](https://wgms.ch/), [2](https://doi.org/10.5194/essd-17-1977-2025)]. The gridded glaciers mass change dataset that is on the CDS is currently one of the most complete dataset of glacier mass change data in terms of its spatial coverage. It is generally considered the main reference dataset to determine long-term glacier mass changes across the globe.\n\nWhen measured over a long period and at extended geographical scales, trends in glacier mass balance can be considered a clear indicator of global climate change [[3](https://doi.org/10.1038/s41586-019-1071-0)]. Despite some known issues, this dataset provides valuable insights into glacier mass changes across spatial and temporal scales. In that regard, this notebook investigates how well the dataset can be used to quantify the link between glacier melt and glacier-related sea level contributions. More specifically, the notebook evaluates whether the dataset is of sufficient maturity and quality for that purpose in terms of its uncertainty (i.e. accuracy and precision).", "text_with_prefix": "EQC Quality Assessment: \"Glacier mass change data from satellite and in-situ observations: uncertainty analysis for glaciological and climate change monitoring\"\nDataset: derived-gridded-glacier-mass-change [CDS]\nAspect: data-completeness_q01 | Category: Satellite_ECVs\nSection: Glacier mass change data from satellite and in-situ observations: uncertainty analysis for glaciological and climate change monitoring > Quality assessment question\n---\n* **\"Is the glacier mass change dataset sufficiently adequate in terms of its uncertainty to effectively monitor and evaluate global (cumulative) glacier mass changes over time, including their associated impacts on global sea level rise?\"**\n\nGlaciers significantly impact global sea-level rise, freshwater resources, natural hazards, hydro-power generation, recreation and tourism. Assessing glacier mass changes due to climate warming is therefore crucial for addressing these issues. The \"[Glacier mass change gridded data from 1976 to present derived from the Fluctuations of Glaciers Database](https://cds.climate.copernicus.eu/datasets/derived-gridded-glacier-mass-change?tab=overview)\" (version WGMS-FOG-2023-09) on the Climate Data Store (CDS), offers a global coverage of glacier mass changes by integrating in-situ, aerial, and satellite data [[1](https://wgms.ch/), [2](https://doi.org/10.5194/essd-17-1977-2025)]. The gridded glaciers mass change dataset that is on the CDS is currently one of the most complete dataset of glacier mass change data in terms of its spatial coverage. It is generally considered the main reference dataset to determine long-term glacier mass changes across the globe.\n\nWhen measured over a long period and at extended geographical scales, trends in glacier mass balance can be considered a clear indicator of global climate change [[3](https://doi.org/10.1038/s41586-019-1071-0)]. Despite some known issues, this dataset provides valuable insights into glacier mass changes across spatial and temporal scales. In that regard, this notebook investigates how well the dataset can be used to quantify the link between glacier melt and glacier-related sea level contributions. More specifically, the notebook evaluates whether the dataset is of sufficient maturity and quality for that purpose in terms of its uncertainty (i.e. accuracy and precision)."} {"chunk_id": "satellite_derived-gridded-glacier-mass-change_data-completeness_q01__33bcb7f87194", "report_id": "satellite_derived-gridded-glacier-mass-change_data-completeness_q01", "dataset_id": "derived-gridded-glacier-mass-change", "store": "CDS", "doc_type": "EQC_QA", "aspect": "data-completeness_q01", "aspect_base": "data-completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Glacier mass change data from satellite and in-situ observations: uncertainty analysis for glaciological and climate change monitoring > Quality assessment statements", "title": "Glacier mass change data from satellite and in-situ observations: uncertainty analysis for glaciological and climate change monitoring", "chunk_index": 2, "token_count": 444, "text_raw": "These are the key outcomes of this assessment\n\n- The C3S glacier mass change dataset is a mature and quality-rich dataset that allows users to capture the changing state of glaciers on a local, regional and global scale. The data come with a quantitative pixel-by-pixel error estimate in the form of precision errors (reported as 1.96 times the standard deviation), that generally decrease in magnitude over time. As a consequence, the most recent years (after 2000) meet proposed international standards of uncertainty thresholds proposed by GCOS. The uncertainty that comes with the data is, however, highly spatially and temporally variable, with areas like the peripheral glaciers in Greenland and Antarctica exhibiting higher glacier mass change errors. Hence, not all pixels meet the minimum uncertainty threshold proposed by GCOS, of which users should take note. However, when seen as a whole, the majority of the pixels meet the proposed thresholds for uncertainty.\n- The dataset is highly suitable to monitor and derive global (cumulative) mass changes over time and to, for example, assess the corresponding impact of their shrinkage. The dataset furthermore has sufficient quality in terms of its consistency with respect to spatial (i.e. global) and temporal (i.e. since 1975-76) coverage for climate change monitoring to become useful and to be able to draw reliable conclusions. Some potential key limitiations include the lack of detailed sampling information, the limited use for individual glacier-scale (smaller than the pixel-scale) applications, and the fact that not all mass changes considered in the dataset directly contribute to sea-level changes (which is nevertheless assumed to be the case). Due to variations in data collection methods and sources, data consistency along the data may furthermore be affected. \n```", "text_with_prefix": "EQC Quality Assessment: \"Glacier mass change data from satellite and in-situ observations: uncertainty analysis for glaciological and climate change monitoring\"\nDataset: derived-gridded-glacier-mass-change [CDS]\nAspect: data-completeness_q01 | Category: Satellite_ECVs\nSection: Glacier mass change data from satellite and in-situ observations: uncertainty analysis for glaciological and climate change monitoring > Quality assessment statements\n---\nThese are the key outcomes of this assessment\n\n- The C3S glacier mass change dataset is a mature and quality-rich dataset that allows users to capture the changing state of glaciers on a local, regional and global scale. The data come with a quantitative pixel-by-pixel error estimate in the form of precision errors (reported as 1.96 times the standard deviation), that generally decrease in magnitude over time. As a consequence, the most recent years (after 2000) meet proposed international standards of uncertainty thresholds proposed by GCOS. The uncertainty that comes with the data is, however, highly spatially and temporally variable, with areas like the peripheral glaciers in Greenland and Antarctica exhibiting higher glacier mass change errors. Hence, not all pixels meet the minimum uncertainty threshold proposed by GCOS, of which users should take note. However, when seen as a whole, the majority of the pixels meet the proposed thresholds for uncertainty.\n- The dataset is highly suitable to monitor and derive global (cumulative) mass changes over time and to, for example, assess the corresponding impact of their shrinkage. The dataset furthermore has sufficient quality in terms of its consistency with respect to spatial (i.e. global) and temporal (i.e. since 1975-76) coverage for climate change monitoring to become useful and to be able to draw reliable conclusions. Some potential key limitiations include the lack of detailed sampling information, the limited use for individual glacier-scale (smaller than the pixel-scale) applications, and the fact that not all mass changes considered in the dataset directly contribute to sea-level changes (which is nevertheless assumed to be the case). Due to variations in data collection methods and sources, data consistency along the data may furthermore be affected. \n```"} {"chunk_id": "satellite_derived-gridded-glacier-mass-change_data-completeness_q01__42c3919208a1", "report_id": "satellite_derived-gridded-glacier-mass-change_data-completeness_q01", "dataset_id": "derived-gridded-glacier-mass-change", "store": "CDS", "doc_type": "EQC_QA", "aspect": "data-completeness_q01", "aspect_base": "data-completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Glacier mass change data from satellite and in-situ observations: uncertainty analysis for glaciological and climate change monitoring > Methodology > Dataset description", "title": "Glacier mass change data from satellite and in-situ observations: uncertainty analysis for glaciological and climate change monitoring", "chunk_index": 3, "token_count": 874, "text_raw": "The mass balance of a glacier is the difference between mass gained (mainly from snow accumulation) and mass lost (by meltwater runoff or solid ice discharge into lakes/the ocean), which is the same as the net mass change of a glacier over its surface area $A$:\n\n$\\Delta{M} = \\dfrac{1}{A} \\int\\limits_{A} b_a dA - \\dfrac{1}{A} \\int\\limits_{P} d_a dP$\n\nwhere:\n\n$b_a = \\int\\limits^{t_1}_{t_0}b_s dt$, with $b_s = ACC - ABL$ [kg m$^{-2}$ yr$^{-1}$]\n\n$d_a = \\int\\limits^{t_1}_{t_0}d_g dt$, with $d_g = \\rho_{i} \\cdot H \\cdot \\overline{V_P}$ [kg m$^{-1}$ yr$^{-1}$]\n\nThe first term represents the the total annual surface mass balance (i.e. the difference between accumulation and ablation), whereas the second term encompasses the total annual ice discharge across the grounding line $g$ with perimeter $P$. Here, $\\rho_{i}$ is the ice density, $H$ ice thickness and $\\overline{V_P}$ the gate-perpendicular vertically averaged horizontal velocity at the grounding line location. The latter term is considered zero for land-terminating glaciers. Other terms can be included as well, such as the basal and internal mass balances.\n\nIn general, the basis for the derived gridded mass changes are individual measurements (mainly glaciological in-situ local annual surface mass balance measurements) and geodetic air or spaceborne elevation change data (a surface elevation/ice thickness change or an ice volume change over time). These data are converted into an averaged specific mass balance value (i.e. mostly reported with units of meter water equivalent and often shortened to m w.e.) for an individual glacier. Afterwards, the data are submitted to the World Glacier Monitoring Service (WGMS). Further processing of the data results in a glacier mass change product reported over a 0.5° global grid dating back until the 1975-76 hydrological year. Each grid cell therefore contains a time series of total glacier mass change (in Gt yr⁻¹) or mass balance data (in m w.e. yr⁻¹) of all glaciers within the specific grid cell. In this notebook, we use version WGMS-FOG-2023-09. For a more detailed description of the data acquisition and processing methods, we refer to the [documentation on the CDS](https://cds.climate.copernicus.eu/datasets/derived-gridded-glacier-mass-change?tab=overview) and the [ECMWF Confluence Wiki](https://confluence.ecmwf.int/pages/viewpage.action?pageId=355349383) (Copernicus Knowledge Base).\n\nIt is important to note that the glaciers with an annual glaciological sample (ca. 500 glaciers) form the basis for the determination of annual mass changes from the remaining glaciers through a complex algorithm of spatial/temporal interpolation and (area-weighted) averaging of these mass change data [[2](https://essd.copernicus.org/preprints/essd-2024-323/)]. Hence, not all glaciers in the dataset exhibit a time series of directly measured annual in-situ mass balance observations, but most glaciers in the dataset are either unobserved or only have (limited and mostly multi-annual) geodetic mass change data available. Glacier mass change data with these geodetic samples are also more prone to uncertainties in general, for example due to uncertainties related to volume to mass conversions.", "text_with_prefix": "EQC Quality Assessment: \"Glacier mass change data from satellite and in-situ observations: uncertainty analysis for glaciological and climate change monitoring\"\nDataset: derived-gridded-glacier-mass-change [CDS]\nAspect: data-completeness_q01 | Category: Satellite_ECVs\nSection: Glacier mass change data from satellite and in-situ observations: uncertainty analysis for glaciological and climate change monitoring > Methodology > Dataset description\n---\nThe mass balance of a glacier is the difference between mass gained (mainly from snow accumulation) and mass lost (by meltwater runoff or solid ice discharge into lakes/the ocean), which is the same as the net mass change of a glacier over its surface area $A$:\n\n$\\Delta{M} = \\dfrac{1}{A} \\int\\limits_{A} b_a dA - \\dfrac{1}{A} \\int\\limits_{P} d_a dP$\n\nwhere:\n\n$b_a = \\int\\limits^{t_1}_{t_0}b_s dt$, with $b_s = ACC - ABL$ [kg m$^{-2}$ yr$^{-1}$]\n\n$d_a = \\int\\limits^{t_1}_{t_0}d_g dt$, with $d_g = \\rho_{i} \\cdot H \\cdot \\overline{V_P}$ [kg m$^{-1}$ yr$^{-1}$]\n\nThe first term represents the the total annual surface mass balance (i.e. the difference between accumulation and ablation), whereas the second term encompasses the total annual ice discharge across the grounding line $g$ with perimeter $P$. Here, $\\rho_{i}$ is the ice density, $H$ ice thickness and $\\overline{V_P}$ the gate-perpendicular vertically averaged horizontal velocity at the grounding line location. The latter term is considered zero for land-terminating glaciers. Other terms can be included as well, such as the basal and internal mass balances.\n\nIn general, the basis for the derived gridded mass changes are individual measurements (mainly glaciological in-situ local annual surface mass balance measurements) and geodetic air or spaceborne elevation change data (a surface elevation/ice thickness change or an ice volume change over time). These data are converted into an averaged specific mass balance value (i.e. mostly reported with units of meter water equivalent and often shortened to m w.e.) for an individual glacier. Afterwards, the data are submitted to the World Glacier Monitoring Service (WGMS). Further processing of the data results in a glacier mass change product reported over a 0.5° global grid dating back until the 1975-76 hydrological year. Each grid cell therefore contains a time series of total glacier mass change (in Gt yr⁻¹) or mass balance data (in m w.e. yr⁻¹) of all glaciers within the specific grid cell. In this notebook, we use version WGMS-FOG-2023-09. For a more detailed description of the data acquisition and processing methods, we refer to the [documentation on the CDS](https://cds.climate.copernicus.eu/datasets/derived-gridded-glacier-mass-change?tab=overview) and the [ECMWF Confluence Wiki](https://confluence.ecmwf.int/pages/viewpage.action?pageId=355349383) (Copernicus Knowledge Base).\n\nIt is important to note that the glaciers with an annual glaciological sample (ca. 500 glaciers) form the basis for the determination of annual mass changes from the remaining glaciers through a complex algorithm of spatial/temporal interpolation and (area-weighted) averaging of these mass change data [[2](https://essd.copernicus.org/preprints/essd-2024-323/)]. Hence, not all glaciers in the dataset exhibit a time series of directly measured annual in-situ mass balance observations, but most glaciers in the dataset are either unobserved or only have (limited and mostly multi-annual) geodetic mass change data available. Glacier mass change data with these geodetic samples are also more prone to uncertainties in general, for example due to uncertainties related to volume to mass conversions."} {"chunk_id": "satellite_derived-gridded-glacier-mass-change_data-completeness_q01__defb3ba9e7ba", "report_id": "satellite_derived-gridded-glacier-mass-change_data-completeness_q01", "dataset_id": "derived-gridded-glacier-mass-change", "store": "CDS", "doc_type": "EQC_QA", "aspect": "data-completeness_q01", "aspect_base": "data-completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Glacier mass change data from satellite and in-situ observations: uncertainty analysis for glaciological and climate change monitoring > Methodology > Structure and (sub)sections", "title": "Glacier mass change data from satellite and in-situ observations: uncertainty analysis for glaciological and climate change monitoring", "chunk_index": 4, "token_count": 221, "text_raw": "**[](section-1)**\n\n* [](section-1-1)\n* [](section-1-2)\n* [](section-1-3)\n\n**[](section-2)**\n* [](section-2-1)\n* [](section-2-2)\n* [](section-2-3)\n\n**[](section-3)**\n* [](section-3-1)\n* [](section-3-2)\n* [](section-3-3)\n\n**[](section-4)**\n* [](section-4-1)\n* [](section-4-2)\n\n**[](section-5)**", "text_with_prefix": "EQC Quality Assessment: \"Glacier mass change data from satellite and in-situ observations: uncertainty analysis for glaciological and climate change monitoring\"\nDataset: derived-gridded-glacier-mass-change [CDS]\nAspect: data-completeness_q01 | Category: Satellite_ECVs\nSection: Glacier mass change data from satellite and in-situ observations: uncertainty analysis for glaciological and climate change monitoring > Methodology > Structure and (sub)sections\n---\n**[](section-1)**\n\n* [](section-1-1)\n* [](section-1-2)\n* [](section-1-3)\n\n**[](section-2)**\n* [](section-2-1)\n* [](section-2-2)\n* [](section-2-3)\n\n**[](section-3)**\n* [](section-3-1)\n* [](section-3-2)\n* [](section-3-3)\n\n**[](section-4)**\n* [](section-4-1)\n* [](section-4-2)\n\n**[](section-5)**"} {"chunk_id": "satellite_derived-gridded-glacier-mass-change_data-completeness_q01__9b120dbed624", "report_id": "satellite_derived-gridded-glacier-mass-change_data-completeness_q01", "dataset_id": "derived-gridded-glacier-mass-change", "store": "CDS", "doc_type": "EQC_QA", "aspect": "data-completeness_q01", "aspect_base": "data-completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Glacier mass change data from satellite and in-situ observations: uncertainty analysis for glaciological and climate change monitoring > Analysis and results > 1. Data preparation and processing > 1.2 Define request and download", "title": "Glacier mass change data from satellite and in-situ observations: uncertainty analysis for glaciological and climate change monitoring", "chunk_index": 5, "token_count": 248, "text_raw": "Then we define the parameters, i.e. for which years we want the glacier mass change data to be downloaded:\n\nThis line validates that both period_start and period_stop meet specific criteria:\nThey contain an underscore (\"_\") and their length is exactly 9 characters.\n\nThen we define requests for download from the CDS and download and transform the glacier mass change data.\n\nDownload glacier mass change data\n\n```text\nDownloading glacier mass change data...\n```\n\n```text\n100%|██████████| 1/1 [00:06<00:00, 6.66s/it]\n```\n\n```text\nDownloading done.\n```\n\n(section-1-3)=", "text_with_prefix": "EQC Quality Assessment: \"Glacier mass change data from satellite and in-situ observations: uncertainty analysis for glaciological and climate change monitoring\"\nDataset: derived-gridded-glacier-mass-change [CDS]\nAspect: data-completeness_q01 | Category: Satellite_ECVs\nSection: Glacier mass change data from satellite and in-situ observations: uncertainty analysis for glaciological and climate change monitoring > Analysis and results > 1. Data preparation and processing > 1.2 Define request and download\n---\nThen we define the parameters, i.e. for which years we want the glacier mass change data to be downloaded:\n\nThis line validates that both period_start and period_stop meet specific criteria:\nThey contain an underscore (\"_\") and their length is exactly 9 characters.\n\nThen we define requests for download from the CDS and download and transform the glacier mass change data.\n\nDownload glacier mass change data\n\n```text\nDownloading glacier mass change data...\n```\n\n```text\n100%|██████████| 1/1 [00:06<00:00, 6.66s/it]\n```\n\n```text\nDownloading done.\n```\n\n(section-1-3)="} {"chunk_id": "satellite_derived-gridded-glacier-mass-change_data-completeness_q01__6086b8f98267", "report_id": "satellite_derived-gridded-glacier-mass-change_data-completeness_q01", "dataset_id": "derived-gridded-glacier-mass-change", "store": "CDS", "doc_type": "EQC_QA", "aspect": "data-completeness_q01", "aspect_base": "data-completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Glacier mass change data from satellite and in-situ observations: uncertainty analysis for glaciological and climate change monitoring > Analysis and results > 1. Data preparation and processing > 1.3 Display and inspect data structure", "title": "Glacier mass change data from satellite and in-situ observations: uncertainty analysis for glaciological and climate change monitoring", "chunk_index": 6, "token_count": 1022, "text_raw": "Lastly, we can read and inspect the glacier mass change data. Let us print out the data to inspect its structure:\n\n```text\n Size: 487MB\nDimensions: (time: 47, latitude: 360, longitude: 720)\nCoordinates:\n * time (time) datetime64[ns] 376B 1976-01-01 ... 2022-0...\n * latitude (latitude) float64 3kB 89.75 89.25 ... -89.75\n * longitude (longitude) float64 6kB -179.8 -179.2 ... 179.8\nData variables:\n glacier_mass_change_gt (time, latitude, longitude) float64 97MB dask.array\n glacier_mass_change_mwe (time, latitude, longitude) float64 97MB dask.array\n glacier_area_km2 (time, latitude, longitude) float64 97MB dask.array\n uncertainty_gt (time, latitude, longitude) float64 97MB dask.array\n uncertainty_mwe (time, latitude, longitude) float64 97MB dask.array\nAttributes:\n title: Global gridded annual glacier mass changes\n data_version: version-wgms-fog-2023-09\n project: Copernicus Climate Change Service (C3S) Essential ...\n institution: World Glacier Monitoring Service - Geography Depar...\n created_by: Dr. Ines Dussaillant - ines.dussaillant@geo.uzh.ch\n references: Fluctuation of Glaciers (FoG) database version wgm...\n citation: Dussaillant, I., Bannwart, J., Paul, F., Zemp, M. ...\n conventions: CF Version CF-1.8\n dataset_description: Horizontal resolution: 0.5° (latitude - longitude)...\n dataset_limitations: Grid-point artefact in polar regions: see Algorith...\n dataset_improvements: Improvements of product version WGMS-FOG-2023-09 w...\n comments: Conversions between annual grid point mass change ...\n```\n\nIt is a gridded dataset at a 0.5 by 0.5 degree spatial resolution containing annual values of the total glacier mass change (in Gt yr$^{-1}$) of a grid cell (`glacier_mass_change_gt`) and its uncertainty (`uncertainty_gt`) since the 1975-76 hydrological year. Mass changes and their uncertainty can also be extracted in units of m w.e. yr$^{-1}$ (`glacier_mass_change_mwe` and `uncertainty_mwe`). The parameterized absolute glacier surface area is furthermore also available in this dataset by the variable `glacier_area_km2`. For conversion purposes, glacier mass changes (in Gt yr$^{-1}$) can be converted into glacier mass balances (in m w.e. yr$^{-1}$) by dividing the mass changes (in Gt yr$^{-1}$ multiplied by $1*10^{12}$ to get values in kg yr$^{-1}$) by the product of the density of water (1000 kg m$^{-3}$) and the glacier area (in km$^{2}$ multiplied by $1*10^6$ to get values in m$^2$).\n\nLet us perform some data handling before getting started with the analysis:\n\nCustomize attributes in glacier mass change file\nAdd glacier RGI region based on slope of glacier area change parameterization\nGroup grid points by RGI region (assuming 'rgi_region' is a coordinate in the dataset)\nCompute regional sums assuming spatial correlation within each region\nCompute regional uncertainties assuming full spatial correlation within each region\nCompute global sums assuming uncorrelated regional estimates\nCompute global uncertainty assuming uncorrelated regions\nSum over time\nAdd results to dataset\n\n(section-2)=", "text_with_prefix": "EQC Quality Assessment: \"Glacier mass change data from satellite and in-situ observations: uncertainty analysis for glaciological and climate change monitoring\"\nDataset: derived-gridded-glacier-mass-change [CDS]\nAspect: data-completeness_q01 | Category: Satellite_ECVs\nSection: Glacier mass change data from satellite and in-situ observations: uncertainty analysis for glaciological and climate change monitoring > Analysis and results > 1. Data preparation and processing > 1.3 Display and inspect data structure\n---\nLastly, we can read and inspect the glacier mass change data. Let us print out the data to inspect its structure:\n\n```text\n Size: 487MB\nDimensions: (time: 47, latitude: 360, longitude: 720)\nCoordinates:\n * time (time) datetime64[ns] 376B 1976-01-01 ... 2022-0...\n * latitude (latitude) float64 3kB 89.75 89.25 ... -89.75\n * longitude (longitude) float64 6kB -179.8 -179.2 ... 179.8\nData variables:\n glacier_mass_change_gt (time, latitude, longitude) float64 97MB dask.array\n glacier_mass_change_mwe (time, latitude, longitude) float64 97MB dask.array\n glacier_area_km2 (time, latitude, longitude) float64 97MB dask.array\n uncertainty_gt (time, latitude, longitude) float64 97MB dask.array\n uncertainty_mwe (time, latitude, longitude) float64 97MB dask.array\nAttributes:\n title: Global gridded annual glacier mass changes\n data_version: version-wgms-fog-2023-09\n project: Copernicus Climate Change Service (C3S) Essential ...\n institution: World Glacier Monitoring Service - Geography Depar...\n created_by: Dr. Ines Dussaillant - ines.dussaillant@geo.uzh.ch\n references: Fluctuation of Glaciers (FoG) database version wgm...\n citation: Dussaillant, I., Bannwart, J., Paul, F., Zemp, M. ...\n conventions: CF Version CF-1.8\n dataset_description: Horizontal resolution: 0.5° (latitude - longitude)...\n dataset_limitations: Grid-point artefact in polar regions: see Algorith...\n dataset_improvements: Improvements of product version WGMS-FOG-2023-09 w...\n comments: Conversions between annual grid point mass change ...\n```\n\nIt is a gridded dataset at a 0.5 by 0.5 degree spatial resolution containing annual values of the total glacier mass change (in Gt yr$^{-1}$) of a grid cell (`glacier_mass_change_gt`) and its uncertainty (`uncertainty_gt`) since the 1975-76 hydrological year. Mass changes and their uncertainty can also be extracted in units of m w.e. yr$^{-1}$ (`glacier_mass_change_mwe` and `uncertainty_mwe`). The parameterized absolute glacier surface area is furthermore also available in this dataset by the variable `glacier_area_km2`. For conversion purposes, glacier mass changes (in Gt yr$^{-1}$) can be converted into glacier mass balances (in m w.e. yr$^{-1}$) by dividing the mass changes (in Gt yr$^{-1}$ multiplied by $1*10^{12}$ to get values in kg yr$^{-1}$) by the product of the density of water (1000 kg m$^{-3}$) and the glacier area (in km$^{2}$ multiplied by $1*10^6$ to get values in m$^2$).\n\nLet us perform some data handling before getting started with the analysis:\n\nCustomize attributes in glacier mass change file\nAdd glacier RGI region based on slope of glacier area change parameterization\nGroup grid points by RGI region (assuming 'rgi_region' is a coordinate in the dataset)\nCompute regional sums assuming spatial correlation within each region\nCompute regional uncertainties assuming full spatial correlation within each region\nCompute global sums assuming uncorrelated regional estimates\nCompute global uncertainty assuming uncorrelated regions\nSum over time\nAdd results to dataset\n\n(section-2)="} {"chunk_id": "satellite_derived-gridded-glacier-mass-change_data-completeness_q01__8d4f98bbca2e", "report_id": "satellite_derived-gridded-glacier-mass-change_data-completeness_q01", "dataset_id": "derived-gridded-glacier-mass-change", "store": "CDS", "doc_type": "EQC_QA", "aspect": "data-completeness_q01", "aspect_base": "data-completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Glacier mass change data from satellite and in-situ observations: uncertainty analysis for glaciological and climate change monitoring > Analysis and results > 2. Glacier mass changes and their uncertainty estimates in space and time > 2.1 Glacier mass change uncertainties: accuracy and precision", "title": "Glacier mass change data from satellite and in-situ observations: uncertainty analysis for glaciological and climate change monitoring", "chunk_index": 7, "token_count": 722, "text_raw": "The total error of an annual glacier mass change estimate is theoretically given by the sum of the precision (random) and the accuracy (systematic) error:\n\n$\n\\varepsilon = a\\sigma + \\delta\n$\nwhere $\\sigma$ is the random error (i.e. standard deviation), $a$ the critical z-score related to a certain confidence interval, and $\\delta$ the systematic error.\n\nIn the glacier mass change dataset, precision errors are reported as 1.96 times the standard deviation and the accuracy error is not considered [[2](https://doi.org/10.5194/essd-17-1977-2025)]. Therefore, in our case, $a$ = 1.96 and $\\delta$ is 0. In the following section below, we will thus consider the uncertainty of the dataset as being $1.96 \\sigma$.\n\nLet us now explore a bit more the uncertainty of the data. The pixel-by-pixel quantitative uncertainties that accompany the data arise from different sources, of which the most important ones are the input data measurement uncertainties, the density conversion for geodetic mass balances, and errors related to absolute glacier areas and their changes over time. All these processes and parameters thus play a key role in determining the final estimate of the glacier mass changes. The errors from these various sources are combined in the final product and the authors report these uncertainty values as being precision errors in the form of $1.96 \\sigma$ (i.e. equivalent to a 95% confidence interval for a normal distribution) with units in Gt yr⁻¹ or m w.e. yr⁻¹.\n\nIn the following, we will, however, change the units of the error to kg m⁻² yr⁻¹ because GCOS advises glacier mass change uncertainty values to be provided in these units [[4](https://library.wmo.int/idurl/4/58111)]:\n\n$\n\\varepsilon_{\\Delta{M}}\n$\n[kg m⁻² yr⁻¹]\n$\n= 1 \\cdot 10^{6} * \\left(\\dfrac{1.96 \\sigma}{A}\\right)\n$\n\nwhere $1.96 \\sigma$ is the pixel mass change uncertainty [Gt yr⁻¹] from `uncertainty_gt` and $A$ the pixel glacier surface area [km²] from `glacier_area_km2`.\n\nWe can plot a histogram of the error term over all pixels and all years to inspect its distribution:\n\nPlot the histogram\n\n*Figure 1. Histogram of pixel-by-pixel glacier mass change errors in the glacier mass change dataset.*\n\nLet us plot some statistics:\n\n```text\nThe overall arithmetic mean precision error over all pixels and all years is 530.69 kg m⁻² yr⁻¹ with a maximum of 5083.96 kg m⁻² yr⁻¹.\n```\n\n(section-2-2)=", "text_with_prefix": "EQC Quality Assessment: \"Glacier mass change data from satellite and in-situ observations: uncertainty analysis for glaciological and climate change monitoring\"\nDataset: derived-gridded-glacier-mass-change [CDS]\nAspect: data-completeness_q01 | Category: Satellite_ECVs\nSection: Glacier mass change data from satellite and in-situ observations: uncertainty analysis for glaciological and climate change monitoring > Analysis and results > 2. Glacier mass changes and their uncertainty estimates in space and time > 2.1 Glacier mass change uncertainties: accuracy and precision\n---\nThe total error of an annual glacier mass change estimate is theoretically given by the sum of the precision (random) and the accuracy (systematic) error:\n\n$\n\\varepsilon = a\\sigma + \\delta\n$\nwhere $\\sigma$ is the random error (i.e. standard deviation), $a$ the critical z-score related to a certain confidence interval, and $\\delta$ the systematic error.\n\nIn the glacier mass change dataset, precision errors are reported as 1.96 times the standard deviation and the accuracy error is not considered [[2](https://doi.org/10.5194/essd-17-1977-2025)]. Therefore, in our case, $a$ = 1.96 and $\\delta$ is 0. In the following section below, we will thus consider the uncertainty of the dataset as being $1.96 \\sigma$.\n\nLet us now explore a bit more the uncertainty of the data. The pixel-by-pixel quantitative uncertainties that accompany the data arise from different sources, of which the most important ones are the input data measurement uncertainties, the density conversion for geodetic mass balances, and errors related to absolute glacier areas and their changes over time. All these processes and parameters thus play a key role in determining the final estimate of the glacier mass changes. The errors from these various sources are combined in the final product and the authors report these uncertainty values as being precision errors in the form of $1.96 \\sigma$ (i.e. equivalent to a 95% confidence interval for a normal distribution) with units in Gt yr⁻¹ or m w.e. yr⁻¹.\n\nIn the following, we will, however, change the units of the error to kg m⁻² yr⁻¹ because GCOS advises glacier mass change uncertainty values to be provided in these units [[4](https://library.wmo.int/idurl/4/58111)]:\n\n$\n\\varepsilon_{\\Delta{M}}\n$\n[kg m⁻² yr⁻¹]\n$\n= 1 \\cdot 10^{6} * \\left(\\dfrac{1.96 \\sigma}{A}\\right)\n$\n\nwhere $1.96 \\sigma$ is the pixel mass change uncertainty [Gt yr⁻¹] from `uncertainty_gt` and $A$ the pixel glacier surface area [km²] from `glacier_area_km2`.\n\nWe can plot a histogram of the error term over all pixels and all years to inspect its distribution:\n\nPlot the histogram\n\n*Figure 1. Histogram of pixel-by-pixel glacier mass change errors in the glacier mass change dataset.*\n\nLet us plot some statistics:\n\n```text\nThe overall arithmetic mean precision error over all pixels and all years is 530.69 kg m⁻² yr⁻¹ with a maximum of 5083.96 kg m⁻² yr⁻¹.\n```\n\n(section-2-2)="} {"chunk_id": "satellite_derived-gridded-glacier-mass-change_data-completeness_q01__c032f8011652", "report_id": "satellite_derived-gridded-glacier-mass-change_data-completeness_q01", "dataset_id": "derived-gridded-glacier-mass-change", "store": "CDS", "doc_type": "EQC_QA", "aspect": "data-completeness_q01", "aspect_base": "data-completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Glacier mass change data from satellite and in-situ observations: uncertainty analysis for glaciological and climate change monitoring > Analysis and results > 2. Glacier mass changes and their uncertainty estimates in space and time > 2.2 Glacier mass change errors over time", "title": "Glacier mass change data from satellite and in-situ observations: uncertainty analysis for glaciological and climate change monitoring", "chunk_index": 8, "token_count": 531, "text_raw": "The overall arithmetic mean error over both space and time exhibits a moderate magnitude, and the majority of errors seem to be situated close to 200-500 kg m⁻² yr⁻¹. The threshold (i.e. the minimum requirement to be met to ensure that data are useful) for glacier mass change uncertainty (expressed in terms of 2$\\sigma$) proposed by the GCOS is 500 kg m⁻² yr⁻¹, while the \"breakthrough\" value (i.e. the level at which specified uses within climate monitoring become possible) would be 200 kg m⁻² yr⁻¹ per grid point [[4](https://library.wmo.int/idurl/4/58111)].\n\nAs an extra step, we can calculate the overall arithmetic mean glacier mass change precision error for each hydrological year to get a general idea of the overall magnitude of the errors. This is simply calculated as the spatial arithmetic mean of all pixels for each year $t$:\n\n$\\overline{\\varepsilon_{\\Delta{M}}}_{x,y} \n$\n[kg m⁻² yr⁻¹]\n$ = \\dfrac{1}{n} \\sum\\limits^{{{x,y}}} (\\varepsilon_{\\Delta{M_{x,y}}})$ with $x,y$ the spatial domain size and $n$ is the total amount of latitude/longitude pixels (NaN pixels with no glaciers are excluded).\n\nThis results in the following:\n\nCalculate area-weighted mean error for each year\nFlatten the arrays to perform the weighted mean calculation\nCalculate the area-weighted mean error\nConvert to a numpy array or xarray DataArray if needed\n\n*Figure 2. Time series of annual spatially averaged (arithmetic) glacier mass change errors in the glacier mass change dataset. The red dotted line represents the minimum uncertainty threshold proposed by GCOS [[4](https://library.wmo.int/idurl/4/58111)].*\n\n(section-2-3)=", "text_with_prefix": "EQC Quality Assessment: \"Glacier mass change data from satellite and in-situ observations: uncertainty analysis for glaciological and climate change monitoring\"\nDataset: derived-gridded-glacier-mass-change [CDS]\nAspect: data-completeness_q01 | Category: Satellite_ECVs\nSection: Glacier mass change data from satellite and in-situ observations: uncertainty analysis for glaciological and climate change monitoring > Analysis and results > 2. Glacier mass changes and their uncertainty estimates in space and time > 2.2 Glacier mass change errors over time\n---\nThe overall arithmetic mean error over both space and time exhibits a moderate magnitude, and the majority of errors seem to be situated close to 200-500 kg m⁻² yr⁻¹. The threshold (i.e. the minimum requirement to be met to ensure that data are useful) for glacier mass change uncertainty (expressed in terms of 2$\\sigma$) proposed by the GCOS is 500 kg m⁻² yr⁻¹, while the \"breakthrough\" value (i.e. the level at which specified uses within climate monitoring become possible) would be 200 kg m⁻² yr⁻¹ per grid point [[4](https://library.wmo.int/idurl/4/58111)].\n\nAs an extra step, we can calculate the overall arithmetic mean glacier mass change precision error for each hydrological year to get a general idea of the overall magnitude of the errors. This is simply calculated as the spatial arithmetic mean of all pixels for each year $t$:\n\n$\\overline{\\varepsilon_{\\Delta{M}}}_{x,y} \n$\n[kg m⁻² yr⁻¹]\n$ = \\dfrac{1}{n} \\sum\\limits^{{{x,y}}} (\\varepsilon_{\\Delta{M_{x,y}}})$ with $x,y$ the spatial domain size and $n$ is the total amount of latitude/longitude pixels (NaN pixels with no glaciers are excluded).\n\nThis results in the following:\n\nCalculate area-weighted mean error for each year\nFlatten the arrays to perform the weighted mean calculation\nCalculate the area-weighted mean error\nConvert to a numpy array or xarray DataArray if needed\n\n*Figure 2. Time series of annual spatially averaged (arithmetic) glacier mass change errors in the glacier mass change dataset. The red dotted line represents the minimum uncertainty threshold proposed by GCOS [[4](https://library.wmo.int/idurl/4/58111)].*\n\n(section-2-3)="} {"chunk_id": "satellite_derived-gridded-glacier-mass-change_data-completeness_q01__6f76fc63a2cc", "report_id": "satellite_derived-gridded-glacier-mass-change_data-completeness_q01", "dataset_id": "derived-gridded-glacier-mass-change", "store": "CDS", "doc_type": "EQC_QA", "aspect": "data-completeness_q01", "aspect_base": "data-completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Glacier mass change data from satellite and in-situ observations: uncertainty analysis for glaciological and climate change monitoring > Analysis and results > 2. Glacier mass changes and their uncertainty estimates in space and time > 2.3 Glacier mass change errors in space", "title": "Glacier mass change data from satellite and in-situ observations: uncertainty analysis for glaciological and climate change monitoring", "chunk_index": 9, "token_count": 704, "text_raw": "Next, let us check the spatial distribution of the mean overall error over time per grid point, which is simply calculated as the temporal arithmetic mean over all years for each pixel $x,y$:\n\n$\\overline{\\varepsilon_{\\Delta{M_{t}}}} \n$\n[kg m⁻² yr⁻¹]\n$ = \\dfrac{1}{n} \\sum\\limits_{i={1976}}^{{{1976+n-1}}} (\\varepsilon_{\\Delta{M_i}})$ with $n$ is the total amount of years in the time series.\n\nThis results in the following:\n\nPlot the gridded data\n\n*
Figure 3. Spatial distribution of temporally averaged glacier mass change errors for each pixel in the glacier mass change dataset.
*\n\nLet us make this plot a bit clearer by grouping these errors with respect to the threshold values proposed by GCOS [[4](https://library.wmo.int/idurl/4/58111)]:\n\nDefine the boundaries for the colorbar\nPlot the data\n\n*
Figure 4. Spatial distribution of temporally averaged glacier mass change errors for each pixel in the glacier mass change dataset, classified by proposed GCOS thresholds [[4](https://library.wmo.int/idurl/4/58111)].
*\n\nGreen pixels indicate areas where the breakthrough value of (an uncertainty of 200 kg m⁻² yr⁻¹ or lower), as proposed by GCOS, has been reached, whereas red grid points exhibit temporally arithmetic mean error values that exceed the threshold value (500 kg m⁻² yr⁻¹ or more) for the data to be useful. This confirms our statement from above: especially the Greenland and Antarctic peripheral glaciers and parts of the Andes exhibit error values that are clearly too high. Extra care from the user is required in these regions since the threshold value has not been reached. Other regions, such as High Mountain Asia and the majority of Alaskan and European glaciers, have precision error values that allow for a reliable and high-quality data analysis. Possible reasons for this pattern are sparse data coverage, complex terrain, and gridded data limitations.\n\nLet us quantify the percentage of data (over all pixels and all years) that do and do not reach the threshold values:\n\nQuantify the percentage of data that do and do not reach the threshold values\n\n```text\nThe percentage of data points with a glacier mass change error value less than 200 kg m⁻² yr⁻¹ is 21.79%.\nThe percentage of data points with a glacier mass change error value more than 500 kg m⁻² yr⁻¹ is 38.36%.\n```\n\nNow, let us plot the global annual mass change data over time and propagate its error throughout the time series.\n\n(section-3)=", "text_with_prefix": "EQC Quality Assessment: \"Glacier mass change data from satellite and in-situ observations: uncertainty analysis for glaciological and climate change monitoring\"\nDataset: derived-gridded-glacier-mass-change [CDS]\nAspect: data-completeness_q01 | Category: Satellite_ECVs\nSection: Glacier mass change data from satellite and in-situ observations: uncertainty analysis for glaciological and climate change monitoring > Analysis and results > 2. Glacier mass changes and their uncertainty estimates in space and time > 2.3 Glacier mass change errors in space\n---\nNext, let us check the spatial distribution of the mean overall error over time per grid point, which is simply calculated as the temporal arithmetic mean over all years for each pixel $x,y$:\n\n$\\overline{\\varepsilon_{\\Delta{M_{t}}}} \n$\n[kg m⁻² yr⁻¹]\n$ = \\dfrac{1}{n} \\sum\\limits_{i={1976}}^{{{1976+n-1}}} (\\varepsilon_{\\Delta{M_i}})$ with $n$ is the total amount of years in the time series.\n\nThis results in the following:\n\nPlot the gridded data\n\n*
Figure 3. Spatial distribution of temporally averaged glacier mass change errors for each pixel in the glacier mass change dataset.
*\n\nLet us make this plot a bit clearer by grouping these errors with respect to the threshold values proposed by GCOS [[4](https://library.wmo.int/idurl/4/58111)]:\n\nDefine the boundaries for the colorbar\nPlot the data\n\n*
Figure 4. Spatial distribution of temporally averaged glacier mass change errors for each pixel in the glacier mass change dataset, classified by proposed GCOS thresholds [[4](https://library.wmo.int/idurl/4/58111)].
*\n\nGreen pixels indicate areas where the breakthrough value of (an uncertainty of 200 kg m⁻² yr⁻¹ or lower), as proposed by GCOS, has been reached, whereas red grid points exhibit temporally arithmetic mean error values that exceed the threshold value (500 kg m⁻² yr⁻¹ or more) for the data to be useful. This confirms our statement from above: especially the Greenland and Antarctic peripheral glaciers and parts of the Andes exhibit error values that are clearly too high. Extra care from the user is required in these regions since the threshold value has not been reached. Other regions, such as High Mountain Asia and the majority of Alaskan and European glaciers, have precision error values that allow for a reliable and high-quality data analysis. Possible reasons for this pattern are sparse data coverage, complex terrain, and gridded data limitations.\n\nLet us quantify the percentage of data (over all pixels and all years) that do and do not reach the threshold values:\n\nQuantify the percentage of data that do and do not reach the threshold values\n\n```text\nThe percentage of data points with a glacier mass change error value less than 200 kg m⁻² yr⁻¹ is 21.79%.\nThe percentage of data points with a glacier mass change error value more than 500 kg m⁻² yr⁻¹ is 38.36%.\n```\n\nNow, let us plot the global annual mass change data over time and propagate its error throughout the time series.\n\n(section-3)="} {"chunk_id": "satellite_derived-gridded-glacier-mass-change_data-completeness_q01__2a012915ed9e", "report_id": "satellite_derived-gridded-glacier-mass-change_data-completeness_q01", "dataset_id": "derived-gridded-glacier-mass-change", "store": "CDS", "doc_type": "EQC_QA", "aspect": "data-completeness_q01", "aspect_base": "data-completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Glacier mass change data from satellite and in-situ observations: uncertainty analysis for glaciological and climate change monitoring > Analysis and results > 3. Quantifying global (cumulative) glacier mass changes and their error estimates since 1975-1976 > 3.1 Global glacier mass changes", "title": "Glacier mass change data from satellite and in-situ observations: uncertainty analysis for glaciological and climate change monitoring", "chunk_index": 10, "token_count": 632, "text_raw": "In the following section, we plot the global annual (and cumulative) glacier mass change over time. For the annual values, we therefore sum the gridded mass change product over the entire spatial domain for each individual year to get spatially summed values in Gt yr⁻¹, and calculate the glacier area-weighted mean to get a mass balance value in m w.e. yr⁻¹:\n\n$\\Delta M_i \n$\n[Gt yr⁻¹]\n$ = \\sum\\limits^{x,y}\\Delta {M_{x,y,i}}$\n\nwhere $\\Delta {M_{x,y,i}}$ is the glacier mass change (from `glacier_mass_change_gt`, in Gt yr⁻¹) at pixel $x,y$ during a certain year $i$, and\n\n${\\Delta B_i} \n$\n[m w.e. yr⁻¹]\n$ = \\textstyle\\dfrac{1}{\\sum\\limits^{x,y} A_{x,y,i}} {\\sum\\limits^{x,y} (A_{x,y,i} * \\Delta {B_{x,y,i}})}$\n\nwhere $\\Delta {B_{x,y,i}}$ is the glacier mass balance (from `glacier_mass_change_mwe`, in m w.e. yr⁻¹) and $A_{x,y,i}$ the glacier area [km$^2$] at pixel $x,y$ (from `glacier_area_km2`) during a certain year $i$.\n\nThis results in the following plot:\n\nPlot the data\n\n*
Figure 5. Annual global (left) glacier mass changes (expressed in Gt yr⁻¹) and (right) mass balances (expressed in m w.e. yr⁻¹) from the glacier mass change dataset.
*\n\nThe mass changes in the left plot above are expressed in Gt yr$^{-1}$ (Gigatonnes per year). Since Gt is a unit of mass, the mass of 1 Gt of ice is exactly the same as the mass of 1 Gt of water. The value can, however, also be translated into a volume. For example, 1 Gt of water (density 1000 kg/m³) is exactly 1 km³ water, while 1 Gt of ice (density 917 kg/m³) becomes 1.091 km³ of ice in volume.\n\n(section-3-2)=", "text_with_prefix": "EQC Quality Assessment: \"Glacier mass change data from satellite and in-situ observations: uncertainty analysis for glaciological and climate change monitoring\"\nDataset: derived-gridded-glacier-mass-change [CDS]\nAspect: data-completeness_q01 | Category: Satellite_ECVs\nSection: Glacier mass change data from satellite and in-situ observations: uncertainty analysis for glaciological and climate change monitoring > Analysis and results > 3. Quantifying global (cumulative) glacier mass changes and their error estimates since 1975-1976 > 3.1 Global glacier mass changes\n---\nIn the following section, we plot the global annual (and cumulative) glacier mass change over time. For the annual values, we therefore sum the gridded mass change product over the entire spatial domain for each individual year to get spatially summed values in Gt yr⁻¹, and calculate the glacier area-weighted mean to get a mass balance value in m w.e. yr⁻¹:\n\n$\\Delta M_i \n$\n[Gt yr⁻¹]\n$ = \\sum\\limits^{x,y}\\Delta {M_{x,y,i}}$\n\nwhere $\\Delta {M_{x,y,i}}$ is the glacier mass change (from `glacier_mass_change_gt`, in Gt yr⁻¹) at pixel $x,y$ during a certain year $i$, and\n\n${\\Delta B_i} \n$\n[m w.e. yr⁻¹]\n$ = \\textstyle\\dfrac{1}{\\sum\\limits^{x,y} A_{x,y,i}} {\\sum\\limits^{x,y} (A_{x,y,i} * \\Delta {B_{x,y,i}})}$\n\nwhere $\\Delta {B_{x,y,i}}$ is the glacier mass balance (from `glacier_mass_change_mwe`, in m w.e. yr⁻¹) and $A_{x,y,i}$ the glacier area [km$^2$] at pixel $x,y$ (from `glacier_area_km2`) during a certain year $i$.\n\nThis results in the following plot:\n\nPlot the data\n\n*
Figure 5. Annual global (left) glacier mass changes (expressed in Gt yr⁻¹) and (right) mass balances (expressed in m w.e. yr⁻¹) from the glacier mass change dataset.
*\n\nThe mass changes in the left plot above are expressed in Gt yr$^{-1}$ (Gigatonnes per year). Since Gt is a unit of mass, the mass of 1 Gt of ice is exactly the same as the mass of 1 Gt of water. The value can, however, also be translated into a volume. For example, 1 Gt of water (density 1000 kg/m³) is exactly 1 km³ water, while 1 Gt of ice (density 917 kg/m³) becomes 1.091 km³ of ice in volume.\n\n(section-3-2)="} {"chunk_id": "satellite_derived-gridded-glacier-mass-change_data-completeness_q01__dc008d672a31", "report_id": "satellite_derived-gridded-glacier-mass-change_data-completeness_q01", "dataset_id": "derived-gridded-glacier-mass-change", "store": "CDS", "doc_type": "EQC_QA", "aspect": "data-completeness_q01", "aspect_base": "data-completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Glacier mass change data from satellite and in-situ observations: uncertainty analysis for glaciological and climate change monitoring > Analysis and results > 3. Quantifying global (cumulative) glacier mass changes and their error estimates since 1975-1976 > 3.2 Global glacier mass change error estimates", "title": "Glacier mass change data from satellite and in-situ observations: uncertainty analysis for glaciological and climate change monitoring", "chunk_index": 11, "token_count": 1097, "text_raw": "By making use of the laws of error propagation, we can also determine how the glacier mass change errors propagate throughout the time series. For error propagation, we assume that:\n* Errors are spatially correlated within a certain RGI region, but are uncorrelated in between RGI regions\n* Errors are uncorrelated in time\n\nThe corresponding annual mass change uncertainty is calculated by assuming that errors are spatially correlated within a certain RGI region. To create the global error estimate, we assume uncorrelated/independent errors between the 19 RGI regions [[2](https://doi.org/10.5194/essd-17-1977-2025)]. For error propagation, we also divide the uncertainty estimates by 1.96 since errors values are reported as 1.96 times the standard deviation in the original dataset:\n\n$\n\\sigma_{\\Delta{M_i}} \n$\n[Gt yr⁻¹]\n$= \\sqrt{\\sum\\limits^{19}_{region=1}({\\sigma_{ \\Delta{M_{region,i}}}})^2}\n$\n, where\n$ {\\sigma_{ \\Delta{M_{region,i}}}} = \\sum\\limits^{x,y \\in \\text{region}}({\\sigma_{ \\Delta{M_{x,y,i}}}})\n$\n\nand $\\sigma_{ \\Delta{M_{x,y,i}}}$ is the mass change standard deviation (from `uncertainty_gt` divided by 1.96, in Gt yr⁻¹) at pixel $x,y$ during a certain year $i$ in a certain RGI region, and\n\n$\n\\sigma_{{\\Delta{B_i}}} \n$\n[m w.e. yr⁻¹]\n$= \\sqrt{\\sum\\limits^{19}_{region=1}({\\sigma_{ \\Delta{B_{region,i}}}})^2}\n$\n, where\n${\\sigma_{ \\Delta{B_{region,i}}}} = {\\sum\\limits^{x,y \\in \\text{region}} \\left( \\dfrac{A_{x,y,i}}{\\sum\\limits^{x,y \\in \\text{region}} A_{x,y,i}} * {\\sigma_{{\\Delta{B}_{x,y,i}}}} \\right) }\n$\n\nwhere $\\sigma_{{ \\Delta{B_{x,y,i}}}}$ is the mass balance standard deviation (from `uncertainty_mwe` divided by 1.96, in m w.e. yr⁻¹) at pixel $x,y$ during a certain year $i$ and $A_{x,y,i}$ the glacier area [km$^2$] (from `glacier_area_km2`).\n\nPlotting these data reveals the following:\n\nPlot the data\n\n*
Figure 6. Annual global glacier mass change errors from the glacier mass change dataset expressed in units of (left) Gt yr⁻¹ (mass changes) and (right) m w.e. yr⁻¹ (mass balances).
*\n\nIt can be seen from the plot that the global glacier mass change uncertainty has been decreasing over time, with a sudden drop around 2000 CE. The data can also be plotted in a cumulative way:\n\n$\n{M} \n$\n[Gt]\n$\n= \\sum\\limits_{i={1976}}^{{{1976+n-1}}} (\\Delta M_i)\n$\n, or\n\n$\n{{B}} \n$\n[m w.e.]\n$\n= \\sum\\limits_{i={1976}}^{{{1976+n-1}}} ({\\Delta B_i})\n$\n\nwith the global annual glacier mass change (in Gt yr⁻¹) or balance (m w.e. yr⁻¹) at a certain year $i$ (as calculated above) and $n$ the number of years in the time series.\n\nThe corresponding uncertainty is calculated by assuming uncorrelated errors over time:\n\n$\n\\sigma_{{M}}\n$\n[Gt]\n$\n= \\sqrt{\\sum\\limits_{i={1976}}^{{1976+n-1}} (\\sigma_{\\Delta{M_i}})^2}\n$\n, or\n\n$\n\\sigma_{{{B}}}\n$\n[m w.e.] = \n$\n\\sqrt{\\sum\\limits_{i={1976}}^{{1976+n-1}} (\\sigma_{{\\Delta{B_i}}})^2}\n$\n\nwith the global annual mass change (in Gt yr⁻¹) or balance (m w.e. yr⁻¹) uncertainty at a certain year $i$ (as calculated above) and $n$ the number of years in the time series.\n\n(section-3-3)=", "text_with_prefix": "EQC Quality Assessment: \"Glacier mass change data from satellite and in-situ observations: uncertainty analysis for glaciological and climate change monitoring\"\nDataset: derived-gridded-glacier-mass-change [CDS]\nAspect: data-completeness_q01 | Category: Satellite_ECVs\nSection: Glacier mass change data from satellite and in-situ observations: uncertainty analysis for glaciological and climate change monitoring > Analysis and results > 3. Quantifying global (cumulative) glacier mass changes and their error estimates since 1975-1976 > 3.2 Global glacier mass change error estimates\n---\nBy making use of the laws of error propagation, we can also determine how the glacier mass change errors propagate throughout the time series. For error propagation, we assume that:\n* Errors are spatially correlated within a certain RGI region, but are uncorrelated in between RGI regions\n* Errors are uncorrelated in time\n\nThe corresponding annual mass change uncertainty is calculated by assuming that errors are spatially correlated within a certain RGI region. To create the global error estimate, we assume uncorrelated/independent errors between the 19 RGI regions [[2](https://doi.org/10.5194/essd-17-1977-2025)]. For error propagation, we also divide the uncertainty estimates by 1.96 since errors values are reported as 1.96 times the standard deviation in the original dataset:\n\n$\n\\sigma_{\\Delta{M_i}} \n$\n[Gt yr⁻¹]\n$= \\sqrt{\\sum\\limits^{19}_{region=1}({\\sigma_{ \\Delta{M_{region,i}}}})^2}\n$\n, where\n$ {\\sigma_{ \\Delta{M_{region,i}}}} = \\sum\\limits^{x,y \\in \\text{region}}({\\sigma_{ \\Delta{M_{x,y,i}}}})\n$\n\nand $\\sigma_{ \\Delta{M_{x,y,i}}}$ is the mass change standard deviation (from `uncertainty_gt` divided by 1.96, in Gt yr⁻¹) at pixel $x,y$ during a certain year $i$ in a certain RGI region, and\n\n$\n\\sigma_{{\\Delta{B_i}}} \n$\n[m w.e. yr⁻¹]\n$= \\sqrt{\\sum\\limits^{19}_{region=1}({\\sigma_{ \\Delta{B_{region,i}}}})^2}\n$\n, where\n${\\sigma_{ \\Delta{B_{region,i}}}} = {\\sum\\limits^{x,y \\in \\text{region}} \\left( \\dfrac{A_{x,y,i}}{\\sum\\limits^{x,y \\in \\text{region}} A_{x,y,i}} * {\\sigma_{{\\Delta{B}_{x,y,i}}}} \\right) }\n$\n\nwhere $\\sigma_{{ \\Delta{B_{x,y,i}}}}$ is the mass balance standard deviation (from `uncertainty_mwe` divided by 1.96, in m w.e. yr⁻¹) at pixel $x,y$ during a certain year $i$ and $A_{x,y,i}$ the glacier area [km$^2$] (from `glacier_area_km2`).\n\nPlotting these data reveals the following:\n\nPlot the data\n\n*
Figure 6. Annual global glacier mass change errors from the glacier mass change dataset expressed in units of (left) Gt yr⁻¹ (mass changes) and (right) m w.e. yr⁻¹ (mass balances).
*\n\nIt can be seen from the plot that the global glacier mass change uncertainty has been decreasing over time, with a sudden drop around 2000 CE. The data can also be plotted in a cumulative way:\n\n$\n{M} \n$\n[Gt]\n$\n= \\sum\\limits_{i={1976}}^{{{1976+n-1}}} (\\Delta M_i)\n$\n, or\n\n$\n{{B}} \n$\n[m w.e.]\n$\n= \\sum\\limits_{i={1976}}^{{{1976+n-1}}} ({\\Delta B_i})\n$\n\nwith the global annual glacier mass change (in Gt yr⁻¹) or balance (m w.e. yr⁻¹) at a certain year $i$ (as calculated above) and $n$ the number of years in the time series.\n\nThe corresponding uncertainty is calculated by assuming uncorrelated errors over time:\n\n$\n\\sigma_{{M}}\n$\n[Gt]\n$\n= \\sqrt{\\sum\\limits_{i={1976}}^{{1976+n-1}} (\\sigma_{\\Delta{M_i}})^2}\n$\n, or\n\n$\n\\sigma_{{{B}}}\n$\n[m w.e.] = \n$\n\\sqrt{\\sum\\limits_{i={1976}}^{{1976+n-1}} (\\sigma_{{\\Delta{B_i}}})^2}\n$\n\nwith the global annual mass change (in Gt yr⁻¹) or balance (m w.e. yr⁻¹) uncertainty at a certain year $i$ (as calculated above) and $n$ the number of years in the time series.\n\n(section-3-3)="} {"chunk_id": "satellite_derived-gridded-glacier-mass-change_data-completeness_q01__370a3c748031", "report_id": "satellite_derived-gridded-glacier-mass-change_data-completeness_q01", "dataset_id": "derived-gridded-glacier-mass-change", "store": "CDS", "doc_type": "EQC_QA", "aspect": "data-completeness_q01", "aspect_base": "data-completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Glacier mass change data from satellite and in-situ observations: uncertainty analysis for glaciological and climate change monitoring > Analysis and results > 3. Quantifying global (cumulative) glacier mass changes and their error estimates since 1975-1976 > 3.3 Time series of global cumulative glacier mass changes and their error estimate", "title": "Glacier mass change data from satellite and in-situ observations: uncertainty analysis for glaciological and climate change monitoring", "chunk_index": 12, "token_count": 381, "text_raw": "The corresponding plot looks as follows:\n\nPlot the data\n\n*
Figure 7. Annual global cumulative (left) glacier mass changes (expressed in Gt yr⁻¹) and (right) mass balances (expressed in m w.e. yr⁻¹), with error propagation (shaded).
*\n\nFrom the image above, it can be seen that glaciers clearly have been losing mass during the observational period, especially since the 1990s, which is in line with findings in the literature (e.g. [[8](https://doi.org/10.1038/s41586-024-08545-z)]). The final estimate of the glaciers mass change at the end of the time series is:\n\n```text\nThe total global cumulative glacier mass change between 1975-1976 and 2021-2022 is -10.72 ± 1.12 m w.e. or -7691.70 ± 858.39 Gt.\n```\n\nNow, let us use the glacier mass change data to quantify the link between glacier melt and glacier-related contributions to global sea level rise in the context of Earth System modeling and climate change monitoring.\n\n(section-4)=", "text_with_prefix": "EQC Quality Assessment: \"Glacier mass change data from satellite and in-situ observations: uncertainty analysis for glaciological and climate change monitoring\"\nDataset: derived-gridded-glacier-mass-change [CDS]\nAspect: data-completeness_q01 | Category: Satellite_ECVs\nSection: Glacier mass change data from satellite and in-situ observations: uncertainty analysis for glaciological and climate change monitoring > Analysis and results > 3. Quantifying global (cumulative) glacier mass changes and their error estimates since 1975-1976 > 3.3 Time series of global cumulative glacier mass changes and their error estimate\n---\nThe corresponding plot looks as follows:\n\nPlot the data\n\n*
Figure 7. Annual global cumulative (left) glacier mass changes (expressed in Gt yr⁻¹) and (right) mass balances (expressed in m w.e. yr⁻¹), with error propagation (shaded).
*\n\nFrom the image above, it can be seen that glaciers clearly have been losing mass during the observational period, especially since the 1990s, which is in line with findings in the literature (e.g. [[8](https://doi.org/10.1038/s41586-024-08545-z)]). The final estimate of the glaciers mass change at the end of the time series is:\n\n```text\nThe total global cumulative glacier mass change between 1975-1976 and 2021-2022 is -10.72 ± 1.12 m w.e. or -7691.70 ± 858.39 Gt.\n```\n\nNow, let us use the glacier mass change data to quantify the link between glacier melt and glacier-related contributions to global sea level rise in the context of Earth System modeling and climate change monitoring.\n\n(section-4)="} {"chunk_id": "satellite_derived-gridded-glacier-mass-change_data-completeness_q01__622a6c695135", "report_id": "satellite_derived-gridded-glacier-mass-change_data-completeness_q01", "dataset_id": "derived-gridded-glacier-mass-change", "store": "CDS", "doc_type": "EQC_QA", "aspect": "data-completeness_q01", "aspect_base": "data-completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Glacier mass change data from satellite and in-situ observations: uncertainty analysis for glaciological and climate change monitoring > Analysis and results > 4. Quantification of glacier-related contributions to global sea level change > 4.1 Computing sea-level rise equivalent from glacier mass loss", "title": "Glacier mass change data from satellite and in-situ observations: uncertainty analysis for glaciological and climate change monitoring", "chunk_index": 13, "token_count": 753, "text_raw": "We can calculate the cumulative contribution of glacier mass changes to global sea level change during the last several decades by making use of the following formula [[5](https://doi.org/10.1038/s41561-019-0300-3)]:\n\n$\nh_{SLE} \n$\n[mm]\n$\n= 1 \\cdot 10^6 * \\sum\\limits_{i=1976}^{1976+n-1} \\left(\\dfrac{\\Delta V_{i}}{A_s}\\right)\n$\n\nwhere \n$\n\\Delta{V_{i}} = \\dfrac{\\rho_w}{\\rho_s} * \\sum\\limits^{x,y}\\left(\\dfrac{\\Delta {B_{x,y,i}}}{1 \\cdot 10^3} * A_{x,y,i}\\right)\n$.\n\nHere, $\\Delta {B}$ is the glacier mass balance (in m w.e. yr⁻¹, divided by 1000 to get units in km) at pixel $x,y$ during a certain year $i$ from `ds[\"glacier_mass_change_mwe\"]`, and $A$ the glacier area [km²] from `ds[\"glacier_area_km2\"]`, so that the annual volume change $\\Delta V$ has units of km$^3$ yr$^{-1}$ of water. Furthermore, ${A_{s}}$ is the ocean surface area [km²], $n$ the total number of years in the time series and $\\rho$ the respective densities of water $w$ and the sea $s$ [kg m⁻³]. Note that this formulation assumes that all glacier mass or volume losses directly contribute to sea level changes, while this is not necessarily the case (e.g. [[6](https://doi.org/10.1088/1748-9326/aac2f0), [7](https://doi.org/10.5194/tc-14-833-2020)]). Also note that by multiplying the sea level contribution by $\\frac{\\rho_w}{\\rho_s}$, we calculate sea level contributions in ocean water column equivalent.\n\nThe corresponding uncertainty is given by:\n\n$\n\\sigma_{h_{SLE}} \n$\n[mm] = \n$\n1 \\cdot 10^6 * \\left(\\dfrac{1}{A_s} \\dfrac{{\\rho_{w}}}{\\rho_{s}}\\right) * \\sqrt{ \\sum\\limits_{i={1976}}^{{{1976+n-1}}} (\\sigma_{{V_{i}}})^2}\n$\n\nwhere $\\sigma_{\\Delta{V}}$ is the corresponding volume change uncertainty [km$^{3}$ yr$^{-1}$] calculated from the mass balance uncertainty and the glacier area, again taking into account that uncertainties at the pixel level are given as 1.96 times the standard deviation and that errors within a certain RGI region are correlated, while uncorrelated between RGI regions and over time, as before.\n\n(section-4-2)=", "text_with_prefix": "EQC Quality Assessment: \"Glacier mass change data from satellite and in-situ observations: uncertainty analysis for glaciological and climate change monitoring\"\nDataset: derived-gridded-glacier-mass-change [CDS]\nAspect: data-completeness_q01 | Category: Satellite_ECVs\nSection: Glacier mass change data from satellite and in-situ observations: uncertainty analysis for glaciological and climate change monitoring > Analysis and results > 4. Quantification of glacier-related contributions to global sea level change > 4.1 Computing sea-level rise equivalent from glacier mass loss\n---\nWe can calculate the cumulative contribution of glacier mass changes to global sea level change during the last several decades by making use of the following formula [[5](https://doi.org/10.1038/s41561-019-0300-3)]:\n\n$\nh_{SLE} \n$\n[mm]\n$\n= 1 \\cdot 10^6 * \\sum\\limits_{i=1976}^{1976+n-1} \\left(\\dfrac{\\Delta V_{i}}{A_s}\\right)\n$\n\nwhere \n$\n\\Delta{V_{i}} = \\dfrac{\\rho_w}{\\rho_s} * \\sum\\limits^{x,y}\\left(\\dfrac{\\Delta {B_{x,y,i}}}{1 \\cdot 10^3} * A_{x,y,i}\\right)\n$.\n\nHere, $\\Delta {B}$ is the glacier mass balance (in m w.e. yr⁻¹, divided by 1000 to get units in km) at pixel $x,y$ during a certain year $i$ from `ds[\"glacier_mass_change_mwe\"]`, and $A$ the glacier area [km²] from `ds[\"glacier_area_km2\"]`, so that the annual volume change $\\Delta V$ has units of km$^3$ yr$^{-1}$ of water. Furthermore, ${A_{s}}$ is the ocean surface area [km²], $n$ the total number of years in the time series and $\\rho$ the respective densities of water $w$ and the sea $s$ [kg m⁻³]. Note that this formulation assumes that all glacier mass or volume losses directly contribute to sea level changes, while this is not necessarily the case (e.g. [[6](https://doi.org/10.1088/1748-9326/aac2f0), [7](https://doi.org/10.5194/tc-14-833-2020)]). Also note that by multiplying the sea level contribution by $\\frac{\\rho_w}{\\rho_s}$, we calculate sea level contributions in ocean water column equivalent.\n\nThe corresponding uncertainty is given by:\n\n$\n\\sigma_{h_{SLE}} \n$\n[mm] = \n$\n1 \\cdot 10^6 * \\left(\\dfrac{1}{A_s} \\dfrac{{\\rho_{w}}}{\\rho_{s}}\\right) * \\sqrt{ \\sum\\limits_{i={1976}}^{{{1976+n-1}}} (\\sigma_{{V_{i}}})^2}\n$\n\nwhere $\\sigma_{\\Delta{V}}$ is the corresponding volume change uncertainty [km$^{3}$ yr$^{-1}$] calculated from the mass balance uncertainty and the glacier area, again taking into account that uncertainties at the pixel level are given as 1.96 times the standard deviation and that errors within a certain RGI region are correlated, while uncorrelated between RGI regions and over time, as before.\n\n(section-4-2)="} {"chunk_id": "satellite_derived-gridded-glacier-mass-change_data-completeness_q01__0ccf5b12e312", "report_id": "satellite_derived-gridded-glacier-mass-change_data-completeness_q01", "dataset_id": "derived-gridded-glacier-mass-change", "store": "CDS", "doc_type": "EQC_QA", "aspect": "data-completeness_q01", "aspect_base": "data-completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Glacier mass change data from satellite and in-situ observations: uncertainty analysis for glaciological and climate change monitoring > Analysis and results > 4. Quantification of glacier-related contributions to global sea level change > 4.2 Time series of glacier-related sea level contribution", "title": "Glacier mass change data from satellite and in-situ observations: uncertainty analysis for glaciological and climate change monitoring", "chunk_index": 14, "token_count": 625, "text_raw": "This ultimately results in the following graph:\n\nDefine constants\nCompute regional volume change assuming spatial correlation\nCompute global volume change by summing over all RGI regions\nCompute global sea level equivalent (SLE)\nCompute regional uncertainty assuming full spatial correlation within regions\nCompute global uncertainty assuming uncorrelated RGI regions\nCompute cumulative uncertainty over time\nCompute global SLE uncertainty\nPlot the data with uncertainty shading\n\n*Figure 8. Cumulative contribution of glacier mass changes to global sea level changes (assuming all glacier mass changes contribute to sea level changes), including the propagation of the error (shaded).*\n\nLet us quantify the total glacier-related contribution to global sea level change since the beginning of the dataset:\n\n```text\nThe total contribution of glacier mass changes to global sea level change between 1975-1976 and 2021-2022 is 20.64 ± 2.30 mm.\n```\n\nThe C3S product can be interpreted as glacier mass changes relevant for sea level contributions, as both the glaciological and geodetic approach do not consider ice melt that occurs under the waterline, which already displaces water. In contrast, processes that do contribute to sea level changes – such as liquid runoff from ice above buoyancy and changes in the ice flux across the grounding line – will result in a surface elevation change, which is observed with DEM differencing and hence incorporated in the geodetic method. Even though this consideration is not exact – a precise separation of contributions from ice above buoyancy (which affects sea level changes) and ice below buoyancy (which does not) would require the actual position of the grounding line, as well as values for the ice thickness and bedrock elevation field (e.g. [[7](https://doi.org/10.5194/tc-14-833-2020)]) – it is assumed that the resulting mass change estimates from this approach fall within acceptable uncertainty bounds to consider them relevant for sea level changes [[2](https://doi.org/10.5194/essd-17-1977-2025)]. However, it should also be considered that not all mass changes affect sea levels directly, for example for landlocked glaciers whose meltwater does not reach the ocean [[6](https://doi.org/10.1088/1748-9326/aac2f0), [7](https://doi.org/10.5194/tc-14-833-2020)]).\n\n(section-5)=", "text_with_prefix": "EQC Quality Assessment: \"Glacier mass change data from satellite and in-situ observations: uncertainty analysis for glaciological and climate change monitoring\"\nDataset: derived-gridded-glacier-mass-change [CDS]\nAspect: data-completeness_q01 | Category: Satellite_ECVs\nSection: Glacier mass change data from satellite and in-situ observations: uncertainty analysis for glaciological and climate change monitoring > Analysis and results > 4. Quantification of glacier-related contributions to global sea level change > 4.2 Time series of glacier-related sea level contribution\n---\nThis ultimately results in the following graph:\n\nDefine constants\nCompute regional volume change assuming spatial correlation\nCompute global volume change by summing over all RGI regions\nCompute global sea level equivalent (SLE)\nCompute regional uncertainty assuming full spatial correlation within regions\nCompute global uncertainty assuming uncorrelated RGI regions\nCompute cumulative uncertainty over time\nCompute global SLE uncertainty\nPlot the data with uncertainty shading\n\n*Figure 8. Cumulative contribution of glacier mass changes to global sea level changes (assuming all glacier mass changes contribute to sea level changes), including the propagation of the error (shaded).*\n\nLet us quantify the total glacier-related contribution to global sea level change since the beginning of the dataset:\n\n```text\nThe total contribution of glacier mass changes to global sea level change between 1975-1976 and 2021-2022 is 20.64 ± 2.30 mm.\n```\n\nThe C3S product can be interpreted as glacier mass changes relevant for sea level contributions, as both the glaciological and geodetic approach do not consider ice melt that occurs under the waterline, which already displaces water. In contrast, processes that do contribute to sea level changes – such as liquid runoff from ice above buoyancy and changes in the ice flux across the grounding line – will result in a surface elevation change, which is observed with DEM differencing and hence incorporated in the geodetic method. Even though this consideration is not exact – a precise separation of contributions from ice above buoyancy (which affects sea level changes) and ice below buoyancy (which does not) would require the actual position of the grounding line, as well as values for the ice thickness and bedrock elevation field (e.g. [[7](https://doi.org/10.5194/tc-14-833-2020)]) – it is assumed that the resulting mass change estimates from this approach fall within acceptable uncertainty bounds to consider them relevant for sea level changes [[2](https://doi.org/10.5194/essd-17-1977-2025)]. However, it should also be considered that not all mass changes affect sea levels directly, for example for landlocked glaciers whose meltwater does not reach the ocean [[6](https://doi.org/10.1088/1748-9326/aac2f0), [7](https://doi.org/10.5194/tc-14-833-2020)]).\n\n(section-5)="} {"chunk_id": "satellite_derived-gridded-glacier-mass-change_data-completeness_q01__39fa0c9961b3", "report_id": "satellite_derived-gridded-glacier-mass-change_data-completeness_q01", "dataset_id": "derived-gridded-glacier-mass-change", "store": "CDS", "doc_type": "EQC_QA", "aspect": "data-completeness_q01", "aspect_base": "data-completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Glacier mass change data from satellite and in-situ observations: uncertainty analysis for glaciological and climate change monitoring > Analysis and results > 5. Short summary and take-home messages", "title": "Glacier mass change data from satellite and in-situ observations: uncertainty analysis for glaciological and climate change monitoring", "chunk_index": 15, "token_count": 629, "text_raw": "To be able to use the glacier mass change dataset for reliable climate change analysis/monitoring purposes, the dataset should at least exhibit a comprehensive spatial coverage (i.e. global), a long and continuous temporal coverage (> 30 years), quantified and transparent pixel-by-pixel uncertainty estimates that meet international proposed thresholds [[4](https://library.wmo.int/idurl/4/58111)], and an adequate spatio-temporal resolution (cfr. the \"Maturity Matrix\" [[9](https://doi.org/10.1175/BAMS-D-21-0109.1)]).\n\nSince these conditions are generally met, the C3S glacier mass change dataset on the CDS is considered highly suitable for monitoring and deriving global cumulative mass changes over time, as well as assessing their impact on global sea level changes. The dataset, for example, includes pixel-by-pixel error estimates, expressed as precision errors (1.96 times the standard deviation). The corresponding uncertainty furthermore decreases over time, with post-2000 data meeting international GCOS standards. There is, however, a high spatial and temporal variation in the error estimates, implying that not all pixels meet these uncertainty thresholds. Users should therefore exercise caution in certain areas, such as the peripheral glaciers of Antarctica and Greenland, where uncertainties are higher. The dataset furthermore meets GCOS requirements in terms of resolution, spatial coverage (global) and temporal continuity (since 1975-76 without gaps), which demostrates its maturity and reliability for Earth system modeling and climate change monitoring. Data quality is hence considered strong, especially after 2000 and outside the high-uncertainty peripheral regions of the ice sheets.\n\nThe glacier mass change data can be considered to be directly transformable to sea level contributions, provided certain assumptions are made [[2](https://doi.org/10.5194/essd-17-1977-2025)]. Additionally, users should note limitations, such as the absence of detailed sampling density information, the fact that data can not be consulted at the individual glacier-scale, and that data are measured and generated by different institutes/research groups and from different methods, which could result in inconsistencies. Moreover, since grid cells are not of the same absolute surface area and decrease in size at higher latitudes, mass change artefacts may arise in polar regions if individual glaciers are larger than the cell size. This may affect glacier mass change representativeness in some areas. Therefore, knowledge of the amount and spatial distribution of glaciers with a consistent and long-term glaciological sample is crucial for assessing the representativeness of the data.", "text_with_prefix": "EQC Quality Assessment: \"Glacier mass change data from satellite and in-situ observations: uncertainty analysis for glaciological and climate change monitoring\"\nDataset: derived-gridded-glacier-mass-change [CDS]\nAspect: data-completeness_q01 | Category: Satellite_ECVs\nSection: Glacier mass change data from satellite and in-situ observations: uncertainty analysis for glaciological and climate change monitoring > Analysis and results > 5. Short summary and take-home messages\n---\nTo be able to use the glacier mass change dataset for reliable climate change analysis/monitoring purposes, the dataset should at least exhibit a comprehensive spatial coverage (i.e. global), a long and continuous temporal coverage (> 30 years), quantified and transparent pixel-by-pixel uncertainty estimates that meet international proposed thresholds [[4](https://library.wmo.int/idurl/4/58111)], and an adequate spatio-temporal resolution (cfr. the \"Maturity Matrix\" [[9](https://doi.org/10.1175/BAMS-D-21-0109.1)]).\n\nSince these conditions are generally met, the C3S glacier mass change dataset on the CDS is considered highly suitable for monitoring and deriving global cumulative mass changes over time, as well as assessing their impact on global sea level changes. The dataset, for example, includes pixel-by-pixel error estimates, expressed as precision errors (1.96 times the standard deviation). The corresponding uncertainty furthermore decreases over time, with post-2000 data meeting international GCOS standards. There is, however, a high spatial and temporal variation in the error estimates, implying that not all pixels meet these uncertainty thresholds. Users should therefore exercise caution in certain areas, such as the peripheral glaciers of Antarctica and Greenland, where uncertainties are higher. The dataset furthermore meets GCOS requirements in terms of resolution, spatial coverage (global) and temporal continuity (since 1975-76 without gaps), which demostrates its maturity and reliability for Earth system modeling and climate change monitoring. Data quality is hence considered strong, especially after 2000 and outside the high-uncertainty peripheral regions of the ice sheets.\n\nThe glacier mass change data can be considered to be directly transformable to sea level contributions, provided certain assumptions are made [[2](https://doi.org/10.5194/essd-17-1977-2025)]. Additionally, users should note limitations, such as the absence of detailed sampling density information, the fact that data can not be consulted at the individual glacier-scale, and that data are measured and generated by different institutes/research groups and from different methods, which could result in inconsistencies. Moreover, since grid cells are not of the same absolute surface area and decrease in size at higher latitudes, mass change artefacts may arise in polar regions if individual glaciers are larger than the cell size. This may affect glacier mass change representativeness in some areas. Therefore, knowledge of the amount and spatial distribution of glaciers with a consistent and long-term glaciological sample is crucial for assessing the representativeness of the data."} {"chunk_id": "satellite_derived-gridded-glacier-mass-change_data-completeness_q01__e35a7e9a5437", "report_id": "satellite_derived-gridded-glacier-mass-change_data-completeness_q01", "dataset_id": "derived-gridded-glacier-mass-change", "store": "CDS", "doc_type": "EQC_QA", "aspect": "data-completeness_q01", "aspect_base": "data-completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Glacier mass change data from satellite and in-situ observations: uncertainty analysis for glaciological and climate change monitoring > ℹ️ If you want to know more > Key resources", "title": "Glacier mass change data from satellite and in-situ observations: uncertainty analysis for glaciological and climate change monitoring", "chunk_index": 16, "token_count": 358, "text_raw": "- [Glacier mass change gridded data from 1976 to present derived from the Fluctuations of Glaciers Database](https://cds.climate.copernicus.eu/datasets/derived-gridded-glacier-mass-change?tab=overview)\n- [Documentation on the CDS](https://cds.climate.copernicus.eu/datasets/derived-gridded-glacier-mass-change?tab=overview) and the [ECMWF Confluence Wiki](https://confluence.ecmwf.int/pages/viewpage.action?pageId=355349383) (Copernicus Knowledge Base)\n- [World Glacier Monitoring Service (WGMS)](https://wgms.ch/)\n- [Glacier Mass Balance Intercomparison Exercise (GlaMBIE)](https://glambie.org/)\n- [Copernicus climate change indicators: glaciers](https://climate.copernicus.eu/climate-indicators/glaciers)\n- [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control` prepared by [B-Open](https://www.bopen.eu)", "text_with_prefix": "EQC Quality Assessment: \"Glacier mass change data from satellite and in-situ observations: uncertainty analysis for glaciological and climate change monitoring\"\nDataset: derived-gridded-glacier-mass-change [CDS]\nAspect: data-completeness_q01 | Category: Satellite_ECVs\nSection: Glacier mass change data from satellite and in-situ observations: uncertainty analysis for glaciological and climate change monitoring > ℹ️ If you want to know more > Key resources\n---\n- [Glacier mass change gridded data from 1976 to present derived from the Fluctuations of Glaciers Database](https://cds.climate.copernicus.eu/datasets/derived-gridded-glacier-mass-change?tab=overview)\n- [Documentation on the CDS](https://cds.climate.copernicus.eu/datasets/derived-gridded-glacier-mass-change?tab=overview) and the [ECMWF Confluence Wiki](https://confluence.ecmwf.int/pages/viewpage.action?pageId=355349383) (Copernicus Knowledge Base)\n- [World Glacier Monitoring Service (WGMS)](https://wgms.ch/)\n- [Glacier Mass Balance Intercomparison Exercise (GlaMBIE)](https://glambie.org/)\n- [Copernicus climate change indicators: glaciers](https://climate.copernicus.eu/climate-indicators/glaciers)\n- [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control` prepared by [B-Open](https://www.bopen.eu)"} {"chunk_id": "satellite_derived-gridded-glacier-mass-change_data-completeness_q01__72012626e891", "report_id": "satellite_derived-gridded-glacier-mass-change_data-completeness_q01", "dataset_id": "derived-gridded-glacier-mass-change", "store": "CDS", "doc_type": "EQC_QA", "aspect": "data-completeness_q01", "aspect_base": "data-completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Glacier mass change data from satellite and in-situ observations: uncertainty analysis for glaciological and climate change monitoring > ℹ️ If you want to know more > References", "title": "Glacier mass change data from satellite and in-situ observations: uncertainty analysis for glaciological and climate change monitoring", "chunk_index": 17, "token_count": 877, "text_raw": "- [[1](https://wgms.ch/)] WGMS (2022). Fluctuations of Glaciers Database. doi: 10.5904/wgms-fog-2022-09.\n\n- [[2](https://doi.org/10.5194/essd-17-1977-2025)] Dussaillant, I., Hugonnet, R., Huss, M., Berthier, E., Bannwart, J., Paul, F., and Zemp, M. (2025). Annual mass changes for each glacier in the world from 1976 to 2023, Earth Syst. Sci. Data, doi: 10.5194/essd-2024-323.\n\n- [[3](https://doi.org/10.1038/s41586-019-1071-0)] Zemp, M., Huss, M., Thibert, E., Eckert, N., McNabb, R., Huber, J., Barandun, M., Machguth, H., Nussbaumer, S. U., Gärtner-Roer, I., Thomson, L., Paul, F., Maussion, F., Kutuzov, S., and Cogley, J. G. (2019). Global glacier mass changes and their contributions to sea-level rise from 1961 to 2016. Nature, 568, 382–386. doi: 10.1038/s41586-019-1071-0.\n\n- [[4](https://library.wmo.int/idurl/4/58111)] GCOS (Global Climate Observing System) (2022). The 2022 GCOS ECVs Requirements (GCOS-245). World Meteorological Organization: Geneva, Switzerland. doi: https://library.wmo.int/idurl/4/58111\n\n- [[5](https://doi.org/10.1038/s41561-019-0300-3)] Farinotti, D., Huss, M., Fürst, J. J., Landmann, J., Machguth, H., Maussion, F., and Pandit, A. (2019). A consensus estimate for the ice thickness distribution of all glaciers on Earth. Nature Geoscience, 12(3), 168-173. doi: 10.1038/s41561-019-0300-3.\n\n- [[6](https://doi.org/10.1088/1748-9326/aac2f0)] Bamber, J. L., Westaway, R. M., Marzeion, B., and Wouters, B. (2018): The land ice contribution to sea level during the satellite era. Environmental Research Letters, 13(6), 063008, doi: 10.1088/1748-9326/aac2f0.\n\n- [[7](https://doi.org/10.5194/tc-14-833-2020)] Goelzer, H., Coulon, V., Pattyn, F., de Boer, B., and van de Wal, R. (2020). Brief communication: On calculating the sea-level contribution in marine ice-sheet models, The Cryosphere, 14, 833–840, doi: 10.5194/tc-14-833-2020.\n\n- [[8](https://doi.org/10.1038/s41586-024-08545-z)] The GlaMBIE Team (2025). Community estimate of global glacier mass changes from 2000 to 2023. Nature (2025). doi: 10.1038/s41586-024-08545-z.", "text_with_prefix": "EQC Quality Assessment: \"Glacier mass change data from satellite and in-situ observations: uncertainty analysis for glaciological and climate change monitoring\"\nDataset: derived-gridded-glacier-mass-change [CDS]\nAspect: data-completeness_q01 | Category: Satellite_ECVs\nSection: Glacier mass change data from satellite and in-situ observations: uncertainty analysis for glaciological and climate change monitoring > ℹ️ If you want to know more > References\n---\n- [[1](https://wgms.ch/)] WGMS (2022). Fluctuations of Glaciers Database. doi: 10.5904/wgms-fog-2022-09.\n\n- [[2](https://doi.org/10.5194/essd-17-1977-2025)] Dussaillant, I., Hugonnet, R., Huss, M., Berthier, E., Bannwart, J., Paul, F., and Zemp, M. (2025). Annual mass changes for each glacier in the world from 1976 to 2023, Earth Syst. Sci. Data, doi: 10.5194/essd-2024-323.\n\n- [[3](https://doi.org/10.1038/s41586-019-1071-0)] Zemp, M., Huss, M., Thibert, E., Eckert, N., McNabb, R., Huber, J., Barandun, M., Machguth, H., Nussbaumer, S. U., Gärtner-Roer, I., Thomson, L., Paul, F., Maussion, F., Kutuzov, S., and Cogley, J. G. (2019). Global glacier mass changes and their contributions to sea-level rise from 1961 to 2016. Nature, 568, 382–386. doi: 10.1038/s41586-019-1071-0.\n\n- [[4](https://library.wmo.int/idurl/4/58111)] GCOS (Global Climate Observing System) (2022). The 2022 GCOS ECVs Requirements (GCOS-245). World Meteorological Organization: Geneva, Switzerland. doi: https://library.wmo.int/idurl/4/58111\n\n- [[5](https://doi.org/10.1038/s41561-019-0300-3)] Farinotti, D., Huss, M., Fürst, J. J., Landmann, J., Machguth, H., Maussion, F., and Pandit, A. (2019). A consensus estimate for the ice thickness distribution of all glaciers on Earth. Nature Geoscience, 12(3), 168-173. doi: 10.1038/s41561-019-0300-3.\n\n- [[6](https://doi.org/10.1088/1748-9326/aac2f0)] Bamber, J. L., Westaway, R. M., Marzeion, B., and Wouters, B. (2018): The land ice contribution to sea level during the satellite era. Environmental Research Letters, 13(6), 063008, doi: 10.1088/1748-9326/aac2f0.\n\n- [[7](https://doi.org/10.5194/tc-14-833-2020)] Goelzer, H., Coulon, V., Pattyn, F., de Boer, B., and van de Wal, R. (2020). Brief communication: On calculating the sea-level contribution in marine ice-sheet models, The Cryosphere, 14, 833–840, doi: 10.5194/tc-14-833-2020.\n\n- [[8](https://doi.org/10.1038/s41586-024-08545-z)] The GlaMBIE Team (2025). Community estimate of global glacier mass changes from 2000 to 2023. Nature (2025). doi: 10.1038/s41586-024-08545-z."} {"chunk_id": "satellite_derived-gridded-glacier-mass-change_data-completeness_q01__74e77cc3a0d7", "report_id": "satellite_derived-gridded-glacier-mass-change_data-completeness_q01", "dataset_id": "derived-gridded-glacier-mass-change", "store": "CDS", "doc_type": "EQC_QA", "aspect": "data-completeness_q01", "aspect_base": "data-completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Glacier mass change data from satellite and in-situ observations: uncertainty analysis for glaciological and climate change monitoring > ℹ️ If you want to know more > References", "title": "Glacier mass change data from satellite and in-situ observations: uncertainty analysis for glaciological and climate change monitoring", "chunk_index": 18, "token_count": 482, "text_raw": "Community estimate of global glacier mass changes from 2000 to 2023. Nature (2025). doi: 10.1038/s41586-024-08545-z.\n\n- [[9](https://doi.org/10.1175/BAMS-D-21-0109.1)] Yang, C. X., Cagnazzo, C., Artale, V., Nardelli, B. B., Buontempo, C., Busatto, J., Caporaso, L., Cesarini, C., Cionni, I., Coll, J., Crezee, B., Cristofanelli, P., de Toma, V., Essa, Y. H., Eyring, V., Fierli, F., Grant, L., Hassler, B., Hirschi, M., Huybrechts, P., Le Merle, E., Leonelli, F. E., Lin, X., Madonna, F., Mason, E., Massonnet, F., Marcos, M., Marullo, S., Muller, B., Obregon, A., Organelli, E., Palacz, A., Pascual, A., Pisano, A., Putero, D., Rana, A., Sanchez-Roman, A., Seneviratne, S. I., Serva, F., Storto, A., Thiery, W., Throne, P., Van Tricht, L., Verhaegen, Y., Volpe, G., and Santoleri, R. (2022). Independent Quality Assessment of Essential Climate Variables: Lessons Learned from the Copernicus Climate Change Service, B. Am. Meteorol. Soc., 103, E2032–E2049, doi: 10.1175/Bams-D-21-0109.1.", "text_with_prefix": "EQC Quality Assessment: \"Glacier mass change data from satellite and in-situ observations: uncertainty analysis for glaciological and climate change monitoring\"\nDataset: derived-gridded-glacier-mass-change [CDS]\nAspect: data-completeness_q01 | Category: Satellite_ECVs\nSection: Glacier mass change data from satellite and in-situ observations: uncertainty analysis for glaciological and climate change monitoring > ℹ️ If you want to know more > References\n---\nCommunity estimate of global glacier mass changes from 2000 to 2023. Nature (2025). doi: 10.1038/s41586-024-08545-z.\n\n- [[9](https://doi.org/10.1175/BAMS-D-21-0109.1)] Yang, C. X., Cagnazzo, C., Artale, V., Nardelli, B. B., Buontempo, C., Busatto, J., Caporaso, L., Cesarini, C., Cionni, I., Coll, J., Crezee, B., Cristofanelli, P., de Toma, V., Essa, Y. H., Eyring, V., Fierli, F., Grant, L., Hassler, B., Hirschi, M., Huybrechts, P., Le Merle, E., Leonelli, F. E., Lin, X., Madonna, F., Mason, E., Massonnet, F., Marcos, M., Marullo, S., Muller, B., Obregon, A., Organelli, E., Palacz, A., Pascual, A., Pisano, A., Putero, D., Rana, A., Sanchez-Roman, A., Seneviratne, S. I., Serva, F., Storto, A., Thiery, W., Throne, P., Van Tricht, L., Verhaegen, Y., Volpe, G., and Santoleri, R. (2022). Independent Quality Assessment of Essential Climate Variables: Lessons Learned from the Copernicus Climate Change Service, B. Am. Meteorol. Soc., 103, E2032–E2049, doi: 10.1175/Bams-D-21-0109.1."} {"chunk_id": "satellite_derived-gridded-glacier-mass-change_trend-assessment_q02__4d4c4377f2cc", "report_id": "satellite_derived-gridded-glacier-mass-change_trend-assessment_q02", "dataset_id": "derived-gridded-glacier-mass-change", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q02", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Glacier mass change data from satellite and in-situ observations: resolution and coverage for trend analysis", "title": "Glacier mass change data from satellite and in-situ observations: resolution and coverage for trend analysis", "chunk_index": 0, "token_count": 130, "text_raw": "Production date: 31-05-2025\n\nDataset version: WGMS-FOG-2023-09\n\nProduced by: Yoni Verhaegen and Philippe Huybrechts (Vrije Universiteit Brussel)", "text_with_prefix": "EQC Quality Assessment: \"Glacier mass change data from satellite and in-situ observations: resolution and coverage for trend analysis\"\nDataset: derived-gridded-glacier-mass-change [CDS]\nAspect: trend-assessment_q02 | Category: Satellite_ECVs\nSection: Glacier mass change data from satellite and in-situ observations: resolution and coverage for trend analysis\n---\nProduction date: 31-05-2025\n\nDataset version: WGMS-FOG-2023-09\n\nProduced by: Yoni Verhaegen and Philippe Huybrechts (Vrije Universiteit Brussel)"} {"chunk_id": "satellite_derived-gridded-glacier-mass-change_trend-assessment_q02__c66e9ff3170c", "report_id": "satellite_derived-gridded-glacier-mass-change_trend-assessment_q02", "dataset_id": "derived-gridded-glacier-mass-change", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q02", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Glacier mass change data from satellite and in-situ observations: resolution and coverage for trend analysis > Quality assessment question", "title": "Glacier mass change data from satellite and in-situ observations: resolution and coverage for trend analysis", "chunk_index": 1, "token_count": 456, "text_raw": "* **\"Does the glacier mass change dataset have adequate spatial/temporal resolution, coverage (extent) and sampling density to derive multi-year trends in glacier mass changes, hereby enabling their use as indicators of climate change?\"**\n\nGlaciers significantly impact global sea-level rise, freshwater resources, natural hazards, hydro-power generation, recreation and tourism. Assessing glacier mass changes due to climate warming is therefore crucial for addressing these issues. The \"[Glacier mass change gridded data from 1976 to present derived from the Fluctuations of Glaciers Database](https://cds.climate.copernicus.eu/datasets/derived-gridded-glacier-mass-change?tab=overview)\" (version WGMS-FOG-2023-09) on the Climate Data Store (CDS), offers a global coverage of glacier mass changes by integrating in-situ, aerial, and satellite data [[1](https://wgms.ch/), [2](https://doi.org/10.5194/essd-17-1977-2025)]. The gridded glaciers mass change dataset that is on the CDS is currently one of the most complete dataset of glacier mass change data in terms of its spatial coverage. It is generally considered the main reference dataset to determine long-term glacier mass changes across the globe.\n\nWhen measured over a long period and at extended geographical scales, trends in glacier mass balance can be considered a clear indicator of global climate change [[3](https://doi.org/10.1038/s41586-019-1071-0)]. Despite some known issues, this dataset provides valuable insights into glacier mass changes across spatial and temporal scales. This notebook examines the dataset's suitability as a global climate change indicator, focusing on its resolution, coverage, and ability to reveal long-term trends in glacier mass balance.", "text_with_prefix": "EQC Quality Assessment: \"Glacier mass change data from satellite and in-situ observations: resolution and coverage for trend analysis\"\nDataset: derived-gridded-glacier-mass-change [CDS]\nAspect: trend-assessment_q02 | Category: Satellite_ECVs\nSection: Glacier mass change data from satellite and in-situ observations: resolution and coverage for trend analysis > Quality assessment question\n---\n* **\"Does the glacier mass change dataset have adequate spatial/temporal resolution, coverage (extent) and sampling density to derive multi-year trends in glacier mass changes, hereby enabling their use as indicators of climate change?\"**\n\nGlaciers significantly impact global sea-level rise, freshwater resources, natural hazards, hydro-power generation, recreation and tourism. Assessing glacier mass changes due to climate warming is therefore crucial for addressing these issues. The \"[Glacier mass change gridded data from 1976 to present derived from the Fluctuations of Glaciers Database](https://cds.climate.copernicus.eu/datasets/derived-gridded-glacier-mass-change?tab=overview)\" (version WGMS-FOG-2023-09) on the Climate Data Store (CDS), offers a global coverage of glacier mass changes by integrating in-situ, aerial, and satellite data [[1](https://wgms.ch/), [2](https://doi.org/10.5194/essd-17-1977-2025)]. The gridded glaciers mass change dataset that is on the CDS is currently one of the most complete dataset of glacier mass change data in terms of its spatial coverage. It is generally considered the main reference dataset to determine long-term glacier mass changes across the globe.\n\nWhen measured over a long period and at extended geographical scales, trends in glacier mass balance can be considered a clear indicator of global climate change [[3](https://doi.org/10.1038/s41586-019-1071-0)]. Despite some known issues, this dataset provides valuable insights into glacier mass changes across spatial and temporal scales. This notebook examines the dataset's suitability as a global climate change indicator, focusing on its resolution, coverage, and ability to reveal long-term trends in glacier mass balance."} {"chunk_id": "satellite_derived-gridded-glacier-mass-change_trend-assessment_q02__3f9f6b577111", "report_id": "satellite_derived-gridded-glacier-mass-change_trend-assessment_q02", "dataset_id": "derived-gridded-glacier-mass-change", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q02", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Glacier mass change data from satellite and in-situ observations: resolution and coverage for trend analysis > Quality assessment statements", "title": "Glacier mass change data from satellite and in-situ observations: resolution and coverage for trend analysis", "chunk_index": 2, "token_count": 436, "text_raw": "These are the key outcomes of this assessment\n\n- With its consistent global coverage since 1975-76, annual temporal resolution and a spatial resolution of 0.5° x 0.5°, the dataset meets the minimum international standards concerning spatio-temporal resolution and extent for the data to be useful in climate change analysis and for climate monitoring to become possible [[4](https://library.wmo.int/records/item/58111-the-2022-gcos-ecvs-requirements-gcos-245)]. The dataset thus enables the detection of reliable climate trends and statistical analyses, whereas the gridded format supports monitoring glacier mass changes at the local (pixel), regional, and global scales. Observed trends in glacier mass loss from the dataset furthermore align with theoretical models of the response of glaciers to climate warming, adding further credibility to the dataset. Additionally, majority of pixel-based linear mass change trends have been found to be statistically significant at $\\alpha$ = 0.05. \n- The product relies on a limited sample of (ca. 500) annual glaciological observations to provide the annual variability of glacier mass changes. However, the annual variability is calibrated to multi-annual or even decadal trends of geodetic mass balance estimates (i.e. glacier elevation change observations, converted to mass change) for more than 200,000 glaciers. Hence, the product may have a lesser pronounced observational coverage with respect to the annual variability of the glacier mass changes, but an excellent coverage with respect to their long-term trend. Product limitations include the fact that mass change are not available at the individual glacier scale, and that mass change artefacts may arise in polar regions if individual glaciers are larger than the cell size. \n```", "text_with_prefix": "EQC Quality Assessment: \"Glacier mass change data from satellite and in-situ observations: resolution and coverage for trend analysis\"\nDataset: derived-gridded-glacier-mass-change [CDS]\nAspect: trend-assessment_q02 | Category: Satellite_ECVs\nSection: Glacier mass change data from satellite and in-situ observations: resolution and coverage for trend analysis > Quality assessment statements\n---\nThese are the key outcomes of this assessment\n\n- With its consistent global coverage since 1975-76, annual temporal resolution and a spatial resolution of 0.5° x 0.5°, the dataset meets the minimum international standards concerning spatio-temporal resolution and extent for the data to be useful in climate change analysis and for climate monitoring to become possible [[4](https://library.wmo.int/records/item/58111-the-2022-gcos-ecvs-requirements-gcos-245)]. The dataset thus enables the detection of reliable climate trends and statistical analyses, whereas the gridded format supports monitoring glacier mass changes at the local (pixel), regional, and global scales. Observed trends in glacier mass loss from the dataset furthermore align with theoretical models of the response of glaciers to climate warming, adding further credibility to the dataset. Additionally, majority of pixel-based linear mass change trends have been found to be statistically significant at $\\alpha$ = 0.05. \n- The product relies on a limited sample of (ca. 500) annual glaciological observations to provide the annual variability of glacier mass changes. However, the annual variability is calibrated to multi-annual or even decadal trends of geodetic mass balance estimates (i.e. glacier elevation change observations, converted to mass change) for more than 200,000 glaciers. Hence, the product may have a lesser pronounced observational coverage with respect to the annual variability of the glacier mass changes, but an excellent coverage with respect to their long-term trend. Product limitations include the fact that mass change are not available at the individual glacier scale, and that mass change artefacts may arise in polar regions if individual glaciers are larger than the cell size. \n```"} {"chunk_id": "satellite_derived-gridded-glacier-mass-change_trend-assessment_q02__42c3919208a1", "report_id": "satellite_derived-gridded-glacier-mass-change_trend-assessment_q02", "dataset_id": "derived-gridded-glacier-mass-change", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q02", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Glacier mass change data from satellite and in-situ observations: resolution and coverage for trend analysis > Methodology > Dataset description", "title": "Glacier mass change data from satellite and in-situ observations: resolution and coverage for trend analysis", "chunk_index": 3, "token_count": 865, "text_raw": "The mass balance of a glacier is the difference between mass gained (mainly from snow accumulation) and mass lost (by meltwater runoff or solid ice discharge into lakes/the ocean), which is the same as the net mass change of a glacier over its surface area $A$:\n\n$\\Delta{M} = \\dfrac{1}{A} \\int\\limits_{A} b_a dA - \\dfrac{1}{A} \\int\\limits_{P} d_a dP$\n\nwhere:\n\n$b_a = \\int\\limits^{t_1}_{t_0}b_s dt$, with $b_s = ACC - ABL$ [kg m$^{-2}$ yr$^{-1}$]\n\n$d_a = \\int\\limits^{t_1}_{t_0}d_g dt$, with $d_g = \\rho_{i} \\cdot H \\cdot \\overline{V_P}$ [kg m$^{-1}$ yr$^{-1}$]\n\nThe first term represents the the total annual surface mass balance (i.e. the difference between accumulation and ablation), whereas the second term encompasses the total annual ice discharge across the grounding line $g$ with perimeter $P$. Here, $\\rho_{i}$ is the ice density, $H$ ice thickness and $\\overline{V_P}$ the gate-perpendicular vertically averaged horizontal velocity at the grounding line location. The latter term is considered zero for land-terminating glaciers. Other terms can be included as well, such as the basal and internal mass balances.\n\nIn general, the basis for the derived gridded mass changes are individual measurements (mainly glaciological in-situ local annual surface mass balance measurements) and geodetic air or spaceborne elevation change data (a surface elevation/ice thickness change or an ice volume change over time). These data are converted into an averaged specific mass balance value (i.e. mostly reported with units of meter water equivalent and often shortened to m w.e.) for an individual glacier. Afterwards, the data are submitted to the World Glacier Monitoring Service (WGMS). Further processing of the data results in a glacier mass change product reported over a 0.5° global grid dating back until the 1975-76 hydrological year. Each grid cell therefore contains a time series of total glacier mass change (in Gt yr⁻¹) or mass balance data (in m w.e. yr⁻¹) of all glaciers within the specific grid cell. In this notebook, we use version WGMS-FOG-2023-09. For a more detailed description of the data acquisition and processing methods, we refer to the [documentation on the CDS](https://cds.climate.copernicus.eu/datasets/derived-gridded-glacier-mass-change?tab=overview) and the [ECMWF Confluence Wiki](https://confluence.ecmwf.int/pages/viewpage.action?pageId=355349383) (Copernicus Knowledge Base).\n\nIt is important to note that the glaciers with an annual glaciological sample (ca. 500 glaciers) form the basis for the determination of annual mass changes from the remaining glaciers through a complex algorithm of spatial/temporal interpolation and (area-weighted) averaging of these mass change data [[2](https://essd.copernicus.org/preprints/essd-2024-323/)]. Hence, not all glaciers in the dataset exhibit a time series of directly measured annual in-situ mass balance observations, but most glaciers in the dataset are either unobserved or only have (limited and mostly multi-annual) geodetic mass change data available. Glacier mass change data with these geodetic samples are also more prone to uncertainties in general, for example due to uncertainties related to volume to mass conversions.", "text_with_prefix": "EQC Quality Assessment: \"Glacier mass change data from satellite and in-situ observations: resolution and coverage for trend analysis\"\nDataset: derived-gridded-glacier-mass-change [CDS]\nAspect: trend-assessment_q02 | Category: Satellite_ECVs\nSection: Glacier mass change data from satellite and in-situ observations: resolution and coverage for trend analysis > Methodology > Dataset description\n---\nThe mass balance of a glacier is the difference between mass gained (mainly from snow accumulation) and mass lost (by meltwater runoff or solid ice discharge into lakes/the ocean), which is the same as the net mass change of a glacier over its surface area $A$:\n\n$\\Delta{M} = \\dfrac{1}{A} \\int\\limits_{A} b_a dA - \\dfrac{1}{A} \\int\\limits_{P} d_a dP$\n\nwhere:\n\n$b_a = \\int\\limits^{t_1}_{t_0}b_s dt$, with $b_s = ACC - ABL$ [kg m$^{-2}$ yr$^{-1}$]\n\n$d_a = \\int\\limits^{t_1}_{t_0}d_g dt$, with $d_g = \\rho_{i} \\cdot H \\cdot \\overline{V_P}$ [kg m$^{-1}$ yr$^{-1}$]\n\nThe first term represents the the total annual surface mass balance (i.e. the difference between accumulation and ablation), whereas the second term encompasses the total annual ice discharge across the grounding line $g$ with perimeter $P$. Here, $\\rho_{i}$ is the ice density, $H$ ice thickness and $\\overline{V_P}$ the gate-perpendicular vertically averaged horizontal velocity at the grounding line location. The latter term is considered zero for land-terminating glaciers. Other terms can be included as well, such as the basal and internal mass balances.\n\nIn general, the basis for the derived gridded mass changes are individual measurements (mainly glaciological in-situ local annual surface mass balance measurements) and geodetic air or spaceborne elevation change data (a surface elevation/ice thickness change or an ice volume change over time). These data are converted into an averaged specific mass balance value (i.e. mostly reported with units of meter water equivalent and often shortened to m w.e.) for an individual glacier. Afterwards, the data are submitted to the World Glacier Monitoring Service (WGMS). Further processing of the data results in a glacier mass change product reported over a 0.5° global grid dating back until the 1975-76 hydrological year. Each grid cell therefore contains a time series of total glacier mass change (in Gt yr⁻¹) or mass balance data (in m w.e. yr⁻¹) of all glaciers within the specific grid cell. In this notebook, we use version WGMS-FOG-2023-09. For a more detailed description of the data acquisition and processing methods, we refer to the [documentation on the CDS](https://cds.climate.copernicus.eu/datasets/derived-gridded-glacier-mass-change?tab=overview) and the [ECMWF Confluence Wiki](https://confluence.ecmwf.int/pages/viewpage.action?pageId=355349383) (Copernicus Knowledge Base).\n\nIt is important to note that the glaciers with an annual glaciological sample (ca. 500 glaciers) form the basis for the determination of annual mass changes from the remaining glaciers through a complex algorithm of spatial/temporal interpolation and (area-weighted) averaging of these mass change data [[2](https://essd.copernicus.org/preprints/essd-2024-323/)]. Hence, not all glaciers in the dataset exhibit a time series of directly measured annual in-situ mass balance observations, but most glaciers in the dataset are either unobserved or only have (limited and mostly multi-annual) geodetic mass change data available. Glacier mass change data with these geodetic samples are also more prone to uncertainties in general, for example due to uncertainties related to volume to mass conversions."} {"chunk_id": "satellite_derived-gridded-glacier-mass-change_trend-assessment_q02__a2bcf7c70da9", "report_id": "satellite_derived-gridded-glacier-mass-change_trend-assessment_q02", "dataset_id": "derived-gridded-glacier-mass-change", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q02", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Glacier mass change data from satellite and in-situ observations: resolution and coverage for trend analysis > Methodology > Structure and (sub)sections", "title": "Glacier mass change data from satellite and in-situ observations: resolution and coverage for trend analysis", "chunk_index": 4, "token_count": 197, "text_raw": "**[](section-1)**\n* [](section-1-1)\n* [](section-1-2)\n* [](section-1-3)\n\n**[](section-2)**\n* [](section-2-1)\n* [](section-2-2)\n* [](section-2-3)\n* [](section-2-4)\n\n**[](section-3)**\n* [](section-3-1)\n* [](section-3-2)\n* [](section-3-3)\n\n**[](section-4)**", "text_with_prefix": "EQC Quality Assessment: \"Glacier mass change data from satellite and in-situ observations: resolution and coverage for trend analysis\"\nDataset: derived-gridded-glacier-mass-change [CDS]\nAspect: trend-assessment_q02 | Category: Satellite_ECVs\nSection: Glacier mass change data from satellite and in-situ observations: resolution and coverage for trend analysis > Methodology > Structure and (sub)sections\n---\n**[](section-1)**\n* [](section-1-1)\n* [](section-1-2)\n* [](section-1-3)\n\n**[](section-2)**\n* [](section-2-1)\n* [](section-2-2)\n* [](section-2-3)\n* [](section-2-4)\n\n**[](section-3)**\n* [](section-3-1)\n* [](section-3-2)\n* [](section-3-3)\n\n**[](section-4)**"} {"chunk_id": "satellite_derived-gridded-glacier-mass-change_trend-assessment_q02__9c3be6f86ad0", "report_id": "satellite_derived-gridded-glacier-mass-change_trend-assessment_q02", "dataset_id": "derived-gridded-glacier-mass-change", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q02", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Glacier mass change data from satellite and in-situ observations: resolution and coverage for trend analysis > Analysis and results > 1. Data preparation and processing > 1.2 Define request and download data", "title": "Glacier mass change data from satellite and in-situ observations: resolution and coverage for trend analysis", "chunk_index": 5, "token_count": 246, "text_raw": "Then we define the parameters, i.e. for which years we want the glacier mass change data to be downloaded:\n\nDefine a function to calculate the sum, linear trend and acceleration:\n\nDefine some functions to calculate the sum, linear trends and quadratic trends\n\nThen we define requests for download from the CDS and download and transform the glacier mass change data.\n\nDownload glacier mass change data\nCustomize some attributes\n\n```text\nDownloading and handling glacier mass change data...\n```\n\n```text\n100%|██████████| 1/1 [00:00<00:00, 6.46it/s]\n```\n\n```text\nDownloading and data handling done.\n```\n\n(section-1-3)=", "text_with_prefix": "EQC Quality Assessment: \"Glacier mass change data from satellite and in-situ observations: resolution and coverage for trend analysis\"\nDataset: derived-gridded-glacier-mass-change [CDS]\nAspect: trend-assessment_q02 | Category: Satellite_ECVs\nSection: Glacier mass change data from satellite and in-situ observations: resolution and coverage for trend analysis > Analysis and results > 1. Data preparation and processing > 1.2 Define request and download data\n---\nThen we define the parameters, i.e. for which years we want the glacier mass change data to be downloaded:\n\nDefine a function to calculate the sum, linear trend and acceleration:\n\nDefine some functions to calculate the sum, linear trends and quadratic trends\n\nThen we define requests for download from the CDS and download and transform the glacier mass change data.\n\nDownload glacier mass change data\nCustomize some attributes\n\n```text\nDownloading and handling glacier mass change data...\n```\n\n```text\n100%|██████████| 1/1 [00:00<00:00, 6.46it/s]\n```\n\n```text\nDownloading and data handling done.\n```\n\n(section-1-3)="} {"chunk_id": "satellite_derived-gridded-glacier-mass-change_trend-assessment_q02__5da0bfeda96b", "report_id": "satellite_derived-gridded-glacier-mass-change_trend-assessment_q02", "dataset_id": "derived-gridded-glacier-mass-change", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q02", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Glacier mass change data from satellite and in-situ observations: resolution and coverage for trend analysis > Analysis and results > 1. Data preparation and processing > 1.3 Display and inspect data", "title": "Glacier mass change data from satellite and in-situ observations: resolution and coverage for trend analysis", "chunk_index": 6, "token_count": 897, "text_raw": "Let us inspect the data:\n\n```text\n Size: 487MB\nDimensions: (time: 47, latitude: 360, longitude: 720)\nCoordinates:\n * time (time) int64 376B 1976 1977 1978 ... 2020 2021 2022\n * latitude (latitude) float64 3kB 89.75 89.25 ... -89.75\n * longitude (longitude) float64 6kB -179.8 -179.2 ... 179.8\nData variables:\n glacier_mass_change_gt (time, latitude, longitude) float64 97MB dask.array\n glacier_mass_change_mwe (time, latitude, longitude) float64 97MB dask.array\n glacier_area_km2 (time, latitude, longitude) float64 97MB dask.array\n uncertainty_gt (time, latitude, longitude) float64 97MB dask.array\n uncertainty_mwe (time, latitude, longitude) float64 97MB dask.array\nAttributes:\n title: Global gridded annual glacier mass changes\n data_version: version-wgms-fog-2023-09\n project: Copernicus Climate Change Service (C3S) Essential ...\n institution: World Glacier Monitoring Service - Geography Depar...\n created_by: Dr. Ines Dussaillant - ines.dussaillant@geo.uzh.ch\n references: Fluctuation of Glaciers (FoG) database version wgm...\n citation: Dussaillant, I., Bannwart, J., Paul, F., Zemp, M. ...\n conventions: CF Version CF-1.8\n dataset_description: Horizontal resolution: 0.5° (latitude - longitude)...\n dataset_limitations: Grid-point artefact in polar regions: see Algorith...\n dataset_improvements: Improvements of product version WGMS-FOG-2023-09 w...\n comments: Conversions between annual grid point mass change ...\n```\n\nIt is a gridded dataset at a 0.5 by 0.5 degree spatial resolution containing annual values of the total glacier mass change (in Gt yr$^{-1}$) of a grid cell (`glacier_mass_change_gt`) and its uncertainty (`uncertainty_gt`) since the 1975-76 hydrological year. Mass changes (mass balances in this case) and their uncertainty can also be extracted in units of m w.e. yr$^{-1}$ (`glacier_mass_change_mwe` and `uncertainty_mwe`). The parameterized absolute glacier surface area is furthermore also available in this dataset by the variable `glacier_area_km2`. For conversion purposes, glacier mass changes (in Gt yr$^{-1}$) can be converted into glacier mass balances (in m w.e. yr$^{-1}$) by dividing the mass changes (in Gt yr$^{-1}$ multiplied by $1*10^{12}$ to get values in kg yr$^{-1}$) by the product of the density of water (1000 kg m$^{-3}$) and the glacier area (in km$^{2}$ multiplied by $1*10^6$ to get values in m$^2$).\n\n(section-2)=", "text_with_prefix": "EQC Quality Assessment: \"Glacier mass change data from satellite and in-situ observations: resolution and coverage for trend analysis\"\nDataset: derived-gridded-glacier-mass-change [CDS]\nAspect: trend-assessment_q02 | Category: Satellite_ECVs\nSection: Glacier mass change data from satellite and in-situ observations: resolution and coverage for trend analysis > Analysis and results > 1. Data preparation and processing > 1.3 Display and inspect data\n---\nLet us inspect the data:\n\n```text\n Size: 487MB\nDimensions: (time: 47, latitude: 360, longitude: 720)\nCoordinates:\n * time (time) int64 376B 1976 1977 1978 ... 2020 2021 2022\n * latitude (latitude) float64 3kB 89.75 89.25 ... -89.75\n * longitude (longitude) float64 6kB -179.8 -179.2 ... 179.8\nData variables:\n glacier_mass_change_gt (time, latitude, longitude) float64 97MB dask.array\n glacier_mass_change_mwe (time, latitude, longitude) float64 97MB dask.array\n glacier_area_km2 (time, latitude, longitude) float64 97MB dask.array\n uncertainty_gt (time, latitude, longitude) float64 97MB dask.array\n uncertainty_mwe (time, latitude, longitude) float64 97MB dask.array\nAttributes:\n title: Global gridded annual glacier mass changes\n data_version: version-wgms-fog-2023-09\n project: Copernicus Climate Change Service (C3S) Essential ...\n institution: World Glacier Monitoring Service - Geography Depar...\n created_by: Dr. Ines Dussaillant - ines.dussaillant@geo.uzh.ch\n references: Fluctuation of Glaciers (FoG) database version wgm...\n citation: Dussaillant, I., Bannwart, J., Paul, F., Zemp, M. ...\n conventions: CF Version CF-1.8\n dataset_description: Horizontal resolution: 0.5° (latitude - longitude)...\n dataset_limitations: Grid-point artefact in polar regions: see Algorith...\n dataset_improvements: Improvements of product version WGMS-FOG-2023-09 w...\n comments: Conversions between annual grid point mass change ...\n```\n\nIt is a gridded dataset at a 0.5 by 0.5 degree spatial resolution containing annual values of the total glacier mass change (in Gt yr$^{-1}$) of a grid cell (`glacier_mass_change_gt`) and its uncertainty (`uncertainty_gt`) since the 1975-76 hydrological year. Mass changes (mass balances in this case) and their uncertainty can also be extracted in units of m w.e. yr$^{-1}$ (`glacier_mass_change_mwe` and `uncertainty_mwe`). The parameterized absolute glacier surface area is furthermore also available in this dataset by the variable `glacier_area_km2`. For conversion purposes, glacier mass changes (in Gt yr$^{-1}$) can be converted into glacier mass balances (in m w.e. yr$^{-1}$) by dividing the mass changes (in Gt yr$^{-1}$ multiplied by $1*10^{12}$ to get values in kg yr$^{-1}$) by the product of the density of water (1000 kg m$^{-3}$) and the glacier area (in km$^{2}$ multiplied by $1*10^6$ to get values in m$^2$).\n\n(section-2)="} {"chunk_id": "satellite_derived-gridded-glacier-mass-change_trend-assessment_q02__8aeeb9aff76a", "report_id": "satellite_derived-gridded-glacier-mass-change_trend-assessment_q02", "dataset_id": "derived-gridded-glacier-mass-change", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q02", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Glacier mass change data from satellite and in-situ observations: resolution and coverage for trend analysis > Analysis and results > 2. Analysis of spatio-temporal resolution and extent of glacier mass changes > 2.1 Dataset attributes: spatial and temporal resolution", "title": "Glacier mass change data from satellite and in-situ observations: resolution and coverage for trend analysis", "chunk_index": 7, "token_count": 425, "text_raw": "Let us print some attributes of the dataset to reveal its spatial and temporal resolution:\n\n```text\nHorizontal resolution: 0.5° (latitude - longitude), GCS_WGS_1984Temporal resolution: Annual, hydrological yearTemporal coverage: 1975/76-2021/22Observational sample: 96% of world glaciers with valid observations\n```\n\nThe dataset has a spatial resolution of 0.5 by 0.5 degrees and a temporal resolution of 1 hydrological year. Although no specific thresholds are given for the spatial resolution of glacier mass change data by GCOS, the annual temporal resolution satisfies the minimum threshold [[4](https://library.wmo.int/records/item/58111-the-2022-gcos-ecvs-requirements-gcos-245)]. The data are not useable to detect mass changes at a finer temporal (e.g. short-term monthly or seasonal changes in glacier mass) or spatial (for individual glaciers) resolution, potentially limiting applications that involve individual glaciers. Note that the usage of a 0.5° grid also implies that not every grid cell has the same absolute surface area. This results in the fact that grid cells are smaller towards the poles, which can result in individual glaciers being larger than the grid surface. This, in turn, may give rise to artificial glacier mass change artefacts in those regions. This would, however, affect only glacier mass changes at local scales (pixels), and not those at the regional or regional spatial scales.\n\n(section-2-2)=", "text_with_prefix": "EQC Quality Assessment: \"Glacier mass change data from satellite and in-situ observations: resolution and coverage for trend analysis\"\nDataset: derived-gridded-glacier-mass-change [CDS]\nAspect: trend-assessment_q02 | Category: Satellite_ECVs\nSection: Glacier mass change data from satellite and in-situ observations: resolution and coverage for trend analysis > Analysis and results > 2. Analysis of spatio-temporal resolution and extent of glacier mass changes > 2.1 Dataset attributes: spatial and temporal resolution\n---\nLet us print some attributes of the dataset to reveal its spatial and temporal resolution:\n\n```text\nHorizontal resolution: 0.5° (latitude - longitude), GCS_WGS_1984Temporal resolution: Annual, hydrological yearTemporal coverage: 1975/76-2021/22Observational sample: 96% of world glaciers with valid observations\n```\n\nThe dataset has a spatial resolution of 0.5 by 0.5 degrees and a temporal resolution of 1 hydrological year. Although no specific thresholds are given for the spatial resolution of glacier mass change data by GCOS, the annual temporal resolution satisfies the minimum threshold [[4](https://library.wmo.int/records/item/58111-the-2022-gcos-ecvs-requirements-gcos-245)]. The data are not useable to detect mass changes at a finer temporal (e.g. short-term monthly or seasonal changes in glacier mass) or spatial (for individual glaciers) resolution, potentially limiting applications that involve individual glaciers. Note that the usage of a 0.5° grid also implies that not every grid cell has the same absolute surface area. This results in the fact that grid cells are smaller towards the poles, which can result in individual glaciers being larger than the grid surface. This, in turn, may give rise to artificial glacier mass change artefacts in those regions. This would, however, affect only glacier mass changes at local scales (pixels), and not those at the regional or regional spatial scales.\n\n(section-2-2)="} {"chunk_id": "satellite_derived-gridded-glacier-mass-change_trend-assessment_q02__b9abb6374b39", "report_id": "satellite_derived-gridded-glacier-mass-change_trend-assessment_q02", "dataset_id": "derived-gridded-glacier-mass-change", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q02", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Glacier mass change data from satellite and in-situ observations: resolution and coverage for trend analysis > Analysis and results > 2. Analysis of spatio-temporal resolution and extent of glacier mass changes > 2.2 Temporal extent/coverage", "title": "Glacier mass change data from satellite and in-situ observations: resolution and coverage for trend analysis", "chunk_index": 8, "token_count": 539, "text_raw": "Let us examine the temporal extent of the dataset:\n\n```text\nThe temporal extent of the glacier mass change dataset is 47 years.\n```\n\nTo determine whether this temporal extent of the glacier mass change dataset is sufficient to capture reliable temporal trends in global glacier mass changes and to use these trends as indicators of climatic changes, we turn to the literature. The Intergovernmental Panel on Climate Change (IPCC) and World Meteorological Orginization (WMO) often use 30 years as a standard period for climate normals and trend analysis to ensure that the analysis captures meaningful climatic changes rather than short-term (intra/interannual) fluctuations. We therefore consider these guidelines to be likewise applicative for glacier mass changes. When measured over a long period (> 30 years), trends in glacier mass changes or mass balances can therefore be considered a clear indicator of global climate change. Longer periods, as is here the case, provide even more robust trend estimates and reduce the influence of short-term variability.\n\nWe can also determine the number of years that hold non-NaN mass change data for each pixel. If the pixel exhibits complete temporal coverage, the value should equal the number of years in the time dimension `time` of the dataset. Let us have this number quantified:\n\nDetermine where values are present\nCheck where the number of values equals the time dimension of the dataset\nCount how many non-NaN values there are\n\n```text\nThe number of pixels that have a time series of valid glacier mass change data of 47 years, which is the total number of years in the dataset, is 100.00%.\n```\n\nConcerning temporal aspects, the dataset thus offers data at regular and consistently spaced annual intervals that are sufficiently long in time to capture long-term glacier mass change trends at a local (individual pixels), regional or global scale. There are no gaps in the temporal data that could affect a trend analysis. In other words, the dataset exhibits a consistently complete temporal resolution and coverage, with no data gaps, which allows for reliable quantifications of glacier mass change trends.\n\n(section-2-3)=", "text_with_prefix": "EQC Quality Assessment: \"Glacier mass change data from satellite and in-situ observations: resolution and coverage for trend analysis\"\nDataset: derived-gridded-glacier-mass-change [CDS]\nAspect: trend-assessment_q02 | Category: Satellite_ECVs\nSection: Glacier mass change data from satellite and in-situ observations: resolution and coverage for trend analysis > Analysis and results > 2. Analysis of spatio-temporal resolution and extent of glacier mass changes > 2.2 Temporal extent/coverage\n---\nLet us examine the temporal extent of the dataset:\n\n```text\nThe temporal extent of the glacier mass change dataset is 47 years.\n```\n\nTo determine whether this temporal extent of the glacier mass change dataset is sufficient to capture reliable temporal trends in global glacier mass changes and to use these trends as indicators of climatic changes, we turn to the literature. The Intergovernmental Panel on Climate Change (IPCC) and World Meteorological Orginization (WMO) often use 30 years as a standard period for climate normals and trend analysis to ensure that the analysis captures meaningful climatic changes rather than short-term (intra/interannual) fluctuations. We therefore consider these guidelines to be likewise applicative for glacier mass changes. When measured over a long period (> 30 years), trends in glacier mass changes or mass balances can therefore be considered a clear indicator of global climate change. Longer periods, as is here the case, provide even more robust trend estimates and reduce the influence of short-term variability.\n\nWe can also determine the number of years that hold non-NaN mass change data for each pixel. If the pixel exhibits complete temporal coverage, the value should equal the number of years in the time dimension `time` of the dataset. Let us have this number quantified:\n\nDetermine where values are present\nCheck where the number of values equals the time dimension of the dataset\nCount how many non-NaN values there are\n\n```text\nThe number of pixels that have a time series of valid glacier mass change data of 47 years, which is the total number of years in the dataset, is 100.00%.\n```\n\nConcerning temporal aspects, the dataset thus offers data at regular and consistently spaced annual intervals that are sufficiently long in time to capture long-term glacier mass change trends at a local (individual pixels), regional or global scale. There are no gaps in the temporal data that could affect a trend analysis. In other words, the dataset exhibits a consistently complete temporal resolution and coverage, with no data gaps, which allows for reliable quantifications of glacier mass change trends.\n\n(section-2-3)="} {"chunk_id": "satellite_derived-gridded-glacier-mass-change_trend-assessment_q02__c5c23e73d9e1", "report_id": "satellite_derived-gridded-glacier-mass-change_trend-assessment_q02", "dataset_id": "derived-gridded-glacier-mass-change", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q02", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Glacier mass change data from satellite and in-situ observations: resolution and coverage for trend analysis > Analysis and results > 2. Analysis of spatio-temporal resolution and extent of glacier mass changes > 2.3 Spatial extent/coverage", "title": "Glacier mass change data from satellite and in-situ observations: resolution and coverage for trend analysis", "chunk_index": 9, "token_count": 501, "text_raw": "To further assess the spatial coverage of the data, we can make use of the `glacier_area_km2` variable within the dataset. Let us plot the global glacier area over time to get an idea of how this variable looks:\n\nMake the plot\n\n*Figure 1. Total global glacier area over time in the glacier mass change dataset, which results from a parameterization of estimated regional glacier area changes [[3](https://doi.org/10.1038/s41586-019-1071-0)]).*\n\nAs can be seen from the plot above, the glacier area results from a linear decrease over time. It is hence not derived from in-situ and/or remote sensing data, but rather inserted into the data as a parameterization (i.e. a linear fit to estimated regional glacier area changes, of which it is assumed that these glacier area change trends remain unchanged over time [[3](https://doi.org/10.1038/s41586-019-1071-0)]). For proper assessment, we can compare the value around 2000 CE to that of the \"[Glaciers distribution data from the Randolph Glacier Inventory (RGI) for year 2000](https://cds.climate.copernicus.eu/datasets/insitu-glaciers-extent?tab=overview)\" dataset, which is 746088.28 km$^2$ from the vector (shape file) product in RGI v6.0 [[5](https://www.glims.org/RGI/randolph60.html)]:\n\nCalculate the total glacier area for the year 2000\nDetermine the percentage difference\n\n```text\nThe total global glacier area in the glacier mass change dataset in 2000 CE is 745153.82 km², which is 99.87% of the glacier area in the RGIv6.0 dataset.\n```\n\n(section-2-4)=", "text_with_prefix": "EQC Quality Assessment: \"Glacier mass change data from satellite and in-situ observations: resolution and coverage for trend analysis\"\nDataset: derived-gridded-glacier-mass-change [CDS]\nAspect: trend-assessment_q02 | Category: Satellite_ECVs\nSection: Glacier mass change data from satellite and in-situ observations: resolution and coverage for trend analysis > Analysis and results > 2. Analysis of spatio-temporal resolution and extent of glacier mass changes > 2.3 Spatial extent/coverage\n---\nTo further assess the spatial coverage of the data, we can make use of the `glacier_area_km2` variable within the dataset. Let us plot the global glacier area over time to get an idea of how this variable looks:\n\nMake the plot\n\n*Figure 1. Total global glacier area over time in the glacier mass change dataset, which results from a parameterization of estimated regional glacier area changes [[3](https://doi.org/10.1038/s41586-019-1071-0)]).*\n\nAs can be seen from the plot above, the glacier area results from a linear decrease over time. It is hence not derived from in-situ and/or remote sensing data, but rather inserted into the data as a parameterization (i.e. a linear fit to estimated regional glacier area changes, of which it is assumed that these glacier area change trends remain unchanged over time [[3](https://doi.org/10.1038/s41586-019-1071-0)]). For proper assessment, we can compare the value around 2000 CE to that of the \"[Glaciers distribution data from the Randolph Glacier Inventory (RGI) for year 2000](https://cds.climate.copernicus.eu/datasets/insitu-glaciers-extent?tab=overview)\" dataset, which is 746088.28 km$^2$ from the vector (shape file) product in RGI v6.0 [[5](https://www.glims.org/RGI/randolph60.html)]:\n\nCalculate the total glacier area for the year 2000\nDetermine the percentage difference\n\n```text\nThe total global glacier area in the glacier mass change dataset in 2000 CE is 745153.82 km², which is 99.87% of the glacier area in the RGIv6.0 dataset.\n```\n\n(section-2-4)="} {"chunk_id": "satellite_derived-gridded-glacier-mass-change_trend-assessment_q02__54eed7d71b53", "report_id": "satellite_derived-gridded-glacier-mass-change_trend-assessment_q02", "dataset_id": "derived-gridded-glacier-mass-change", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q02", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Glacier mass change data from satellite and in-situ observations: resolution and coverage for trend analysis > Analysis and results > 2. Analysis of spatio-temporal resolution and extent of glacier mass changes > 2.4 Sampling density", "title": "Glacier mass change data from satellite and in-situ observations: resolution and coverage for trend analysis", "chunk_index": 10, "token_count": 521, "text_raw": "Quantitative details about the sampling density (i.e. the number of sampled glaciers per grid point) are not given in the dataset. Nevertheless, a potential scarce observational sample of annual in-situ glaciological mass balance observations can introduce significant biases and uncertainties into the analysis of long-term trends in glacier mass changes. It is thus not possible to determine whether the data are densely sampled enough to ensure that the observed trends are reliable and not affected by sparse data. An indication of the global distribution of glaciological mass change and geodetic elevation change observations from the WGMS Fluctuations of Glaciers database is, however, available from the [ECMWF Confluence Wiki](https://confluence.ecmwf.int/pages/viewpage.action?pageId=355349383) (see figure below). While the glaciological sample covers ca. 500 glaciers only, the multi-annual to decadal geodetic sample covers ca. 200,000 glaciers around the globe (in total ca. 96% of the glaciers present in the RGIv6.0 dataset). The other remaining glaciers are unobserved by any means and require the complex algorithm to assign them a mass change value based on the behavior of neighboring observed glaciers.\n\n![alternatvie text](https://confluence.ecmwf.int/download/attachments/400511869/worddav17d9d5fbc2dc155690254b907459af22.png?version=1&modificationDate=1713359738989&api=v2)\n\n*Figure 2. Distribution of glacier mass change records from the glaciological (red crosses) and geodetic (blue dots) samples over the 19 RGI 1st order regions (shown as black boxes with region number as label) used in the glacier mass change dataset. From: [ECMWF Confluence Wiki](https://confluence.ecmwf.int/pages/viewpage.action?pageId=355349383) (Copernicus Knowledge Base).*\n\n(section-3)=", "text_with_prefix": "EQC Quality Assessment: \"Glacier mass change data from satellite and in-situ observations: resolution and coverage for trend analysis\"\nDataset: derived-gridded-glacier-mass-change [CDS]\nAspect: trend-assessment_q02 | Category: Satellite_ECVs\nSection: Glacier mass change data from satellite and in-situ observations: resolution and coverage for trend analysis > Analysis and results > 2. Analysis of spatio-temporal resolution and extent of glacier mass changes > 2.4 Sampling density\n---\nQuantitative details about the sampling density (i.e. the number of sampled glaciers per grid point) are not given in the dataset. Nevertheless, a potential scarce observational sample of annual in-situ glaciological mass balance observations can introduce significant biases and uncertainties into the analysis of long-term trends in glacier mass changes. It is thus not possible to determine whether the data are densely sampled enough to ensure that the observed trends are reliable and not affected by sparse data. An indication of the global distribution of glaciological mass change and geodetic elevation change observations from the WGMS Fluctuations of Glaciers database is, however, available from the [ECMWF Confluence Wiki](https://confluence.ecmwf.int/pages/viewpage.action?pageId=355349383) (see figure below). While the glaciological sample covers ca. 500 glaciers only, the multi-annual to decadal geodetic sample covers ca. 200,000 glaciers around the globe (in total ca. 96% of the glaciers present in the RGIv6.0 dataset). The other remaining glaciers are unobserved by any means and require the complex algorithm to assign them a mass change value based on the behavior of neighboring observed glaciers.\n\n![alternatvie text](https://confluence.ecmwf.int/download/attachments/400511869/worddav17d9d5fbc2dc155690254b907459af22.png?version=1&modificationDate=1713359738989&api=v2)\n\n*Figure 2. Distribution of glacier mass change records from the glaciological (red crosses) and geodetic (blue dots) samples over the 19 RGI 1st order regions (shown as black boxes with region number as label) used in the glacier mass change dataset. From: [ECMWF Confluence Wiki](https://confluence.ecmwf.int/pages/viewpage.action?pageId=355349383) (Copernicus Knowledge Base).*\n\n(section-3)="} {"chunk_id": "satellite_derived-gridded-glacier-mass-change_trend-assessment_q02__60f224c5af71", "report_id": "satellite_derived-gridded-glacier-mass-change_trend-assessment_q02", "dataset_id": "derived-gridded-glacier-mass-change", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q02", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Glacier mass change data from satellite and in-situ observations: resolution and coverage for trend analysis > Analysis and results > 3. Glacier mass change trends > 3.1 Spatial distribution of linear mass change trends", "title": "Glacier mass change data from satellite and in-situ observations: resolution and coverage for trend analysis", "chunk_index": 11, "token_count": 447, "text_raw": "Let us now express the time series of glacier mass changes at each pixel in the form of linear trends and plot them on a world map:\n\nPlot the gridded data\n\n*
Figure 3. Spatial distribution of linear mass change trends (Gt/yr) for each pixel in the glacier mass change dataset.
*\n\nProminent mass loss is evident in regions such as Alaska, the western United States, the Andes, Svalbard, the Alps, and the Himalayas, where red tones (i.e. negative linear trends) dominate. In contrast, areas with mass gain are minimal, shown by isolated blue patches (i.e. positive linear trends). The plot of linear trends above underscores the persistent decrease in glacier mass over recent decades. Let us now check the statistical significance of these trends at $\\alpha$ = 0.05:\n\nPlot the gridded data\nDefine the boundaries for the colorbar\nPlot the data\n\n*
Figure 4. Statistical significance (alpha=0.05) of linear mass change trends for each pixel in the glacier mass change dataset.
*\n\nLet us quantify the number of pixels where a statistically significant negative linear trend prevails:\n\nCheck for pixels with negative trend\nCheck for pixels with statistically significant trend\nCount how many there are\n\n```text\nThe number of pixels with both a negative and statistically significant trend is 3590, which is 76.30% of the total pixels that hold glacier mass change data.\n```\n\nThe above analysis indicates that the majority of the derived linear mass change trends are negative and statistically significant, adding further credibility to the glacier mass change dataset.\n\n(section-3-2)=", "text_with_prefix": "EQC Quality Assessment: \"Glacier mass change data from satellite and in-situ observations: resolution and coverage for trend analysis\"\nDataset: derived-gridded-glacier-mass-change [CDS]\nAspect: trend-assessment_q02 | Category: Satellite_ECVs\nSection: Glacier mass change data from satellite and in-situ observations: resolution and coverage for trend analysis > Analysis and results > 3. Glacier mass change trends > 3.1 Spatial distribution of linear mass change trends\n---\nLet us now express the time series of glacier mass changes at each pixel in the form of linear trends and plot them on a world map:\n\nPlot the gridded data\n\n*
Figure 3. Spatial distribution of linear mass change trends (Gt/yr) for each pixel in the glacier mass change dataset.
*\n\nProminent mass loss is evident in regions such as Alaska, the western United States, the Andes, Svalbard, the Alps, and the Himalayas, where red tones (i.e. negative linear trends) dominate. In contrast, areas with mass gain are minimal, shown by isolated blue patches (i.e. positive linear trends). The plot of linear trends above underscores the persistent decrease in glacier mass over recent decades. Let us now check the statistical significance of these trends at $\\alpha$ = 0.05:\n\nPlot the gridded data\nDefine the boundaries for the colorbar\nPlot the data\n\n*
Figure 4. Statistical significance (alpha=0.05) of linear mass change trends for each pixel in the glacier mass change dataset.
*\n\nLet us quantify the number of pixels where a statistically significant negative linear trend prevails:\n\nCheck for pixels with negative trend\nCheck for pixels with statistically significant trend\nCount how many there are\n\n```text\nThe number of pixels with both a negative and statistically significant trend is 3590, which is 76.30% of the total pixels that hold glacier mass change data.\n```\n\nThe above analysis indicates that the majority of the derived linear mass change trends are negative and statistically significant, adding further credibility to the glacier mass change dataset.\n\n(section-3-2)="} {"chunk_id": "satellite_derived-gridded-glacier-mass-change_trend-assessment_q02__6e04c011d7f8", "report_id": "satellite_derived-gridded-glacier-mass-change_trend-assessment_q02", "dataset_id": "derived-gridded-glacier-mass-change", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q02", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Glacier mass change data from satellite and in-situ observations: resolution and coverage for trend analysis > Analysis and results > 3. Glacier mass change trends > 3.2 Trends in global total glacier mass changes", "title": "Glacier mass change data from satellite and in-situ observations: resolution and coverage for trend analysis", "chunk_index": 12, "token_count": 788, "text_raw": "In the following section, we plot the global annual (and cumulative) glacier mass change over time. For the annual values, we therefore sum the gridded mass change product over the entire spatial domain for each individual year to get spatially summed values in Gt yr⁻¹:\n\n$\\Delta M_i \n$\n[Gt yr⁻¹]\n$ = \\sum\\limits^{x,y}\\Delta {M_{x,y}}$\n\nwhere $\\Delta {M_{x,y}}$ is the glacier mass change (in Gt yr⁻¹) at pixel $x,y$ during a certain year $i$.\n\nThe data can also be plotted in a cumulative way:\n\n$\n{\\Delta M} \n$\n[Gt]\n$\n= \\sum\\limits_{i={1976}}^{{{1976+n-1}}} (\\Delta M_i)\n$\n\nwith $n$ the total number of years in the time series.\n\nThis results in the following plot:\n\nCompute cumulative glacier mass change/balance and area\n\n*
Figure 5. Annual (left) and cumulative (right) global glacier mass changes in the glacier mass change dataset.
*\n\nThe values in the plots above align well with findings in the literature (e.g. [[9](https://doi.org/10.1038/s41586-024-08545-z)]). We can also plot the annual global glacier mass changes in the style of the popular climate [warming stripes](https://showyourstripes.info) for a better visual representation:\n\nConfigure the figure\nDefine a red (negative) to blue (positive) colorscale\nSaturate colors\nCreate a colored rectangle for each year\n\n*Figure 6. Annual global glacier mass changes from the glacier mass change dataset expressed in the form of 'warming stripes'.*\n\nAt last, we can calculate the linear and quadratic trends of the global glacier mass change time series:\n\nExtract time and glacier data\nExtract most recent 30 years\nCalculate linear trend\ncoeffs[0] is the slope, coeffs[1] is the intercept\n\n```text\nThe linear trend of the global glacier mass changes between 1975-1976 and 2021-2022 is -167.30 Gt yr⁻¹. During the most recent 30 years, the trend is -257.64 Gt yr⁻¹.\n```\n\nLet us have this plotted:\n\nPlot the data\n\n*Figure 7. Cumulative global glacier mass changes from the glacier mass change dataset and the corresponding linear trend.*\n\nNow for the acceleration of the global glacier mass changes:\n\nExtract time and glacier data\nCalculate quadratic trend\nExtract most recent 30 years\nCalculate quadratic trend\ncoeffs[0] is the quadratic term, coeffs[1] is the linear term, coeffs[2] is the intercept\n\n```text\nThe acceleration of the global glacier mass changes between 1975-1976 and 2021-2022 is -10.52 Gt yr⁻². During the most recent 30 years, the acceleration is -7.60 Gt yr⁻².\n```\n\nLet us have the quadratic trend plotted:\n\nPlot the data\n\n*Figure 8. Cumulative global glacier mass changes from the glacier mass change dataset and the corresponding quadratic trend (acceleration).*\n\n(section-3-3)=", "text_with_prefix": "EQC Quality Assessment: \"Glacier mass change data from satellite and in-situ observations: resolution and coverage for trend analysis\"\nDataset: derived-gridded-glacier-mass-change [CDS]\nAspect: trend-assessment_q02 | Category: Satellite_ECVs\nSection: Glacier mass change data from satellite and in-situ observations: resolution and coverage for trend analysis > Analysis and results > 3. Glacier mass change trends > 3.2 Trends in global total glacier mass changes\n---\nIn the following section, we plot the global annual (and cumulative) glacier mass change over time. For the annual values, we therefore sum the gridded mass change product over the entire spatial domain for each individual year to get spatially summed values in Gt yr⁻¹:\n\n$\\Delta M_i \n$\n[Gt yr⁻¹]\n$ = \\sum\\limits^{x,y}\\Delta {M_{x,y}}$\n\nwhere $\\Delta {M_{x,y}}$ is the glacier mass change (in Gt yr⁻¹) at pixel $x,y$ during a certain year $i$.\n\nThe data can also be plotted in a cumulative way:\n\n$\n{\\Delta M} \n$\n[Gt]\n$\n= \\sum\\limits_{i={1976}}^{{{1976+n-1}}} (\\Delta M_i)\n$\n\nwith $n$ the total number of years in the time series.\n\nThis results in the following plot:\n\nCompute cumulative glacier mass change/balance and area\n\n*
Figure 5. Annual (left) and cumulative (right) global glacier mass changes in the glacier mass change dataset.
*\n\nThe values in the plots above align well with findings in the literature (e.g. [[9](https://doi.org/10.1038/s41586-024-08545-z)]). We can also plot the annual global glacier mass changes in the style of the popular climate [warming stripes](https://showyourstripes.info) for a better visual representation:\n\nConfigure the figure\nDefine a red (negative) to blue (positive) colorscale\nSaturate colors\nCreate a colored rectangle for each year\n\n*Figure 6. Annual global glacier mass changes from the glacier mass change dataset expressed in the form of 'warming stripes'.*\n\nAt last, we can calculate the linear and quadratic trends of the global glacier mass change time series:\n\nExtract time and glacier data\nExtract most recent 30 years\nCalculate linear trend\ncoeffs[0] is the slope, coeffs[1] is the intercept\n\n```text\nThe linear trend of the global glacier mass changes between 1975-1976 and 2021-2022 is -167.30 Gt yr⁻¹. During the most recent 30 years, the trend is -257.64 Gt yr⁻¹.\n```\n\nLet us have this plotted:\n\nPlot the data\n\n*Figure 7. Cumulative global glacier mass changes from the glacier mass change dataset and the corresponding linear trend.*\n\nNow for the acceleration of the global glacier mass changes:\n\nExtract time and glacier data\nCalculate quadratic trend\nExtract most recent 30 years\nCalculate quadratic trend\ncoeffs[0] is the quadratic term, coeffs[1] is the linear term, coeffs[2] is the intercept\n\n```text\nThe acceleration of the global glacier mass changes between 1975-1976 and 2021-2022 is -10.52 Gt yr⁻². During the most recent 30 years, the acceleration is -7.60 Gt yr⁻².\n```\n\nLet us have the quadratic trend plotted:\n\nPlot the data\n\n*Figure 8. Cumulative global glacier mass changes from the glacier mass change dataset and the corresponding quadratic trend (acceleration).*\n\n(section-3-3)="} {"chunk_id": "satellite_derived-gridded-glacier-mass-change_trend-assessment_q02__4c56f6631b6d", "report_id": "satellite_derived-gridded-glacier-mass-change_trend-assessment_q02", "dataset_id": "derived-gridded-glacier-mass-change", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q02", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Glacier mass change data from satellite and in-situ observations: resolution and coverage for trend analysis > Analysis and results > 3. Glacier mass change trends > 3.3 Comparison with theoretical models", "title": "Glacier mass change data from satellite and in-situ observations: resolution and coverage for trend analysis", "chunk_index": 13, "token_count": 672, "text_raw": "The linear trend of global glacier mass changes from 1975-1976 to 2021-2022 indicates a consistent decline. The quadratic trend shows an accelerating rate of these glacier mass losses. This suggests that the rate of global glacier mass loss has intensified over time, particularly during the more recent decades. The quadratic trend furthermore fits the observed data more closely than the linear trend, highlighting the increasing and non-linear impact of climate change on glacier mass changes. This acceleration was also noted in other glacier mass changes products [[6](https://doi.org/10.1038/s41586-021-03436-z), [7](https://doi.org/10.3389/feart.2019.00096)] and is also what can be expected from theory for a linear climate warming over time (e.g. [[8](https://doi.org/10.1007/s003820050222)]):\n\n$\\dfrac{dM}{dt} = -S_T A_{t_0} \\rho_i T'$\n\nwhere $S_T$ is the mass balance sensitivity to temperature changes (m yr$^{-1}$ K$^{-1}$), $A_{t_0}$ the initial glacier surface area (m$^2$), $\\rho_i$ the density of ice (kg m$^{-3}$) and $T'$ the temperature perturbation (K), which gives final units of kg yr$^{-1}$.\n\nBy assuming a linear temperature perturbation over time (i.e. $T' = \\alpha (t - t_0)$), the equation can be rewritten as (with $\\alpha$ in units of K yr$^{-1}$):\n\n$dM = -S_T A_{t_0} \\rho_i T' dt = -S_T A_{t_0} \\rho_i (\\alpha (t - t_0)) dt$\n\nor when integrating over a time period between $t$ and $t_0$:\n\n$ \\int\\limits^{t}_{t_0} dM = -S_T A_{t_0} \\rho_i \\int\\limits^{t}_{t_0} \\alpha (t - t_0)dt$\n\nand by finally assuming ${t_0}$ = 0 and $M(t_0) = 0$, we get:\n\n$M(t) = -\\dfrac{1}{2} S_T A_{t_0} \\rho_i \\alpha t^2$\n\nor $M(t) \\approx f(-t^2)$\n\nwhich thus quantifies a quadratic (accelerated) glacier mass decrease over time, as is also noted in the glacier mass change dataset, highlighting the non-linear temperature-driven effects on global glacier mass changes.\n\n(section-4)=", "text_with_prefix": "EQC Quality Assessment: \"Glacier mass change data from satellite and in-situ observations: resolution and coverage for trend analysis\"\nDataset: derived-gridded-glacier-mass-change [CDS]\nAspect: trend-assessment_q02 | Category: Satellite_ECVs\nSection: Glacier mass change data from satellite and in-situ observations: resolution and coverage for trend analysis > Analysis and results > 3. Glacier mass change trends > 3.3 Comparison with theoretical models\n---\nThe linear trend of global glacier mass changes from 1975-1976 to 2021-2022 indicates a consistent decline. The quadratic trend shows an accelerating rate of these glacier mass losses. This suggests that the rate of global glacier mass loss has intensified over time, particularly during the more recent decades. The quadratic trend furthermore fits the observed data more closely than the linear trend, highlighting the increasing and non-linear impact of climate change on glacier mass changes. This acceleration was also noted in other glacier mass changes products [[6](https://doi.org/10.1038/s41586-021-03436-z), [7](https://doi.org/10.3389/feart.2019.00096)] and is also what can be expected from theory for a linear climate warming over time (e.g. [[8](https://doi.org/10.1007/s003820050222)]):\n\n$\\dfrac{dM}{dt} = -S_T A_{t_0} \\rho_i T'$\n\nwhere $S_T$ is the mass balance sensitivity to temperature changes (m yr$^{-1}$ K$^{-1}$), $A_{t_0}$ the initial glacier surface area (m$^2$), $\\rho_i$ the density of ice (kg m$^{-3}$) and $T'$ the temperature perturbation (K), which gives final units of kg yr$^{-1}$.\n\nBy assuming a linear temperature perturbation over time (i.e. $T' = \\alpha (t - t_0)$), the equation can be rewritten as (with $\\alpha$ in units of K yr$^{-1}$):\n\n$dM = -S_T A_{t_0} \\rho_i T' dt = -S_T A_{t_0} \\rho_i (\\alpha (t - t_0)) dt$\n\nor when integrating over a time period between $t$ and $t_0$:\n\n$ \\int\\limits^{t}_{t_0} dM = -S_T A_{t_0} \\rho_i \\int\\limits^{t}_{t_0} \\alpha (t - t_0)dt$\n\nand by finally assuming ${t_0}$ = 0 and $M(t_0) = 0$, we get:\n\n$M(t) = -\\dfrac{1}{2} S_T A_{t_0} \\rho_i \\alpha t^2$\n\nor $M(t) \\approx f(-t^2)$\n\nwhich thus quantifies a quadratic (accelerated) glacier mass decrease over time, as is also noted in the glacier mass change dataset, highlighting the non-linear temperature-driven effects on global glacier mass changes.\n\n(section-4)="} {"chunk_id": "satellite_derived-gridded-glacier-mass-change_trend-assessment_q02__335a89fce9c6", "report_id": "satellite_derived-gridded-glacier-mass-change_trend-assessment_q02", "dataset_id": "derived-gridded-glacier-mass-change", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q02", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Glacier mass change data from satellite and in-situ observations: resolution and coverage for trend analysis > Analysis and results > 4. Short summary and take-home messages", "title": "Glacier mass change data from satellite and in-situ observations: resolution and coverage for trend analysis", "chunk_index": 14, "token_count": 829, "text_raw": "When measured over long periods and extensive geographical scales, trends in glacier mass changes are clear indicators of global climate change [[3](https://doi.org/10.1038/s41586-019-1071-0), [9](https://doi.org/10.1038/s41586-024-08545-z)]. To be able to derive such trends of glacier mass changes and for climate change analysis/monitoring to become reliable and possible, the glacier mass change dataset should at least exhibit a comprehensive spatial coverage (i.e. global), a long and continuous temporal coverage (> 30 years), quantified and transparent pixel-by-pixel uncertainty estimates that meet international proposed thresholds [[4](https://library.wmo.int/records/item/58111-the-2022-gcos-ecvs-requirements-gcos-245)], a validation effort or a comparison to theoretical models, and an adequate spatio-temporal resolution (cfr. the \"Maturity Matrix\" [[10](https://doi.org/10.1175/BAMS-D-21-0109.1)]).\n\nThe [glacier mass change dataset on the CDS](https://cds.climate.copernicus.eu/datasets/derived-gridded-glacier-mass-change?tab=overview) is found to exhibit a consistent suitable spatial (0.5° x 0.5°) and temporal (annual) resolution, as well as an extensive coverage (> 30 years globally), to conduct a meaningful analysis of linear and quadratic trends in glacier mass changes at local (pixel-by-pixel), regional, and global scales. The spatial and temporal resolution/extent of the data furthermore align with international standards such as those proposed by GCOS [[4](https://library.wmo.int/records/item/58111-the-2022-gcos-ecvs-requirements-gcos-245)] and data gaps are practically non-existant in this dataset. The resulting linear glacier mass change trends show consistent mass loss across most pixels across the globe, while quadratic trends highlight a general accelerating rate of loss over the last several decades. These observations align well with theoretical considerations of the response of glaciers to a linear warming trend [[8](https://doi.org/10.1007/s003820050222)]. The majority of the pixel-based trends are statistically significant at $\\alpha$ = 0.05, further enhancing the dataset's credibility and reliability for global climate monitoring. The corresponding glacier mass changes are all assumed to occur above sea level.\n\nUsers should, however, keep in mind that mass change artefacts may occur in polar regions due to a variable absolute surface area of the pixels, and that the dataset lacks detailed information on sampling density. The latter is important because annual in-situ glaciological samples (~500 glaciers) form the basis to infer mass changes for a much larger sample of geodetic mass change observations and unobserved glaciers [[2](https://doi.org/10.5194/essd-17-1977-2025)]. This, for example, results in the fact that the product has a less well-pronounced observational coverage with respect to the annual variability of glacier mass changes (which may be less well represented), but an excellent coverage with respect to their long-term trend. Therefore, knowledge of the amount and spatial distribution of glaciers with a consistent and long-term glaciological sample is crucial for assessing the representativeness of the data. Despite sampling density limitations and the related uncertainties, the dataset’s robust temporal/spatial coverage and resolution make it a reliable tool for assessing glacier mass changes as an indicator of global climatic changes.", "text_with_prefix": "EQC Quality Assessment: \"Glacier mass change data from satellite and in-situ observations: resolution and coverage for trend analysis\"\nDataset: derived-gridded-glacier-mass-change [CDS]\nAspect: trend-assessment_q02 | Category: Satellite_ECVs\nSection: Glacier mass change data from satellite and in-situ observations: resolution and coverage for trend analysis > Analysis and results > 4. Short summary and take-home messages\n---\nWhen measured over long periods and extensive geographical scales, trends in glacier mass changes are clear indicators of global climate change [[3](https://doi.org/10.1038/s41586-019-1071-0), [9](https://doi.org/10.1038/s41586-024-08545-z)]. To be able to derive such trends of glacier mass changes and for climate change analysis/monitoring to become reliable and possible, the glacier mass change dataset should at least exhibit a comprehensive spatial coverage (i.e. global), a long and continuous temporal coverage (> 30 years), quantified and transparent pixel-by-pixel uncertainty estimates that meet international proposed thresholds [[4](https://library.wmo.int/records/item/58111-the-2022-gcos-ecvs-requirements-gcos-245)], a validation effort or a comparison to theoretical models, and an adequate spatio-temporal resolution (cfr. the \"Maturity Matrix\" [[10](https://doi.org/10.1175/BAMS-D-21-0109.1)]).\n\nThe [glacier mass change dataset on the CDS](https://cds.climate.copernicus.eu/datasets/derived-gridded-glacier-mass-change?tab=overview) is found to exhibit a consistent suitable spatial (0.5° x 0.5°) and temporal (annual) resolution, as well as an extensive coverage (> 30 years globally), to conduct a meaningful analysis of linear and quadratic trends in glacier mass changes at local (pixel-by-pixel), regional, and global scales. The spatial and temporal resolution/extent of the data furthermore align with international standards such as those proposed by GCOS [[4](https://library.wmo.int/records/item/58111-the-2022-gcos-ecvs-requirements-gcos-245)] and data gaps are practically non-existant in this dataset. The resulting linear glacier mass change trends show consistent mass loss across most pixels across the globe, while quadratic trends highlight a general accelerating rate of loss over the last several decades. These observations align well with theoretical considerations of the response of glaciers to a linear warming trend [[8](https://doi.org/10.1007/s003820050222)]. The majority of the pixel-based trends are statistically significant at $\\alpha$ = 0.05, further enhancing the dataset's credibility and reliability for global climate monitoring. The corresponding glacier mass changes are all assumed to occur above sea level.\n\nUsers should, however, keep in mind that mass change artefacts may occur in polar regions due to a variable absolute surface area of the pixels, and that the dataset lacks detailed information on sampling density. The latter is important because annual in-situ glaciological samples (~500 glaciers) form the basis to infer mass changes for a much larger sample of geodetic mass change observations and unobserved glaciers [[2](https://doi.org/10.5194/essd-17-1977-2025)]. This, for example, results in the fact that the product has a less well-pronounced observational coverage with respect to the annual variability of glacier mass changes (which may be less well represented), but an excellent coverage with respect to their long-term trend. Therefore, knowledge of the amount and spatial distribution of glaciers with a consistent and long-term glaciological sample is crucial for assessing the representativeness of the data. Despite sampling density limitations and the related uncertainties, the dataset’s robust temporal/spatial coverage and resolution make it a reliable tool for assessing glacier mass changes as an indicator of global climatic changes."} {"chunk_id": "satellite_derived-gridded-glacier-mass-change_trend-assessment_q02__e35a7e9a5437", "report_id": "satellite_derived-gridded-glacier-mass-change_trend-assessment_q02", "dataset_id": "derived-gridded-glacier-mass-change", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q02", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Glacier mass change data from satellite and in-situ observations: resolution and coverage for trend analysis > ℹ️ If you want to know more > Key resources", "title": "Glacier mass change data from satellite and in-situ observations: resolution and coverage for trend analysis", "chunk_index": 15, "token_count": 349, "text_raw": "- [Glacier mass change gridded data from 1976 to present derived from the Fluctuations of Glaciers Database](https://cds.climate.copernicus.eu/datasets/derived-gridded-glacier-mass-change?tab=overview)\n- [Documentation on the CDS](https://cds.climate.copernicus.eu/datasets/derived-gridded-glacier-mass-change?tab=overview) and the [ECMWF Confluence Wiki](https://confluence.ecmwf.int/pages/viewpage.action?pageId=355349383) (Copernicus Knowledge Base)\n- [World Glacier Monitoring Service (WGMS)](https://wgms.ch/)\n- [Glacier Mass Balance Intercomparison Exercise (GlaMBIE)](https://glambie.org/)\n- [Copernicus climate change indicators: glaciers](https://climate.copernicus.eu/climate-indicators/glaciers)\n- [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control` prepared by [B-Open](https://www.bopen.eu)", "text_with_prefix": "EQC Quality Assessment: \"Glacier mass change data from satellite and in-situ observations: resolution and coverage for trend analysis\"\nDataset: derived-gridded-glacier-mass-change [CDS]\nAspect: trend-assessment_q02 | Category: Satellite_ECVs\nSection: Glacier mass change data from satellite and in-situ observations: resolution and coverage for trend analysis > ℹ️ If you want to know more > Key resources\n---\n- [Glacier mass change gridded data from 1976 to present derived from the Fluctuations of Glaciers Database](https://cds.climate.copernicus.eu/datasets/derived-gridded-glacier-mass-change?tab=overview)\n- [Documentation on the CDS](https://cds.climate.copernicus.eu/datasets/derived-gridded-glacier-mass-change?tab=overview) and the [ECMWF Confluence Wiki](https://confluence.ecmwf.int/pages/viewpage.action?pageId=355349383) (Copernicus Knowledge Base)\n- [World Glacier Monitoring Service (WGMS)](https://wgms.ch/)\n- [Glacier Mass Balance Intercomparison Exercise (GlaMBIE)](https://glambie.org/)\n- [Copernicus climate change indicators: glaciers](https://climate.copernicus.eu/climate-indicators/glaciers)\n- [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control` prepared by [B-Open](https://www.bopen.eu)"} {"chunk_id": "satellite_derived-gridded-glacier-mass-change_trend-assessment_q02__662ea1c70408", "report_id": "satellite_derived-gridded-glacier-mass-change_trend-assessment_q02", "dataset_id": "derived-gridded-glacier-mass-change", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q02", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Glacier mass change data from satellite and in-situ observations: resolution and coverage for trend analysis > ℹ️ If you want to know more > References", "title": "Glacier mass change data from satellite and in-situ observations: resolution and coverage for trend analysis", "chunk_index": 16, "token_count": 989, "text_raw": "- [[1](https://wgms.ch/)] WGMS (2022). Fluctuations of Glaciers Database. doi: 10.5904/wgms-fog-2022-09.\n\n- [[2](https://doi.org/10.5194/essd-17-1977-2025)] Dussaillant, I., Hugonnet, R., Huss, M., Berthier, E., Bannwart, J., Paul, F., and Zemp, M. (2025). Annual mass changes for each glacier in the world from 1976 to 2023, Earth Syst. Sci. Data, https://doi.org/10.5194/essd-2024-323\n\n- [[3](https://doi.org/10.1038/s41586-019-1071-0)] Zemp, M., Huss, M., Thibert, E., Eckert, N., McNabb, R., Huber, J., Barandun, M., Machguth, H., Nussbaumer, S. U., Gärtner-Roer, I., Thomson, L., Paul, F., Maussion, F., Kutuzov, S., and Cogley, J. G. (2019). Global glacier mass changes and their contributions to sea-level rise from 1961 to 2016. Nature, 568, 382–386. doi: 10.1038/s41586-019-1071-0.\n\n- [[4](https://library.wmo.int/records/item/58111-the-2022-gcos-ecvs-requirements-gcos-245)] GCOS (Global Climate Observing System) (2022). The 2022 GCOS ECVs Requirements (GCOS-245). World Meteorological Organization: Geneva, Switzerland. doi: https://library.wmo.int/idurl/4/58111\n\n- [[5](https://www.glims.org/RGI/randolph60.html)] RGI Consortium (2017). Randolph Glacier Inventory – A Dataset of Global Glacier Outlines: Version 6.0: Technical Report, Global Land Ice Measurements from Space, Colorado, USA. Digital Media. doi: 10.7265/N5-RGI-60.\n\n- [[6](https://doi.org/10.1038/s41586-021-03436-z)] Hugonnet, R., McNabb, R., Berthier, E., Menounos, B., Nuth, C., Girod, L., Huss, M., Farinotti, D., Dussaillant, I., Brun, F., and Kääb, A. (2021). Accelerated global glacier mass loss in the early twenty-first century. Nature 592, 726–731. doi: 10.1038/s41586-021-03436-z.\n\n- [[7](https://doi.org/10.3389/feart.2019.00096)] Wouters, B., Gardner, A. S., and Moholdt, G. (2019). Global Glacier Mass Loss During the GRACE Satellite Mission (2002-2016). Front. Earth Sci. 7. doi: 10.3389/feart.2019.00096.\n\n- [[8](https://doi.org/10.1007/s003820050222)] Oerlemans, J., Anderson, B., Hubbard, A., Huybrechts, P., Jóhannesson, T., W. H. Knap, M. Schmeits, A. P. Stroeven, R. S. W. van de Wal, J. Wallinga and Z. Zuo (1998). Modelling the response of glaciers to climate warming. Climate Dynamics 14, 267–274. doi: 10.1007/s003820050222\n\n- [[9](https://doi.org/10.1038/s41586-024-08545-z)] The GlaMBIE Team (2025). Community estimate of global glacier mass changes from 2000 to 2023. Nature (2025). doi: 10.1038/s41586-024-08545-z.", "text_with_prefix": "EQC Quality Assessment: \"Glacier mass change data from satellite and in-situ observations: resolution and coverage for trend analysis\"\nDataset: derived-gridded-glacier-mass-change [CDS]\nAspect: trend-assessment_q02 | Category: Satellite_ECVs\nSection: Glacier mass change data from satellite and in-situ observations: resolution and coverage for trend analysis > ℹ️ If you want to know more > References\n---\n- [[1](https://wgms.ch/)] WGMS (2022). Fluctuations of Glaciers Database. doi: 10.5904/wgms-fog-2022-09.\n\n- [[2](https://doi.org/10.5194/essd-17-1977-2025)] Dussaillant, I., Hugonnet, R., Huss, M., Berthier, E., Bannwart, J., Paul, F., and Zemp, M. (2025). Annual mass changes for each glacier in the world from 1976 to 2023, Earth Syst. Sci. Data, https://doi.org/10.5194/essd-2024-323\n\n- [[3](https://doi.org/10.1038/s41586-019-1071-0)] Zemp, M., Huss, M., Thibert, E., Eckert, N., McNabb, R., Huber, J., Barandun, M., Machguth, H., Nussbaumer, S. U., Gärtner-Roer, I., Thomson, L., Paul, F., Maussion, F., Kutuzov, S., and Cogley, J. G. (2019). Global glacier mass changes and their contributions to sea-level rise from 1961 to 2016. Nature, 568, 382–386. doi: 10.1038/s41586-019-1071-0.\n\n- [[4](https://library.wmo.int/records/item/58111-the-2022-gcos-ecvs-requirements-gcos-245)] GCOS (Global Climate Observing System) (2022). The 2022 GCOS ECVs Requirements (GCOS-245). World Meteorological Organization: Geneva, Switzerland. doi: https://library.wmo.int/idurl/4/58111\n\n- [[5](https://www.glims.org/RGI/randolph60.html)] RGI Consortium (2017). Randolph Glacier Inventory – A Dataset of Global Glacier Outlines: Version 6.0: Technical Report, Global Land Ice Measurements from Space, Colorado, USA. Digital Media. doi: 10.7265/N5-RGI-60.\n\n- [[6](https://doi.org/10.1038/s41586-021-03436-z)] Hugonnet, R., McNabb, R., Berthier, E., Menounos, B., Nuth, C., Girod, L., Huss, M., Farinotti, D., Dussaillant, I., Brun, F., and Kääb, A. (2021). Accelerated global glacier mass loss in the early twenty-first century. Nature 592, 726–731. doi: 10.1038/s41586-021-03436-z.\n\n- [[7](https://doi.org/10.3389/feart.2019.00096)] Wouters, B., Gardner, A. S., and Moholdt, G. (2019). Global Glacier Mass Loss During the GRACE Satellite Mission (2002-2016). Front. Earth Sci. 7. doi: 10.3389/feart.2019.00096.\n\n- [[8](https://doi.org/10.1007/s003820050222)] Oerlemans, J., Anderson, B., Hubbard, A., Huybrechts, P., Jóhannesson, T., W. H. Knap, M. Schmeits, A. P. Stroeven, R. S. W. van de Wal, J. Wallinga and Z. Zuo (1998). Modelling the response of glaciers to climate warming. Climate Dynamics 14, 267–274. doi: 10.1007/s003820050222\n\n- [[9](https://doi.org/10.1038/s41586-024-08545-z)] The GlaMBIE Team (2025). Community estimate of global glacier mass changes from 2000 to 2023. Nature (2025). doi: 10.1038/s41586-024-08545-z."} {"chunk_id": "satellite_derived-gridded-glacier-mass-change_trend-assessment_q02__0ffc7ce0acd2", "report_id": "satellite_derived-gridded-glacier-mass-change_trend-assessment_q02", "dataset_id": "derived-gridded-glacier-mass-change", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q02", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Glacier mass change data from satellite and in-situ observations: resolution and coverage for trend analysis > ℹ️ If you want to know more > References", "title": "Glacier mass change data from satellite and in-situ observations: resolution and coverage for trend analysis", "chunk_index": 17, "token_count": 479, "text_raw": "IE Team (2025). Community estimate of global glacier mass changes from 2000 to 2023. Nature (2025). doi: 10.1038/s41586-024-08545-z.\n\n- [[10](https://doi.org/10.1175/BAMS-D-21-0109.1)] Yang, C. X., Cagnazzo, C., Artale, V., Nardelli, B. B., Buontempo, C., Busatto, J., Caporaso, L., Cesarini, C., Cionni, I., Coll, J., Crezee, B., Cristofanelli, P., de Toma, V., Essa, Y. H., Eyring, V., Fierli, F., Grant, L., Hassler, B., Hirschi, M., Huybrechts, P., Le Merle, E., Leonelli, F. E., Lin, X., Madonna, F., Mason, E., Massonnet, F., Marcos, M., Marullo, S., Muller, B., Obregon, A., Organelli, E., Palacz, A., Pascual, A., Pisano, A., Putero, D., Rana, A., Sanchez-Roman, A., Seneviratne, S. I., Serva, F., Storto, A., Thiery, W., Throne, P., Van Tricht, L., Verhaegen, Y., Volpe, G., and Santoleri, R. (2022). Independent Quality Assessment of Essential Climate Variables: Lessons Learned from the Copernicus Climate Change Service, B. Am. Meteorol. Soc., 103, E2032–E2049, doi: 10.1175/Bams-D-21-0109.1.", "text_with_prefix": "EQC Quality Assessment: \"Glacier mass change data from satellite and in-situ observations: resolution and coverage for trend analysis\"\nDataset: derived-gridded-glacier-mass-change [CDS]\nAspect: trend-assessment_q02 | Category: Satellite_ECVs\nSection: Glacier mass change data from satellite and in-situ observations: resolution and coverage for trend analysis > ℹ️ If you want to know more > References\n---\nIE Team (2025). Community estimate of global glacier mass changes from 2000 to 2023. Nature (2025). doi: 10.1038/s41586-024-08545-z.\n\n- [[10](https://doi.org/10.1175/BAMS-D-21-0109.1)] Yang, C. X., Cagnazzo, C., Artale, V., Nardelli, B. B., Buontempo, C., Busatto, J., Caporaso, L., Cesarini, C., Cionni, I., Coll, J., Crezee, B., Cristofanelli, P., de Toma, V., Essa, Y. H., Eyring, V., Fierli, F., Grant, L., Hassler, B., Hirschi, M., Huybrechts, P., Le Merle, E., Leonelli, F. E., Lin, X., Madonna, F., Mason, E., Massonnet, F., Marcos, M., Marullo, S., Muller, B., Obregon, A., Organelli, E., Palacz, A., Pascual, A., Pisano, A., Putero, D., Rana, A., Sanchez-Roman, A., Seneviratne, S. I., Serva, F., Storto, A., Thiery, W., Throne, P., Van Tricht, L., Verhaegen, Y., Volpe, G., and Santoleri, R. (2022). Independent Quality Assessment of Essential Climate Variables: Lessons Learned from the Copernicus Climate Change Service, B. Am. Meteorol. Soc., 103, E2032–E2049, doi: 10.1175/Bams-D-21-0109.1."} {"chunk_id": "satellite_insitu-glaciers-extent_uncertainty_q01__19b394282c3e", "report_id": "satellite_insitu-glaciers-extent_uncertainty_q01", "dataset_id": "insitu-glaciers-extent", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty_q01", "aspect_base": "uncertainty", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates", "title": "Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates", "chunk_index": 0, "token_count": 117, "text_raw": "Production date: 31-05-2025\n\nDataset version: 6.0\n\nProduced by: Yoni Verhaegen and Philippe Huybrechts (Vrije Universiteit Brussel)", "text_with_prefix": "EQC Quality Assessment: \"Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates\"\nDataset: insitu-glaciers-extent [CDS]\nAspect: uncertainty_q01 | Category: Satellite_ECVs\nSection: Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates\n---\nProduction date: 31-05-2025\n\nDataset version: 6.0\n\nProduced by: Yoni Verhaegen and Philippe Huybrechts (Vrije Universiteit Brussel)"} {"chunk_id": "satellite_insitu-glaciers-extent_uncertainty_q01__9db448850bf4", "report_id": "satellite_insitu-glaciers-extent_uncertainty_q01", "dataset_id": "insitu-glaciers-extent", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty_q01", "aspect_base": "uncertainty", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates > Quality assessment question", "title": "Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates", "chunk_index": 1, "token_count": 542, "text_raw": "**\"What is the temporal distribution of the delineated glacier outlines, nominally provided for the year 2000, and how well can they be used to estimate regional and global glacier ice volumes?\"**\n\nGlaciers are a major contributor to current global sea-level rise, a resource of fresh water, a potential threat of natural hazards, and an important factor for hydro-power production, recreation and tourism. A proper assessment of glacier areas, glacier characteristics, as well as their changes due to warming climatic conditions therefore plays a crucial role in dealing with these issues. In that regard, the \"[Glaciers distribution data from the Randolph Glacier Inventory (RGI) for year 2000](https://cds.climate.copernicus.eu/datasets/insitu-glaciers-extent?tab=overview)\" dataset on the Climate Data Store (CDS) provides key information with respect to glacier extent and their characteristics. The RGI dataset is a collection of digital glacier and ice cap outlines at the global scale, nominally provided for the year 2000 [[1](https://www.glims.org/RGI/randolph60.html), [2](https://doi.org/10.3189/2014JoG13J176)]. The data are available in both vector (a shapefile with polygons of individual glacier outlines) and raster (as gridded data with the aggregated fractional glacier areas per pixel) format. Although the goal of the dataset is to provide glacier outlines of all glaciers on Earth as close as possible to the year 2000, one of the main known issues of the dataset is the fact that the acquisition date of the delineated glacier data varies substantially. This notebook investigates the corresponding temporal distribution of delineated glacier data in the vector version of the dataset and evaluates its implications for the estimation of current global glacier ice volumes using the examples of Farinotti et al. (2019) [[3](https://doi.org/10.1038/s41561-019-0300-3)] and GlaMBIE (2025) [[14](https://doi.org/10.1038/s41586-024-08545-z)]. Since both studies made use of the RGIv6.0, we also use this version in our notebook.", "text_with_prefix": "EQC Quality Assessment: \"Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates\"\nDataset: insitu-glaciers-extent [CDS]\nAspect: uncertainty_q01 | Category: Satellite_ECVs\nSection: Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates > Quality assessment question\n---\n**\"What is the temporal distribution of the delineated glacier outlines, nominally provided for the year 2000, and how well can they be used to estimate regional and global glacier ice volumes?\"**\n\nGlaciers are a major contributor to current global sea-level rise, a resource of fresh water, a potential threat of natural hazards, and an important factor for hydro-power production, recreation and tourism. A proper assessment of glacier areas, glacier characteristics, as well as their changes due to warming climatic conditions therefore plays a crucial role in dealing with these issues. In that regard, the \"[Glaciers distribution data from the Randolph Glacier Inventory (RGI) for year 2000](https://cds.climate.copernicus.eu/datasets/insitu-glaciers-extent?tab=overview)\" dataset on the Climate Data Store (CDS) provides key information with respect to glacier extent and their characteristics. The RGI dataset is a collection of digital glacier and ice cap outlines at the global scale, nominally provided for the year 2000 [[1](https://www.glims.org/RGI/randolph60.html), [2](https://doi.org/10.3189/2014JoG13J176)]. The data are available in both vector (a shapefile with polygons of individual glacier outlines) and raster (as gridded data with the aggregated fractional glacier areas per pixel) format. Although the goal of the dataset is to provide glacier outlines of all glaciers on Earth as close as possible to the year 2000, one of the main known issues of the dataset is the fact that the acquisition date of the delineated glacier data varies substantially. This notebook investigates the corresponding temporal distribution of delineated glacier data in the vector version of the dataset and evaluates its implications for the estimation of current global glacier ice volumes using the examples of Farinotti et al. (2019) [[3](https://doi.org/10.1038/s41561-019-0300-3)] and GlaMBIE (2025) [[14](https://doi.org/10.1038/s41586-024-08545-z)]. Since both studies made use of the RGIv6.0, we also use this version in our notebook."} {"chunk_id": "satellite_insitu-glaciers-extent_uncertainty_q01__211caf1209f6", "report_id": "satellite_insitu-glaciers-extent_uncertainty_q01", "dataset_id": "insitu-glaciers-extent", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty_q01", "aspect_base": "uncertainty", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates > Quality assessment statements", "title": "Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates", "chunk_index": 2, "token_count": 375, "text_raw": "These are the key outcomes of this assessment\n\n- The RGI dataset, available on the Climate Data Store, is the most comprehensive source of digital glacier area data and provides consistent, static glacier outlines for climatological, hydrological, and glaciological applications at regional to global scales. It holds a snapshot of global glacier area data that is assumed to be representative for the year 2000. However, in reality there is a variation in delineation dates (a certain amount of glaciers were delineated before or after the year 2000). The acquired glacier outlines that deviate from the year 2000 can lead to systematic errors that are apparant for certain applications, such as the use of RGI glacier outlines as input for regional/global glacier volume estimates. \n- Nevertheless, the assessment reveals that the (area-weighted) mean acquisition date of all glacier outlines in the dataset is relatively close to 2000, and that the systematic error at the global scale is relatively small compared to the total glacier volume estimates from the glacier community. This makes the dataset reliable to be used as input data for regional and global glacier volume estimates. However, regional and glacier-specific errors can vary significantly due to factors like strongly deviating delineation dates in some areas, varying spatial glacier mass loss rates and glacier response times, and the presence of a supraglacial debris cover (which may cause significant omission errors in glacierized areas). Users should carefully evaluate these factors for specific applications.\n```", "text_with_prefix": "EQC Quality Assessment: \"Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates\"\nDataset: insitu-glaciers-extent [CDS]\nAspect: uncertainty_q01 | Category: Satellite_ECVs\nSection: Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates > Quality assessment statements\n---\nThese are the key outcomes of this assessment\n\n- The RGI dataset, available on the Climate Data Store, is the most comprehensive source of digital glacier area data and provides consistent, static glacier outlines for climatological, hydrological, and glaciological applications at regional to global scales. It holds a snapshot of global glacier area data that is assumed to be representative for the year 2000. However, in reality there is a variation in delineation dates (a certain amount of glaciers were delineated before or after the year 2000). The acquired glacier outlines that deviate from the year 2000 can lead to systematic errors that are apparant for certain applications, such as the use of RGI glacier outlines as input for regional/global glacier volume estimates. \n- Nevertheless, the assessment reveals that the (area-weighted) mean acquisition date of all glacier outlines in the dataset is relatively close to 2000, and that the systematic error at the global scale is relatively small compared to the total glacier volume estimates from the glacier community. This makes the dataset reliable to be used as input data for regional and global glacier volume estimates. However, regional and glacier-specific errors can vary significantly due to factors like strongly deviating delineation dates in some areas, varying spatial glacier mass loss rates and glacier response times, and the presence of a supraglacial debris cover (which may cause significant omission errors in glacierized areas). Users should carefully evaluate these factors for specific applications.\n```"} {"chunk_id": "satellite_insitu-glaciers-extent_uncertainty_q01__e73f5f3f90f4", "report_id": "satellite_insitu-glaciers-extent_uncertainty_q01", "dataset_id": "insitu-glaciers-extent", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty_q01", "aspect_base": "uncertainty", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates > Methodology > Dataset description", "title": "Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates", "chunk_index": 3, "token_count": 415, "text_raw": "The dataset of the glaciers distribution on the Climate Data Store (CDS) is an almost complete collection of digital glacier and ice cap outlines and their geometrical/hypsometrical characteristics from various data sources at the global scale [[1](https://www.glims.org/RGI/randolph60.html), [2](https://doi.org/10.3189/2014JoG13J176)]. The RGI dataset on the CDS is considered a snapshot of glacier outlines around the year 2000, assembled mainly from satellite images, and is currently the most complete dataset of glacier outlines. Simply stated, the glaciers in the dataset were automatically classified using the distinctive spectral reflectance signatures of bedrock and ice. During post-processing, raw glacier outlines are quality checked and manually corrected if required (e.g. in the case of a supraglacial debris cover). The vector part of the dataset divides the glaciers into separate RGI regions, i.e. there are 19 of those regions (\"clusters\" of glaciers) in RGIv6.0. The raster version of the data contains aggregated fractional glacier areas for each pixel of 1 by 1 degree but does not provide sufficient information to classify glaciers into distinct RGI regions. For a more detailed description of the data acquisition and processing methods, we refer to the [documentation on the CDS](https://cds.climate.copernicus.eu/datasets/insitu-glaciers-extent?tab=overview) and the [ECMWF Confluence Wiki](https://confluence.ecmwf.int/display/CKB/Glacier+Area) (Copernicus Knowledge Base).", "text_with_prefix": "EQC Quality Assessment: \"Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates\"\nDataset: insitu-glaciers-extent [CDS]\nAspect: uncertainty_q01 | Category: Satellite_ECVs\nSection: Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates > Methodology > Dataset description\n---\nThe dataset of the glaciers distribution on the Climate Data Store (CDS) is an almost complete collection of digital glacier and ice cap outlines and their geometrical/hypsometrical characteristics from various data sources at the global scale [[1](https://www.glims.org/RGI/randolph60.html), [2](https://doi.org/10.3189/2014JoG13J176)]. The RGI dataset on the CDS is considered a snapshot of glacier outlines around the year 2000, assembled mainly from satellite images, and is currently the most complete dataset of glacier outlines. Simply stated, the glaciers in the dataset were automatically classified using the distinctive spectral reflectance signatures of bedrock and ice. During post-processing, raw glacier outlines are quality checked and manually corrected if required (e.g. in the case of a supraglacial debris cover). The vector part of the dataset divides the glaciers into separate RGI regions, i.e. there are 19 of those regions (\"clusters\" of glaciers) in RGIv6.0. The raster version of the data contains aggregated fractional glacier areas for each pixel of 1 by 1 degree but does not provide sufficient information to classify glaciers into distinct RGI regions. For a more detailed description of the data acquisition and processing methods, we refer to the [documentation on the CDS](https://cds.climate.copernicus.eu/datasets/insitu-glaciers-extent?tab=overview) and the [ECMWF Confluence Wiki](https://confluence.ecmwf.int/display/CKB/Glacier+Area) (Copernicus Knowledge Base)."} {"chunk_id": "satellite_insitu-glaciers-extent_uncertainty_q01__f01d50276372", "report_id": "satellite_insitu-glaciers-extent_uncertainty_q01", "dataset_id": "insitu-glaciers-extent", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty_q01", "aspect_base": "uncertainty", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates > Methodology > Structure and (sub)sections", "title": "Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates", "chunk_index": 4, "token_count": 182, "text_raw": "**[](section-1)**\n\n* [](section-1-1)\n* [](section-1-2)\n* [](section-1-3)\n\n**[](section-2)**\n* [](section-2-1)\n* [](section-2-2)\n* [](section-2-3)\n\n**[](section-3)**\n* [](section-3-1)\n* [](section-3-2)\n* [](section-3-3)\n\n**[](section-4)**", "text_with_prefix": "EQC Quality Assessment: \"Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates\"\nDataset: insitu-glaciers-extent [CDS]\nAspect: uncertainty_q01 | Category: Satellite_ECVs\nSection: Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates > Methodology > Structure and (sub)sections\n---\n**[](section-1)**\n\n* [](section-1-1)\n* [](section-1-2)\n* [](section-1-3)\n\n**[](section-2)**\n* [](section-2-1)\n* [](section-2-2)\n* [](section-2-3)\n\n**[](section-3)**\n* [](section-3-1)\n* [](section-3-2)\n* [](section-3-3)\n\n**[](section-4)**"} {"chunk_id": "satellite_insitu-glaciers-extent_uncertainty_q01__138ff6006f03", "report_id": "satellite_insitu-glaciers-extent_uncertainty_q01", "dataset_id": "insitu-glaciers-extent", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty_q01", "aspect_base": "uncertainty", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates > Analysis and results > 1. Data preparation and processing > 1.2 Define request and download", "title": "Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates", "chunk_index": 5, "token_count": 333, "text_raw": "Then we define requests for download from the CDS:\n\n🚨 **The files can be large! Since the data files to be downloaded and manipulated have a considerable size, this may take a couple of minutes.**\\\n🚨 **Insert the correct RGI version number to be downloaded below:**\n\nSELECT THE RGI VERSION TO DOWNLOAD #############\nSelect correct column names based on RGI version\nGlacier extent data\nGlacier mass change data\n\nNext, we download the data:\n\nGet glacier extent data: takes a couple of minutes\nDownload glacier mass change data\n\n```text\nDownloading and handling glacier extent data from the CDS, this may take a couple of minutes...\n```\n\n```text\n100%|██████████| 1/1 [00:00<00:00, 67.66it/s]\n```\n\n```text\nDownloading glacier mass change data...\n```\n\n```text\n100%|██████████| 1/1 [00:01<00:00, 2.00s/it]\n```\n\n```text\nDownloading done.\n```\n\nWe define the functions to be used:\n\n(section-1-3)=", "text_with_prefix": "EQC Quality Assessment: \"Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates\"\nDataset: insitu-glaciers-extent [CDS]\nAspect: uncertainty_q01 | Category: Satellite_ECVs\nSection: Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates > Analysis and results > 1. Data preparation and processing > 1.2 Define request and download\n---\nThen we define requests for download from the CDS:\n\n🚨 **The files can be large! Since the data files to be downloaded and manipulated have a considerable size, this may take a couple of minutes.**\\\n🚨 **Insert the correct RGI version number to be downloaded below:**\n\nSELECT THE RGI VERSION TO DOWNLOAD #############\nSelect correct column names based on RGI version\nGlacier extent data\nGlacier mass change data\n\nNext, we download the data:\n\nGet glacier extent data: takes a couple of minutes\nDownload glacier mass change data\n\n```text\nDownloading and handling glacier extent data from the CDS, this may take a couple of minutes...\n```\n\n```text\n100%|██████████| 1/1 [00:00<00:00, 67.66it/s]\n```\n\n```text\nDownloading glacier mass change data...\n```\n\n```text\n100%|██████████| 1/1 [00:01<00:00, 2.00s/it]\n```\n\n```text\nDownloading done.\n```\n\nWe define the functions to be used:\n\n(section-1-3)="} {"chunk_id": "satellite_insitu-glaciers-extent_uncertainty_q01__bb7d77e557ee", "report_id": "satellite_insitu-glaciers-extent_uncertainty_q01", "dataset_id": "insitu-glaciers-extent", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty_q01", "aspect_base": "uncertainty", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates > Analysis and results > 1. Data preparation and processing > 1.3 Display and inspect data", "title": "Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates", "chunk_index": 6, "token_count": 980, "text_raw": "Lastly, we can read and inspect the data. Let us print out the data to inspect its structure:\n\n```text\nC3S_ID RGIID GLIMSID BGNDATE ENDDATE \\\nindex \n0 C3S_000001 RGI60-01.00001 G213177E63689N 20090703 -9999999 \n1 C3S_000002 RGI60-01.00002 G213332E63404N 20090703 -9999999 \n2 C3S_000003 RGI60-01.00003 G213920E63376N 20090703 -9999999 \n3 C3S_000004 RGI60-01.00004 G213880E63381N 20090703 -9999999 \n4 C3S_000005 RGI60-01.00005 G212943E63551N 20090703 -9999999 \n... ... ... ... ... ... \n216424 C3S_216425 RGI60-19.02748 G322268E53986S 20020502 -9999999 \n216425 C3S_216426 RGI60-19.02749 G323864E54831S 20030207 -9999999 \n216426 C3S_216427 RGI60-19.02750 G322698E54188S 20030207 -9999999 \n216427 C3S_216428 RGI60-19.02751 G269573E68866S 19870101 -9999999 \n216428 C3S_216429 RGI60-19.02752 G037714E46897S 19660301 -9999999\n\nCENLON CENLAT O1REGION O2REGION AREA ZMIN ZMAX ZMED SLOPE \\\nindex \n0 -146.8230 63.6890 1 2 0.360 1936 2725 2385 42.0 \n1 -146.6680 63.4040 1 2 0.558 1713 2144 2005 16.0 \n2 -146.0800 63.3760 1 2 1.685 1609 2182 1868 18.0 \n3 -146.1200 63.3810 1 2 3.681 1273 2317 1944 19.0 \n4 -147.0570 63.5510 1 2 2.573 1494 2317 1914 16.0 \n... ... ... ... ... ... ... ... ... ... \n216424 -37.7325 -53.9860 19 3 0.042 310 510 -999 29.9 \n216425 -36.1361 -54.8310 19 3 0.567 330 830 -999 23.6 \n216426 -37.3018 -54.1884 19 3 4.118 10 1110 -999 16.8 \n216427 -90.4266 -68.8656 19 1 0.011 170 270 -999 0.4 \n216428 37.7140 -46.8972 19 4 0.528 970 1170 -999 9.6", "text_with_prefix": "EQC Quality Assessment: \"Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates\"\nDataset: insitu-glaciers-extent [CDS]\nAspect: uncertainty_q01 | Category: Satellite_ECVs\nSection: Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates > Analysis and results > 1. Data preparation and processing > 1.3 Display and inspect data\n---\nLastly, we can read and inspect the data. Let us print out the data to inspect its structure:\n\n```text\nC3S_ID RGIID GLIMSID BGNDATE ENDDATE \\\nindex \n0 C3S_000001 RGI60-01.00001 G213177E63689N 20090703 -9999999 \n1 C3S_000002 RGI60-01.00002 G213332E63404N 20090703 -9999999 \n2 C3S_000003 RGI60-01.00003 G213920E63376N 20090703 -9999999 \n3 C3S_000004 RGI60-01.00004 G213880E63381N 20090703 -9999999 \n4 C3S_000005 RGI60-01.00005 G212943E63551N 20090703 -9999999 \n... ... ... ... ... ... \n216424 C3S_216425 RGI60-19.02748 G322268E53986S 20020502 -9999999 \n216425 C3S_216426 RGI60-19.02749 G323864E54831S 20030207 -9999999 \n216426 C3S_216427 RGI60-19.02750 G322698E54188S 20030207 -9999999 \n216427 C3S_216428 RGI60-19.02751 G269573E68866S 19870101 -9999999 \n216428 C3S_216429 RGI60-19.02752 G037714E46897S 19660301 -9999999\n\nCENLON CENLAT O1REGION O2REGION AREA ZMIN ZMAX ZMED SLOPE \\\nindex \n0 -146.8230 63.6890 1 2 0.360 1936 2725 2385 42.0 \n1 -146.6680 63.4040 1 2 0.558 1713 2144 2005 16.0 \n2 -146.0800 63.3760 1 2 1.685 1609 2182 1868 18.0 \n3 -146.1200 63.3810 1 2 3.681 1273 2317 1944 19.0 \n4 -147.0570 63.5510 1 2 2.573 1494 2317 1914 16.0 \n... ... ... ... ... ... ... ... ... ... \n216424 -37.7325 -53.9860 19 3 0.042 310 510 -999 29.9 \n216425 -36.1361 -54.8310 19 3 0.567 330 830 -999 23.6 \n216426 -37.3018 -54.1884 19 3 4.118 10 1110 -999 16.8 \n216427 -90.4266 -68.8656 19 1 0.011 170 270 -999 0.4 \n216428 37.7140 -46.8972 19 4 0.528 970 1170 -999 9.6"} {"chunk_id": "satellite_insitu-glaciers-extent_uncertainty_q01__cfe81b8108c9", "report_id": "satellite_insitu-glaciers-extent_uncertainty_q01", "dataset_id": "insitu-glaciers-extent", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty_q01", "aspect_base": "uncertainty", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates > Analysis and results > 1. Data preparation and processing > 1.3 Display and inspect data", "title": "Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates", "chunk_index": 7, "token_count": 691, "text_raw": "0.4 \n216428 37.7140 -46.8972 19 4 0.528 970 1170 -999 9.6\n\nASPECT LMAX NAME \\\nindex \n0 346 839 None \n1 162 1197 None \n2 175 2106 None \n3 195 4175 None \n4 181 2981 None \n... ... ... ... \n216424 315 255 None \n216425 200 1130 None \n216426 308 4329 None \n216427 122 106 AQ6C10200013 \n216428 35 -9 ZA6C40100001 Ice Plateau\n\ngeometry \nindex \n0 POLYGON ((-146.81804 63.69081, -146.81768 63.6... \n1 POLYGON ((-146.66354 63.40764, -146.66344 63.4... \n2 POLYGON ((-146.07232 63.38348, -146.07232 63.3... \n3 POLYGON ((-146.14895 63.37919, -146.14881 63.3... \n4 POLYGON ((-147.04307 63.55024, -147.04483 63.5... \n... ... \n216424 POLYGON ((-37.73275 -53.98779, -37.73328 -53.9... \n216425 POLYGON ((-36.13834 -54.82735, -36.13826 -54.8... \n216426 POLYGON ((-37.29309 -54.17506, -37.29308 -54.1... \n216427 POLYGON ((-90.42751 -68.8649, -90.42719 -68.86... \n216428 POLYGON ((37.7162 -46.89403, 37.71674 -46.894,...\n\n[216429 rows x 18 columns]\n```\n\nAs can be seen above, the data includes attribute information for each individual glacier (i.e. delineated polygon) in the vector-type dataset. Important for this notebook are the `BGNDATE` and `ENDDATE` columns (for version 6.0) or `src_date` (for version 7.0), which contain information about the time of delineation of the specific glacier. If for some glaciers (part of) the attributes are missing, the data are filled by `-9999999`. We will use this information below.\n\n(section-2)=", "text_with_prefix": "EQC Quality Assessment: \"Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates\"\nDataset: insitu-glaciers-extent [CDS]\nAspect: uncertainty_q01 | Category: Satellite_ECVs\nSection: Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates > Analysis and results > 1. Data preparation and processing > 1.3 Display and inspect data\n---\n0.4 \n216428 37.7140 -46.8972 19 4 0.528 970 1170 -999 9.6\n\nASPECT LMAX NAME \\\nindex \n0 346 839 None \n1 162 1197 None \n2 175 2106 None \n3 195 4175 None \n4 181 2981 None \n... ... ... ... \n216424 315 255 None \n216425 200 1130 None \n216426 308 4329 None \n216427 122 106 AQ6C10200013 \n216428 35 -9 ZA6C40100001 Ice Plateau\n\ngeometry \nindex \n0 POLYGON ((-146.81804 63.69081, -146.81768 63.6... \n1 POLYGON ((-146.66354 63.40764, -146.66344 63.4... \n2 POLYGON ((-146.07232 63.38348, -146.07232 63.3... \n3 POLYGON ((-146.14895 63.37919, -146.14881 63.3... \n4 POLYGON ((-147.04307 63.55024, -147.04483 63.5... \n... ... \n216424 POLYGON ((-37.73275 -53.98779, -37.73328 -53.9... \n216425 POLYGON ((-36.13834 -54.82735, -36.13826 -54.8... \n216426 POLYGON ((-37.29309 -54.17506, -37.29308 -54.1... \n216427 POLYGON ((-90.42751 -68.8649, -90.42719 -68.86... \n216428 POLYGON ((37.7162 -46.89403, 37.71674 -46.894,...\n\n[216429 rows x 18 columns]\n```\n\nAs can be seen above, the data includes attribute information for each individual glacier (i.e. delineated polygon) in the vector-type dataset. Important for this notebook are the `BGNDATE` and `ENDDATE` columns (for version 6.0) or `src_date` (for version 7.0), which contain information about the time of delineation of the specific glacier. If for some glaciers (part of) the attributes are missing, the data are filled by `-9999999`. We will use this information below.\n\n(section-2)="} {"chunk_id": "satellite_insitu-glaciers-extent_uncertainty_q01__2c2ff8b6a357", "report_id": "satellite_insitu-glaciers-extent_uncertainty_q01", "dataset_id": "insitu-glaciers-extent", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty_q01", "aspect_base": "uncertainty", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates > Analysis and results > 2. The (area-weighted) mean year of delineation of glacier data > 2.1 Determination of date of delineation for every glacier", "title": "Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates", "chunk_index": 8, "token_count": 428, "text_raw": "We can begin to answer the user question by extracting information from the attribute table of the downloaded shapefile. For some glaciers the delineation of the outline is composed from several scenes over multiple years in v6.0, implying that a begin date (attribute \"`BGNDATE`\") and an end date (attribute \"`ENDDATE`\") of the outline delineation is given in the attribute table in YYYYMMDD format (if both are available). For v7.0, the acquisition or delineation date is simply given by \"`src_date`\" in YYYY-MM-DD format:\n\n$\nyear_t = \\textstyle\\dfrac{BGNDATE_t + ENDDATE_t}{2}\\ \n$ (version 6.0), or\n\n$\nyear_t = srcdate_t\n$ (version 7.0)\n\nThis results in the following plot:\n\nConvert dates from string to datetime, and add digitalization year\nGet data from years\nGet some statistics\n\n*
Figure 1. Time of delineation (date of acquisition) for the glaciers in the glacier extent dataset.
*\n\nLet us calculate the percentage of years between some specific periods centered around the year 2000 to stronger emphasize that the large temporal deviations from the year 2000 are not frequent:\n\n```text\nIn the RGIv6.0 dataset, 92.2% of all acquisitions occurred between 1990 and 2010, of which 65.1% took place between 1995 and 2005.\n```\n\n(section-2-2)=", "text_with_prefix": "EQC Quality Assessment: \"Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates\"\nDataset: insitu-glaciers-extent [CDS]\nAspect: uncertainty_q01 | Category: Satellite_ECVs\nSection: Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates > Analysis and results > 2. The (area-weighted) mean year of delineation of glacier data > 2.1 Determination of date of delineation for every glacier\n---\nWe can begin to answer the user question by extracting information from the attribute table of the downloaded shapefile. For some glaciers the delineation of the outline is composed from several scenes over multiple years in v6.0, implying that a begin date (attribute \"`BGNDATE`\") and an end date (attribute \"`ENDDATE`\") of the outline delineation is given in the attribute table in YYYYMMDD format (if both are available). For v7.0, the acquisition or delineation date is simply given by \"`src_date`\" in YYYY-MM-DD format:\n\n$\nyear_t = \\textstyle\\dfrac{BGNDATE_t + ENDDATE_t}{2}\\ \n$ (version 6.0), or\n\n$\nyear_t = srcdate_t\n$ (version 7.0)\n\nThis results in the following plot:\n\nConvert dates from string to datetime, and add digitalization year\nGet data from years\nGet some statistics\n\n*
Figure 1. Time of delineation (date of acquisition) for the glaciers in the glacier extent dataset.
*\n\nLet us calculate the percentage of years between some specific periods centered around the year 2000 to stronger emphasize that the large temporal deviations from the year 2000 are not frequent:\n\n```text\nIn the RGIv6.0 dataset, 92.2% of all acquisitions occurred between 1990 and 2010, of which 65.1% took place between 1995 and 2005.\n```\n\n(section-2-2)="} {"chunk_id": "satellite_insitu-glaciers-extent_uncertainty_q01__46aae7dc5d7f", "report_id": "satellite_insitu-glaciers-extent_uncertainty_q01", "dataset_id": "insitu-glaciers-extent", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty_q01", "aspect_base": "uncertainty", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates > Analysis and results > 2. The (area-weighted) mean year of delineation of glacier data > 2.2 Spatial distribution of date of delineation", "title": "Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates", "chunk_index": 9, "token_count": 360, "text_raw": "Now that we calculated the acquisition dates, let us check the spatial distribution of the year of delineation for each glacier in the dataset. We therefore produce a world map where every dot represents a glacier, that is colored according to its year of delineation. Glaciers with missing data are not plotted:\n\n*
Figure 2. Spatial distribution of the time of delineation (date of acquisition) for the glaciers in the glacier extent dataset.
*\n\nFrom the above plots, it becomes clear that not all glaciers are delineated in the year 2000. Although the dataset is intended to be a snapshot of the world’s glaciers as they were near the beginning of the 21$^{st}$ century, the user must keep in mind that the range of delineation dates is substantial. This means that the outlines do not all lie in between the same time frame, which is a main issue of the dataset. The figures above, however, demonstrate that most outlines originate from the period between 2000 and 2010. Nevertheless, for some areas like the western United States or some peripheral glaciers around northern Greenland and the Antarctic Penninsula, the glacier outlines date back to the 1970s or earlier.\n\n(section-2-3)=", "text_with_prefix": "EQC Quality Assessment: \"Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates\"\nDataset: insitu-glaciers-extent [CDS]\nAspect: uncertainty_q01 | Category: Satellite_ECVs\nSection: Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates > Analysis and results > 2. The (area-weighted) mean year of delineation of glacier data > 2.2 Spatial distribution of date of delineation\n---\nNow that we calculated the acquisition dates, let us check the spatial distribution of the year of delineation for each glacier in the dataset. We therefore produce a world map where every dot represents a glacier, that is colored according to its year of delineation. Glaciers with missing data are not plotted:\n\n*
Figure 2. Spatial distribution of the time of delineation (date of acquisition) for the glaciers in the glacier extent dataset.
*\n\nFrom the above plots, it becomes clear that not all glaciers are delineated in the year 2000. Although the dataset is intended to be a snapshot of the world’s glaciers as they were near the beginning of the 21$^{st}$ century, the user must keep in mind that the range of delineation dates is substantial. This means that the outlines do not all lie in between the same time frame, which is a main issue of the dataset. The figures above, however, demonstrate that most outlines originate from the period between 2000 and 2010. Nevertheless, for some areas like the western United States or some peripheral glaciers around northern Greenland and the Antarctic Penninsula, the glacier outlines date back to the 1970s or earlier.\n\n(section-2-3)="} {"chunk_id": "satellite_insitu-glaciers-extent_uncertainty_q01__bf31efbfe825", "report_id": "satellite_insitu-glaciers-extent_uncertainty_q01", "dataset_id": "insitu-glaciers-extent", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty_q01", "aspect_base": "uncertainty", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates > Analysis and results > 2. The (area-weighted) mean year of delineation of glacier data > 2.3 The (area-weighted) mean year of delineation per RGI region", "title": "Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates", "chunk_index": 10, "token_count": 635, "text_raw": "Let us now calculate the (area-weighted) mean time of delineation for the 19 different RGI regions (see Figure 5 for the locations of these regions). We do that as follows:\n\n- Arithmetic mean:\n$\n\\overline{year}_a = \\textstyle\\dfrac{1}{{n\\in \\text{region}}}{\\sum\\limits_{i=1}^{n\\in \\text{region}} year_{t,i}}\n$\nwith \n$\nn\n$\nthe total number of glaciers and $year_t$ the time of delineation for glacier $i$ as calculated above.\n\n- Area-weighted mean:\n$\n\\overline{year}_w = \\left(\\dfrac{1}{\\sum\\limits_{i=1}^{n\\in \\text{region}} A_i}\\right) \\sum\\limits_{i=1}^{n\\in \\text{region}} \\left(A_i * year_{t,i}\\right)\n$\nwhere\n$\nA_i\n$\nis the glacier area for glacier $i$ [km²].\n\nThis results in the following plot:\n\nExtract region number from RGIID\nGroup by region and calculate total area\nCalculate means\nExtract region number from o1region\nGroup by region and calculate total area\nCalculate means\nEnsure regions are correctly defined as an array\nExtract the means\nDefine bar width and positions\nCreate the bar plot\nAdd labels, title, and legend\nAdd grid and show plot\n\n```text\nThe global arithmetic mean year of delineation (RGI v6.0) is 2002.8.\nThe global area-weighted mean year of delineation (RGI v6.0) is 2001.3.\n```\n\n*
Figure 3. Arithmetic and area-weighted mean year of delineation for the glaciers in each RGI region of the glacier extent dataset.
*\n\nWe observe that at the global scale, the arithmetic area-weighted average of the year of delineation is close to 2002, which only slightly differs in time from the nominally stated dataset reference time (i.e. 2000). However, the (area-weighted) average of the time of delineation varies notably across the different RGI regions: for RGI region 18, for example, the area-weighted average dates back to the late 1970s, but for RGI regions 1 and 10, the average origin dates are closer to the 2010s.\n\n(section-3)=", "text_with_prefix": "EQC Quality Assessment: \"Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates\"\nDataset: insitu-glaciers-extent [CDS]\nAspect: uncertainty_q01 | Category: Satellite_ECVs\nSection: Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates > Analysis and results > 2. The (area-weighted) mean year of delineation of glacier data > 2.3 The (area-weighted) mean year of delineation per RGI region\n---\nLet us now calculate the (area-weighted) mean time of delineation for the 19 different RGI regions (see Figure 5 for the locations of these regions). We do that as follows:\n\n- Arithmetic mean:\n$\n\\overline{year}_a = \\textstyle\\dfrac{1}{{n\\in \\text{region}}}{\\sum\\limits_{i=1}^{n\\in \\text{region}} year_{t,i}}\n$\nwith \n$\nn\n$\nthe total number of glaciers and $year_t$ the time of delineation for glacier $i$ as calculated above.\n\n- Area-weighted mean:\n$\n\\overline{year}_w = \\left(\\dfrac{1}{\\sum\\limits_{i=1}^{n\\in \\text{region}} A_i}\\right) \\sum\\limits_{i=1}^{n\\in \\text{region}} \\left(A_i * year_{t,i}\\right)\n$\nwhere\n$\nA_i\n$\nis the glacier area for glacier $i$ [km²].\n\nThis results in the following plot:\n\nExtract region number from RGIID\nGroup by region and calculate total area\nCalculate means\nExtract region number from o1region\nGroup by region and calculate total area\nCalculate means\nEnsure regions are correctly defined as an array\nExtract the means\nDefine bar width and positions\nCreate the bar plot\nAdd labels, title, and legend\nAdd grid and show plot\n\n```text\nThe global arithmetic mean year of delineation (RGI v6.0) is 2002.8.\nThe global area-weighted mean year of delineation (RGI v6.0) is 2001.3.\n```\n\n*
Figure 3. Arithmetic and area-weighted mean year of delineation for the glaciers in each RGI region of the glacier extent dataset.
*\n\nWe observe that at the global scale, the arithmetic area-weighted average of the year of delineation is close to 2002, which only slightly differs in time from the nominally stated dataset reference time (i.e. 2000). However, the (area-weighted) average of the time of delineation varies notably across the different RGI regions: for RGI region 18, for example, the area-weighted average dates back to the late 1970s, but for RGI regions 1 and 10, the average origin dates are closer to the 2010s.\n\n(section-3)="} {"chunk_id": "satellite_insitu-glaciers-extent_uncertainty_q01__86c09a3316c9", "report_id": "satellite_insitu-glaciers-extent_uncertainty_q01", "dataset_id": "insitu-glaciers-extent", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty_q01", "aspect_base": "uncertainty", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates > Analysis and results > 3. Estimation of global glacier ice volumes around the year 2000 with RGI data > 3.1 Recent glacier area and mass/volume change trends", "title": "Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates", "chunk_index": 11, "token_count": 718, "text_raw": "Going back to our specific use case, it is worth noting that several methods exist in the literature to derive ice thickness and ice volume estimates of glaciers at different spatial scales, of which the most well-known procedures are the volume-area scaling method, numerical modelling, and the estimation of ice thicknesses from surface characteristics and the principle of ice dynamics. These methods, however, strongly rely on the precise determination of the glacier surface area (and hypsometry) as input data to derive total ice volume estimates.\n\nFarinotti et al. (2019), for example, used the RGIv6.0 dataset to derive a global glacier ice volume of 158.17 ± 41.03 × 10³ km³, noting the heavy reliance on accurate glacier surface area data [[3](https://doi.org/10.1038/s41561-019-0300-3)]. Other studies, including Huss and Farinotti (2012), Marzeion et al. (2012) and GlaMBIE (2025), have similarly highlighted input data uncertainties with respect to the glacier area in ice volume estimates [[4](https://doi.org/10.1029/2012JF002523), [5](https://doi.org/10.5194/tc-6-1295-2012), [14](https://doi.org/10.1038/s41586-024-08545-z)]. The main concern here is that the glacierized area has decreased significantly during the last several decades, with global glacier area decreasing at rates of -0.18% to -0.34% per year and median retreat rates of 7.4 m per year [[6](https://doi.org/10.3189/2015JoG15J017), [7](https://doi.org/10.1016/j.accre.2020.03.003), [8](https://doi.org/10.5194/tc-8-659-2014)]. These changes suggest a potential overestimation of volumes for glaciers delineated before 2000 and an underestimation elsewhere. The suitability of RGI outlines for current volume estimates therefore primarily depends on (1) the deviation of the time since delineation of the outline from 2000 and (2) the significance of glacier area/volume changes during that period, which are influenced by the local climate and specific glacier geometries.\n\nLet us plot the regional glacier mass changes for the different RGI regions from the \"[Glacier mass change gridded data from 1976 to present derived from the Fluctuations of Glaciers Database](https://cds.climate.copernicus.eu/datasets/derived-gridded-glacier-mass-change?tab=overview)\" dataset on the CDS:\n\nMask data for RGI regions\nMask data for RGI regions\nCompute cumulative fields\nSelect first and last year explicitly (robust)\nTotal mass change per region", "text_with_prefix": "EQC Quality Assessment: \"Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates\"\nDataset: insitu-glaciers-extent [CDS]\nAspect: uncertainty_q01 | Category: Satellite_ECVs\nSection: Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates > Analysis and results > 3. Estimation of global glacier ice volumes around the year 2000 with RGI data > 3.1 Recent glacier area and mass/volume change trends\n---\nGoing back to our specific use case, it is worth noting that several methods exist in the literature to derive ice thickness and ice volume estimates of glaciers at different spatial scales, of which the most well-known procedures are the volume-area scaling method, numerical modelling, and the estimation of ice thicknesses from surface characteristics and the principle of ice dynamics. These methods, however, strongly rely on the precise determination of the glacier surface area (and hypsometry) as input data to derive total ice volume estimates.\n\nFarinotti et al. (2019), for example, used the RGIv6.0 dataset to derive a global glacier ice volume of 158.17 ± 41.03 × 10³ km³, noting the heavy reliance on accurate glacier surface area data [[3](https://doi.org/10.1038/s41561-019-0300-3)]. Other studies, including Huss and Farinotti (2012), Marzeion et al. (2012) and GlaMBIE (2025), have similarly highlighted input data uncertainties with respect to the glacier area in ice volume estimates [[4](https://doi.org/10.1029/2012JF002523), [5](https://doi.org/10.5194/tc-6-1295-2012), [14](https://doi.org/10.1038/s41586-024-08545-z)]. The main concern here is that the glacierized area has decreased significantly during the last several decades, with global glacier area decreasing at rates of -0.18% to -0.34% per year and median retreat rates of 7.4 m per year [[6](https://doi.org/10.3189/2015JoG15J017), [7](https://doi.org/10.1016/j.accre.2020.03.003), [8](https://doi.org/10.5194/tc-8-659-2014)]. These changes suggest a potential overestimation of volumes for glaciers delineated before 2000 and an underestimation elsewhere. The suitability of RGI outlines for current volume estimates therefore primarily depends on (1) the deviation of the time since delineation of the outline from 2000 and (2) the significance of glacier area/volume changes during that period, which are influenced by the local climate and specific glacier geometries.\n\nLet us plot the regional glacier mass changes for the different RGI regions from the \"[Glacier mass change gridded data from 1976 to present derived from the Fluctuations of Glaciers Database](https://cds.climate.copernicus.eu/datasets/derived-gridded-glacier-mass-change?tab=overview)\" dataset on the CDS:\n\nMask data for RGI regions\nMask data for RGI regions\nCompute cumulative fields\nSelect first and last year explicitly (robust)\nTotal mass change per region"} {"chunk_id": "satellite_insitu-glaciers-extent_uncertainty_q01__ac1ca1ca2a7c", "report_id": "satellite_insitu-glaciers-extent_uncertainty_q01", "dataset_id": "insitu-glaciers-extent", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty_q01", "aspect_base": "uncertainty", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates > Analysis and results > 3. Estimation of global glacier ice volumes around the year 2000 with RGI data > 3.1 Recent glacier area and mass/volume change trends", "title": "Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates", "chunk_index": 12, "token_count": 737, "text_raw": "RGI regions\nMask data for RGI regions\nCompute cumulative fields\nSelect first and last year explicitly (robust)\nTotal mass change per region\n\n```text\nThe total glacier mass change in RGI region 1 between 1976 and 2022 is -2445.24 Gt.\nThe total glacier mass change in RGI region 2 between 1976 and 2022 is -370.94 Gt.\nThe total glacier mass change in RGI region 3 between 1976 and 2022 is -889.74 Gt.\nThe total glacier mass change in RGI region 4 between 1976 and 2022 is -794.43 Gt.\nThe total glacier mass change in RGI region 5 between 1976 and 2022 is -675.07 Gt.\nThe total glacier mass change in RGI region 6 between 1976 and 2022 is -263.90 Gt.\nThe total glacier mass change in RGI region 7 between 1976 and 2022 is -424.88 Gt.\nThe total glacier mass change in RGI region 8 between 1976 and 2022 is -36.70 Gt.\nThe total glacier mass change in RGI region 9 between 1976 and 2022 is -404.35 Gt.\nThe total glacier mass change in RGI region 10 between 1976 and 2022 is -50.55 Gt.\nThe total glacier mass change in RGI region 11 between 1976 and 2022 is -45.34 Gt.\nThe total glacier mass change in RGI region 12 between 1976 and 2022 is -16.86 Gt.\nThe total glacier mass change in RGI region 13 between 1976 and 2022 is -415.13 Gt.\nThe total glacier mass change in RGI region 14 between 1976 and 2022 is -314.62 Gt.\nThe total glacier mass change in RGI region 15 between 1976 and 2022 is -278.79 Gt.\nThe total glacier mass change in RGI region 16 between 1976 and 2022 is -24.09 Gt.\nThe total glacier mass change in RGI region 17 between 1976 and 2022 is -899.90 Gt.\nThe total glacier mass change in RGI region 18 between 1976 and 2022 is -9.49 Gt.\nThe total glacier mass change in RGI region 19 between 1976 and 2022 is 601.53 Gt.\n```\n\nThe above emphasizes the strong negative glacier mass change trends that have been observed across the globe. It, however, creates a systematic bias when using the RGI data for the estimation of regional and global glacier volumes if the acquisition date does not coincide with the year 2000. Let us try to quantify this error with all of the information above.\n\n(section-3-2)=", "text_with_prefix": "EQC Quality Assessment: \"Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates\"\nDataset: insitu-glaciers-extent [CDS]\nAspect: uncertainty_q01 | Category: Satellite_ECVs\nSection: Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates > Analysis and results > 3. Estimation of global glacier ice volumes around the year 2000 with RGI data > 3.1 Recent glacier area and mass/volume change trends\n---\nRGI regions\nMask data for RGI regions\nCompute cumulative fields\nSelect first and last year explicitly (robust)\nTotal mass change per region\n\n```text\nThe total glacier mass change in RGI region 1 between 1976 and 2022 is -2445.24 Gt.\nThe total glacier mass change in RGI region 2 between 1976 and 2022 is -370.94 Gt.\nThe total glacier mass change in RGI region 3 between 1976 and 2022 is -889.74 Gt.\nThe total glacier mass change in RGI region 4 between 1976 and 2022 is -794.43 Gt.\nThe total glacier mass change in RGI region 5 between 1976 and 2022 is -675.07 Gt.\nThe total glacier mass change in RGI region 6 between 1976 and 2022 is -263.90 Gt.\nThe total glacier mass change in RGI region 7 between 1976 and 2022 is -424.88 Gt.\nThe total glacier mass change in RGI region 8 between 1976 and 2022 is -36.70 Gt.\nThe total glacier mass change in RGI region 9 between 1976 and 2022 is -404.35 Gt.\nThe total glacier mass change in RGI region 10 between 1976 and 2022 is -50.55 Gt.\nThe total glacier mass change in RGI region 11 between 1976 and 2022 is -45.34 Gt.\nThe total glacier mass change in RGI region 12 between 1976 and 2022 is -16.86 Gt.\nThe total glacier mass change in RGI region 13 between 1976 and 2022 is -415.13 Gt.\nThe total glacier mass change in RGI region 14 between 1976 and 2022 is -314.62 Gt.\nThe total glacier mass change in RGI region 15 between 1976 and 2022 is -278.79 Gt.\nThe total glacier mass change in RGI region 16 between 1976 and 2022 is -24.09 Gt.\nThe total glacier mass change in RGI region 17 between 1976 and 2022 is -899.90 Gt.\nThe total glacier mass change in RGI region 18 between 1976 and 2022 is -9.49 Gt.\nThe total glacier mass change in RGI region 19 between 1976 and 2022 is 601.53 Gt.\n```\n\nThe above emphasizes the strong negative glacier mass change trends that have been observed across the globe. It, however, creates a systematic bias when using the RGI data for the estimation of regional and global glacier volumes if the acquisition date does not coincide with the year 2000. Let us try to quantify this error with all of the information above.\n\n(section-3-2)="} {"chunk_id": "satellite_insitu-glaciers-extent_uncertainty_q01__29c3077f7699", "report_id": "satellite_insitu-glaciers-extent_uncertainty_q01", "dataset_id": "insitu-glaciers-extent", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty_q01", "aspect_base": "uncertainty", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates > Analysis and results > 3. Estimation of global glacier ice volumes around the year 2000 with RGI data > 3.2 An attempt to quantify potential glacier volume over or underestimation", "title": "Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates", "chunk_index": 13, "token_count": 901, "text_raw": "It is difficult to quantify the potential volume over or underestimation due to deviating outlines for each glacier in the dataset. However, we can get an idea of the regional glacier volume under/overestimation for 2000 CE by using the glacier mass change dataset that is available on the CDS. Our goal here is to obtain values for the regional glacier volume change between the year 2000 and the area-weighted mean year of delineation of that region $\\overline{year}_w$ (which is thus equivalent to the potential volume over or underestimation relative to the year 2000), denoted with the symbol $dV_{\\text{region}}^{2000,\\overline{\\text{year}}_w}$.\n\nTo get an idea of these potential glacier volume over- or underestimations, we can make use of the glacier mass change dataset that is on the CDS. This dataset estimates the gridded yearly mass change of glacier ice ($dM/dt$) in Gt yr$^{-1}$, which can be converted to glacier volume changes ($dV/dt$) with units of km$^3$ yr$^{-1}$ of ice by making use of the appropriate density. We thus calculate the mass and volume change between 2000 and the area-weighted average year of delineation for a certain RGI region ($\\overline{year}_w$) to get a total volume change for that specific RGI region between those two dates $dV_{\\text{region}}^{2000,\\overline{year}_w}$. If ${{\\overline{year}_w}}$ is a decimal (fractional) year, we use linear interpolation to find the interpolated glacier mass and volume change:\n\n$ \ndV_{\\text{region}}^{2000,\\overline{year}_w} \\text{ [km}^3\\text{]} = \n\\begin{cases} \n\\left(\\sum\\limits_{i=2000}^{\\lceil \\overline{year}_w \\rceil} \\sum\\limits^{\\substack{\\text{x,y} \\\\ \\text{{region}}}} \\left(\\frac{1000}{\\rho_i} \\frac{dM}{dt}\\right)\\right) - \\\\ \\left[ \\left( \\lceil \\overline{year}_w \\rceil - \\overline{year}_w \\right) * \\left( \\sum\\limits^{\\substack{\\text{x,y} \\\\ \\text{region}}} \\left(\\frac{1000}{\\rho_i} \\frac{dM}{dt}\\right)_{\\lceil \\overline{year}_w \\rceil} \\right) \\right] & \\text{if } \\overline{year}_w > 2000 \\\\ \n\\left(\\sum\\limits_{i=\\lfloor \\overline{year}_w \\rfloor}^{2000} \\sum\\limits^{\\substack{\\text{x,y} \\\\ \\text{region}}} \\left(\\frac{1000}{\\rho_i} \\frac{dM}{dt}\\right)\\right) - \\\\ \\left[ \\left( \\overline{year}_w - \\lfloor \\overline{year}_w \\rfloor \\right) * \\left( \\sum\\limits^{\\substack{\\text{x,y} \\\\ \\text{region}}} \\left(\\frac{1000}{\\rho_i} \\frac{dM}{dt}\\right)_{\\lfloor \\overline{year}_w \\rfloor} \\right) \\right] & \\text{if } \\overline{year}_w < 2000 \n\\end{cases} \n$", "text_with_prefix": "EQC Quality Assessment: \"Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates\"\nDataset: insitu-glaciers-extent [CDS]\nAspect: uncertainty_q01 | Category: Satellite_ECVs\nSection: Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates > Analysis and results > 3. Estimation of global glacier ice volumes around the year 2000 with RGI data > 3.2 An attempt to quantify potential glacier volume over or underestimation\n---\nIt is difficult to quantify the potential volume over or underestimation due to deviating outlines for each glacier in the dataset. However, we can get an idea of the regional glacier volume under/overestimation for 2000 CE by using the glacier mass change dataset that is available on the CDS. Our goal here is to obtain values for the regional glacier volume change between the year 2000 and the area-weighted mean year of delineation of that region $\\overline{year}_w$ (which is thus equivalent to the potential volume over or underestimation relative to the year 2000), denoted with the symbol $dV_{\\text{region}}^{2000,\\overline{\\text{year}}_w}$.\n\nTo get an idea of these potential glacier volume over- or underestimations, we can make use of the glacier mass change dataset that is on the CDS. This dataset estimates the gridded yearly mass change of glacier ice ($dM/dt$) in Gt yr$^{-1}$, which can be converted to glacier volume changes ($dV/dt$) with units of km$^3$ yr$^{-1}$ of ice by making use of the appropriate density. We thus calculate the mass and volume change between 2000 and the area-weighted average year of delineation for a certain RGI region ($\\overline{year}_w$) to get a total volume change for that specific RGI region between those two dates $dV_{\\text{region}}^{2000,\\overline{year}_w}$. If ${{\\overline{year}_w}}$ is a decimal (fractional) year, we use linear interpolation to find the interpolated glacier mass and volume change:\n\n$ \ndV_{\\text{region}}^{2000,\\overline{year}_w} \\text{ [km}^3\\text{]} = \n\\begin{cases} \n\\left(\\sum\\limits_{i=2000}^{\\lceil \\overline{year}_w \\rceil} \\sum\\limits^{\\substack{\\text{x,y} \\\\ \\text{{region}}}} \\left(\\frac{1000}{\\rho_i} \\frac{dM}{dt}\\right)\\right) - \\\\ \\left[ \\left( \\lceil \\overline{year}_w \\rceil - \\overline{year}_w \\right) * \\left( \\sum\\limits^{\\substack{\\text{x,y} \\\\ \\text{region}}} \\left(\\frac{1000}{\\rho_i} \\frac{dM}{dt}\\right)_{\\lceil \\overline{year}_w \\rceil} \\right) \\right] & \\text{if } \\overline{year}_w > 2000 \\\\ \n\\left(\\sum\\limits_{i=\\lfloor \\overline{year}_w \\rfloor}^{2000} \\sum\\limits^{\\substack{\\text{x,y} \\\\ \\text{region}}} \\left(\\frac{1000}{\\rho_i} \\frac{dM}{dt}\\right)\\right) - \\\\ \\left[ \\left( \\overline{year}_w - \\lfloor \\overline{year}_w \\rfloor \\right) * \\left( \\sum\\limits^{\\substack{\\text{x,y} \\\\ \\text{region}}} \\left(\\frac{1000}{\\rho_i} \\frac{dM}{dt}\\right)_{\\lfloor \\overline{year}_w \\rfloor} \\right) \\right] & \\text{if } \\overline{year}_w < 2000 \n\\end{cases} \n$"} {"chunk_id": "satellite_insitu-glaciers-extent_uncertainty_q01__d7a73d562aa7", "report_id": "satellite_insitu-glaciers-extent_uncertainty_q01", "dataset_id": "insitu-glaciers-extent", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty_q01", "aspect_base": "uncertainty", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates > Analysis and results > 3. Estimation of global glacier ice volumes around the year 2000 with RGI data > 3.2 An attempt to quantify potential glacier volume over or underestimation", "title": "Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates", "chunk_index": 14, "token_count": 1149, "text_raw": "{year}_w \\rfloor} \\right) \\right] & \\text{if } \\overline{year}_w < 2000 \n\\end{cases} \n$\n\nBasically, this comes down to \"reprojecting\" glacier volumes back to the year 2000 from the area-weighted year of data acquisition by making use of known glacier changes in between these two dates. Here, we further assume a density $\\rho_i$ of 900 kg m$^{-3}$ for the mass to volume conservation of glaciers, as also chosen by Farinotti et al. (2019) [[3](https://doi.org/10.1038/s41561-019-0300-3), [9](https://doi.org/10.5194/tc-7-877-2013)]. This may look like a complicated formula, but the principle is simple: for each RGI region, we simply sum the mass (and hence volume) changes over all spatial pixels between the year 2000 and the area-weighted year of delineation $(\\overline{year}_w)$:\n- If $(\\overline{year}_w)$ is an integer, the second part of the formula (after the minus sign) drops out.\n- If $(\\overline{year}_w)$ is not an integer, the formula adjusts for the fractional part of the year:\n - If $(\\overline{year}_w)$ > 2000, it uses the fraction between $(\\overline{year}_w)$ and $(\\lceil \\overline{year}_w \\rceil)$, where $\\lceil \\cdot \\rceil$ denotes the ceiling function, which rounds a number up to the nearest integer.\n - If $(\\overline{year}_w)$ < 2000, it uses the fraction between $(\\lfloor \\overline{year}_w \\rfloor)$ and $(\\overline{year}_w)$, where $\\lfloor \\cdot \\rfloor$ denotes the floor function, which rounds a number down to the nearest integer.\n\nNow that we have all information to derive values for $ dV_{\\text{region}}^{2000,\\overline{year}_w}$, let us apply this to the 19 different RGI regions in the RGIv7.0 data:\n\nCalculate volume under or overestimation\nPrinting loop\n\n```text\nThe volume estimate in RGI region 1 for the year 2000 is underestimated by 531.53 km³.\nThe volume estimate in RGI region 2 for the year 2000 is underestimated by 27.16 km³.\nThe volume estimate in RGI region 3 for the year 2000 is underestimated by 14.62 km³.\nThe volume estimate in RGI region 4 for the year 2000 is underestimated by 21.82 km³.\nThe volume estimate in RGI region 5 for the year 2000 is underestimated by 34.71 km³.\nThe volume estimate in RGI region 6 for the year 2000 is underestimated by 0.07 km³.\nThe volume estimate in RGI region 7 for the year 2000 is underestimated by 85.89 km³.\nThe volume estimate in RGI region 8 for the year 2000 is underestimated by 11.39 km³.\nThe volume estimate in RGI region 9 for the year 2000 is underestimated by 74.83 km³.\nThe volume estimate in RGI region 10 for the year 2000 is underestimated by 9.88 km³.\nThe volume estimate in RGI region 11 for the year 2000 is underestimated by 5.09 km³.\nThe volume estimate in RGI region 12 for the year 2000 is overestimated by 1.73 km³.\nThe volume estimate in RGI region 13 for the year 2000 is underestimated by 49.42 km³.\nThe volume estimate in RGI region 14 for the year 2000 is underestimated by 2.72 km³.\nThe volume estimate in RGI region 15 for the year 2000 is underestimated by 25.80 km³.\nThe volume estimate in RGI region 16 for the year 2000 is underestimated by 2.96 km³.\nThe volume estimate in RGI region 17 for the year 2000 is underestimated by 17.51 km³.\nThe volume estimate in RGI region 18 for the year 2000 is underestimated by 2.91 km³.\nThe volume estimate in RGI region 19 for the year 2000 is underestimated by 219.79 km³.\n\nThe volume estimate at the global scale for the year 2000 is underestimated by 1136.37 km³.\n```\n\nOr when plotted:\n\nArea-weighted estimate\nPlot 1 — Raw estimates\nPlot 2 — Area-weighted", "text_with_prefix": "EQC Quality Assessment: \"Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates\"\nDataset: insitu-glaciers-extent [CDS]\nAspect: uncertainty_q01 | Category: Satellite_ECVs\nSection: Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates > Analysis and results > 3. Estimation of global glacier ice volumes around the year 2000 with RGI data > 3.2 An attempt to quantify potential glacier volume over or underestimation\n---\n{year}_w \\rfloor} \\right) \\right] & \\text{if } \\overline{year}_w < 2000 \n\\end{cases} \n$\n\nBasically, this comes down to \"reprojecting\" glacier volumes back to the year 2000 from the area-weighted year of data acquisition by making use of known glacier changes in between these two dates. Here, we further assume a density $\\rho_i$ of 900 kg m$^{-3}$ for the mass to volume conservation of glaciers, as also chosen by Farinotti et al. (2019) [[3](https://doi.org/10.1038/s41561-019-0300-3), [9](https://doi.org/10.5194/tc-7-877-2013)]. This may look like a complicated formula, but the principle is simple: for each RGI region, we simply sum the mass (and hence volume) changes over all spatial pixels between the year 2000 and the area-weighted year of delineation $(\\overline{year}_w)$:\n- If $(\\overline{year}_w)$ is an integer, the second part of the formula (after the minus sign) drops out.\n- If $(\\overline{year}_w)$ is not an integer, the formula adjusts for the fractional part of the year:\n - If $(\\overline{year}_w)$ > 2000, it uses the fraction between $(\\overline{year}_w)$ and $(\\lceil \\overline{year}_w \\rceil)$, where $\\lceil \\cdot \\rceil$ denotes the ceiling function, which rounds a number up to the nearest integer.\n - If $(\\overline{year}_w)$ < 2000, it uses the fraction between $(\\lfloor \\overline{year}_w \\rfloor)$ and $(\\overline{year}_w)$, where $\\lfloor \\cdot \\rfloor$ denotes the floor function, which rounds a number down to the nearest integer.\n\nNow that we have all information to derive values for $ dV_{\\text{region}}^{2000,\\overline{year}_w}$, let us apply this to the 19 different RGI regions in the RGIv7.0 data:\n\nCalculate volume under or overestimation\nPrinting loop\n\n```text\nThe volume estimate in RGI region 1 for the year 2000 is underestimated by 531.53 km³.\nThe volume estimate in RGI region 2 for the year 2000 is underestimated by 27.16 km³.\nThe volume estimate in RGI region 3 for the year 2000 is underestimated by 14.62 km³.\nThe volume estimate in RGI region 4 for the year 2000 is underestimated by 21.82 km³.\nThe volume estimate in RGI region 5 for the year 2000 is underestimated by 34.71 km³.\nThe volume estimate in RGI region 6 for the year 2000 is underestimated by 0.07 km³.\nThe volume estimate in RGI region 7 for the year 2000 is underestimated by 85.89 km³.\nThe volume estimate in RGI region 8 for the year 2000 is underestimated by 11.39 km³.\nThe volume estimate in RGI region 9 for the year 2000 is underestimated by 74.83 km³.\nThe volume estimate in RGI region 10 for the year 2000 is underestimated by 9.88 km³.\nThe volume estimate in RGI region 11 for the year 2000 is underestimated by 5.09 km³.\nThe volume estimate in RGI region 12 for the year 2000 is overestimated by 1.73 km³.\nThe volume estimate in RGI region 13 for the year 2000 is underestimated by 49.42 km³.\nThe volume estimate in RGI region 14 for the year 2000 is underestimated by 2.72 km³.\nThe volume estimate in RGI region 15 for the year 2000 is underestimated by 25.80 km³.\nThe volume estimate in RGI region 16 for the year 2000 is underestimated by 2.96 km³.\nThe volume estimate in RGI region 17 for the year 2000 is underestimated by 17.51 km³.\nThe volume estimate in RGI region 18 for the year 2000 is underestimated by 2.91 km³.\nThe volume estimate in RGI region 19 for the year 2000 is underestimated by 219.79 km³.\n\nThe volume estimate at the global scale for the year 2000 is underestimated by 1136.37 km³.\n```\n\nOr when plotted:\n\nArea-weighted estimate\nPlot 1 — Raw estimates\nPlot 2 — Area-weighted"} {"chunk_id": "satellite_insitu-glaciers-extent_uncertainty_q01__582c1c4bfe47", "report_id": "satellite_insitu-glaciers-extent_uncertainty_q01", "dataset_id": "insitu-glaciers-extent", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty_q01", "aspect_base": "uncertainty", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates > Analysis and results > 3. Estimation of global glacier ice volumes around the year 2000 with RGI data > 3.2 An attempt to quantify potential glacier volume over or underestimation", "title": "Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates", "chunk_index": 15, "token_count": 698, "text_raw": "The volume estimate at the global scale for the year 2000 is underestimated by 1136.37 km³.\n```\n\nOr when plotted:\n\nArea-weighted estimate\nPlot 1 — Raw estimates\nPlot 2 — Area-weighted\n\n*Figure 4. Total (above) and relative, i.e. volume divided by the total glacier surface area, (below) volume over or underestimation compared to the year 2000 for the glaciers in each RGI region of the glacier extent dataset.*\n\nIn the plot above, negative values indicate an underestimation. Underestimations of glacier volumes are noted in all but one regions, which is because the (area-weighted) average origin dates of the outlines are generally later than 2000, together with a general trend of glacier shrinkage during the last several decades [[6](https://doi.org/10.3189/2015JoG15J017)]. It is clear that especially RGI regions 1 (Alaska) and 19 (Antarctic and Subantarctic) exhibit a relatively large underestimation of the total glacier ice volume, mainly due to their large deviation of the area-weighted average year of delineation to the year 2000, as well as the large glacier ice volume changes during this corresponding period. Even though the largest temporal deviation of acquisition dates is noted for RGI region 18 (New Zealand), the total volume underestimation is not significant due to the relatively small glacier changes.\n\nAt the global scale, glacier ice volume estimates would exhibit an underestimation of ca. 1150 km³ due to misdated glacier outlines, which is only ca. 0.75% and 0.85% of the total glacier ice volume estimate by Farinotti et al. (2019) and GlaMBIE (2025) respectively [[3](https://doi.org/10.1038/s41561-019-0300-3), [14](https://doi.org/10.1038/s41586-024-08545-z)]. At the global scale, the above obtained deviation related to the misdated glacier outlines from RGIv6.0 is thus relatively low when compared to the total global glacier ice volume estimate from the glacier community, implying that the dataset is well-suited to estimate current global glacier ice volumes for the year 2000. The corresponding systematic error is thus small, but glacier- and region-dependent.\n\n![alternatvie text](https://www.glims.org/rgi_user_guide/_images/global_map_small.jpeg)\n\n*Figure 5. The 19 first-order regions of the RGI version 7.0 and glacier locations in red. From: [GLIMS RGI](https://www.glims.org/rgi_user_guide/welcome.html).*\n\n(section-3-3)=", "text_with_prefix": "EQC Quality Assessment: \"Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates\"\nDataset: insitu-glaciers-extent [CDS]\nAspect: uncertainty_q01 | Category: Satellite_ECVs\nSection: Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates > Analysis and results > 3. Estimation of global glacier ice volumes around the year 2000 with RGI data > 3.2 An attempt to quantify potential glacier volume over or underestimation\n---\nThe volume estimate at the global scale for the year 2000 is underestimated by 1136.37 km³.\n```\n\nOr when plotted:\n\nArea-weighted estimate\nPlot 1 — Raw estimates\nPlot 2 — Area-weighted\n\n*Figure 4. Total (above) and relative, i.e. volume divided by the total glacier surface area, (below) volume over or underestimation compared to the year 2000 for the glaciers in each RGI region of the glacier extent dataset.*\n\nIn the plot above, negative values indicate an underestimation. Underestimations of glacier volumes are noted in all but one regions, which is because the (area-weighted) average origin dates of the outlines are generally later than 2000, together with a general trend of glacier shrinkage during the last several decades [[6](https://doi.org/10.3189/2015JoG15J017)]. It is clear that especially RGI regions 1 (Alaska) and 19 (Antarctic and Subantarctic) exhibit a relatively large underestimation of the total glacier ice volume, mainly due to their large deviation of the area-weighted average year of delineation to the year 2000, as well as the large glacier ice volume changes during this corresponding period. Even though the largest temporal deviation of acquisition dates is noted for RGI region 18 (New Zealand), the total volume underestimation is not significant due to the relatively small glacier changes.\n\nAt the global scale, glacier ice volume estimates would exhibit an underestimation of ca. 1150 km³ due to misdated glacier outlines, which is only ca. 0.75% and 0.85% of the total glacier ice volume estimate by Farinotti et al. (2019) and GlaMBIE (2025) respectively [[3](https://doi.org/10.1038/s41561-019-0300-3), [14](https://doi.org/10.1038/s41586-024-08545-z)]. At the global scale, the above obtained deviation related to the misdated glacier outlines from RGIv6.0 is thus relatively low when compared to the total global glacier ice volume estimate from the glacier community, implying that the dataset is well-suited to estimate current global glacier ice volumes for the year 2000. The corresponding systematic error is thus small, but glacier- and region-dependent.\n\n![alternatvie text](https://www.glims.org/rgi_user_guide/_images/global_map_small.jpeg)\n\n*Figure 5. The 19 first-order regions of the RGI version 7.0 and glacier locations in red. From: [GLIMS RGI](https://www.glims.org/rgi_user_guide/welcome.html).*\n\n(section-3-3)="} {"chunk_id": "satellite_insitu-glaciers-extent_uncertainty_q01__a82a1ad29b31", "report_id": "satellite_insitu-glaciers-extent_uncertainty_q01", "dataset_id": "insitu-glaciers-extent", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty_q01", "aspect_base": "uncertainty", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates > Analysis and results > 3. Estimation of global glacier ice volumes around the year 2000 with RGI data > 3.3 Other potential sources of error and uncertainty", "title": "Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates", "chunk_index": 16, "token_count": 673, "text_raw": "Several additional factors contribute to the uncertainty in glacier data derived from the RGI dataset. These include low-resolution and poor-quality source data, errors by delineators, and challenges such as shadows, clouds, proglacial lakes, seasonal snowfields, and limited multi-temporal satellite imagery. The most significant source of misdelineation is, however, supraglacial debris, which often results in glacier area underestimation in regions with extensive debris cover (e.g. High Mountain Asia and the Caucasus), due to the spectral similarity between debris and surrounding moraines or bedrock [[10](https://doi.org/10.3189/2013AoG63A296)]. Additionally, ice bodies smaller than 0.01 km², considered below the threshold for ice flow, were excluded, leading to further underestimation of glacier area in some regions [[1](https://www.glims.org/RGI/randolph60.html)].\n\nThe interpretation of what constitutes a glacier also varies due to the diverse global community that contributed to the dataset, causing inconsistencies in how features like debris-covered glaciers, tributaries, and peripheral ice bodies in Antarctica and Greenland were handled [[11](https://doi.org/10.1017/jog.2023.1)]. For example, distinguishing debris-covered ice from bedrock or separating outlet glaciers from main ice sheets remains problematic. Other problems are related to geolocation issues, outline artefacts, and ice divides at wrong locations.\n\nQuantitative error estimates are not provided with the downloaded data, but [[2](https://doi.org/10.3189/2014JoG13J176)] noted that relative errors are higher for small glaciers due to their larger outline-to-area ratio. Consequently, the uncertainty in an individual glacier outline depends on the quality of the source material and the specific glacier's characteristics. Users should thus exercise caution, particularly in regions with significant debris cover, small glacier features or around the peripheral glaciers of the Greenland and Antarctic ice sheets (where a clear separation of outlet glaciers from the main ice sheet body may be difficult).\n\nAlthough this notebook makes use of the RGIv6.0, some notable improvements have been made in the newer RGIv7.0 with respect to area and mapping year estimates. This has ensured that the quality of outlines has substantially improved in many regions due to, amongst others, the inclusion of new updated inventory data and a new largely automated workflow. For example, 35% of all RGIv6.0 outlines were dated to five or more years away from the target year 2000, while this number is down to 23% in RGIv7.0. It is therefore advised for users to utilize the RGIv7.0 in future assessments.\n\n(section-4)=", "text_with_prefix": "EQC Quality Assessment: \"Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates\"\nDataset: insitu-glaciers-extent [CDS]\nAspect: uncertainty_q01 | Category: Satellite_ECVs\nSection: Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates > Analysis and results > 3. Estimation of global glacier ice volumes around the year 2000 with RGI data > 3.3 Other potential sources of error and uncertainty\n---\nSeveral additional factors contribute to the uncertainty in glacier data derived from the RGI dataset. These include low-resolution and poor-quality source data, errors by delineators, and challenges such as shadows, clouds, proglacial lakes, seasonal snowfields, and limited multi-temporal satellite imagery. The most significant source of misdelineation is, however, supraglacial debris, which often results in glacier area underestimation in regions with extensive debris cover (e.g. High Mountain Asia and the Caucasus), due to the spectral similarity between debris and surrounding moraines or bedrock [[10](https://doi.org/10.3189/2013AoG63A296)]. Additionally, ice bodies smaller than 0.01 km², considered below the threshold for ice flow, were excluded, leading to further underestimation of glacier area in some regions [[1](https://www.glims.org/RGI/randolph60.html)].\n\nThe interpretation of what constitutes a glacier also varies due to the diverse global community that contributed to the dataset, causing inconsistencies in how features like debris-covered glaciers, tributaries, and peripheral ice bodies in Antarctica and Greenland were handled [[11](https://doi.org/10.1017/jog.2023.1)]. For example, distinguishing debris-covered ice from bedrock or separating outlet glaciers from main ice sheets remains problematic. Other problems are related to geolocation issues, outline artefacts, and ice divides at wrong locations.\n\nQuantitative error estimates are not provided with the downloaded data, but [[2](https://doi.org/10.3189/2014JoG13J176)] noted that relative errors are higher for small glaciers due to their larger outline-to-area ratio. Consequently, the uncertainty in an individual glacier outline depends on the quality of the source material and the specific glacier's characteristics. Users should thus exercise caution, particularly in regions with significant debris cover, small glacier features or around the peripheral glaciers of the Greenland and Antarctic ice sheets (where a clear separation of outlet glaciers from the main ice sheet body may be difficult).\n\nAlthough this notebook makes use of the RGIv6.0, some notable improvements have been made in the newer RGIv7.0 with respect to area and mapping year estimates. This has ensured that the quality of outlines has substantially improved in many regions due to, amongst others, the inclusion of new updated inventory data and a new largely automated workflow. For example, 35% of all RGIv6.0 outlines were dated to five or more years away from the target year 2000, while this number is down to 23% in RGIv7.0. It is therefore advised for users to utilize the RGIv7.0 in future assessments.\n\n(section-4)="} {"chunk_id": "satellite_insitu-glaciers-extent_uncertainty_q01__91942b244b91", "report_id": "satellite_insitu-glaciers-extent_uncertainty_q01", "dataset_id": "insitu-glaciers-extent", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty_q01", "aspect_base": "uncertainty", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates > Analysis and results > 4. Short summary and take-home messages", "title": "Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates", "chunk_index": 17, "token_count": 535, "text_raw": "The RGI glaciers distribution dataset that is on the CDS is currently the most complete dataset of glacier (and ice cap) outlines, outside of the Greenland and Antarctic Ice Sheets, in terms of its spatial coverage. It aims to capture the spatial distribution and global state of glacier outlines representative for the year 2000. However, the date of delineation (i.e. the time for which the outlines are representative) can significantly deviate from 2000, resulting in a systematic error.\n\nWhen using the data for certain applications, such as for quantifying current global glacier volumes with RGI glacier areas as input data, this sytematic error has to be taken into account. It is found that the suitability of a RGI outline for current glacier ice volume estimates hence depends on (1) the deviation of the time since delineation from 2000 and (2) the significance of glacier area/volume changes during that period, which are influenced by changes of the local climate and the specific glacier geometries. Nevertheless, the assessment in this notebook revealed that (area-weighted) mean acquisition date of the glacier outlines is relatively close to 2000 at the global scale, although a tendency for delineation dates that are slightly more recent than 2000 is noted. Moreover, when compared to the total global ice volume estimate by Farinotti et al. (2019) and GlaMBIE (2025) [[3](https://doi.org/10.1038/s41561-019-0300-3), [14](https://doi.org/10.1038/s41586-024-08545-z)], the resulting systematic error calculated in this notebook is in the order of <1% only. This makes the dataset reliable to be used as input data for current global glacier volume estimates, but the systematic error is glacier- and region-dependent.\n\nAdditional factors to take into account are, amongst others, supraglacial debris, which often leads to an underestimation of the glacierized area in highly debris-covered regions such as High Mountain Asia and the Caucasus. When fed into glaciological and/or hydrological models, users are thus advized to take into account these considerations.", "text_with_prefix": "EQC Quality Assessment: \"Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates\"\nDataset: insitu-glaciers-extent [CDS]\nAspect: uncertainty_q01 | Category: Satellite_ECVs\nSection: Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates > Analysis and results > 4. Short summary and take-home messages\n---\nThe RGI glaciers distribution dataset that is on the CDS is currently the most complete dataset of glacier (and ice cap) outlines, outside of the Greenland and Antarctic Ice Sheets, in terms of its spatial coverage. It aims to capture the spatial distribution and global state of glacier outlines representative for the year 2000. However, the date of delineation (i.e. the time for which the outlines are representative) can significantly deviate from 2000, resulting in a systematic error.\n\nWhen using the data for certain applications, such as for quantifying current global glacier volumes with RGI glacier areas as input data, this sytematic error has to be taken into account. It is found that the suitability of a RGI outline for current glacier ice volume estimates hence depends on (1) the deviation of the time since delineation from 2000 and (2) the significance of glacier area/volume changes during that period, which are influenced by changes of the local climate and the specific glacier geometries. Nevertheless, the assessment in this notebook revealed that (area-weighted) mean acquisition date of the glacier outlines is relatively close to 2000 at the global scale, although a tendency for delineation dates that are slightly more recent than 2000 is noted. Moreover, when compared to the total global ice volume estimate by Farinotti et al. (2019) and GlaMBIE (2025) [[3](https://doi.org/10.1038/s41561-019-0300-3), [14](https://doi.org/10.1038/s41586-024-08545-z)], the resulting systematic error calculated in this notebook is in the order of <1% only. This makes the dataset reliable to be used as input data for current global glacier volume estimates, but the systematic error is glacier- and region-dependent.\n\nAdditional factors to take into account are, amongst others, supraglacial debris, which often leads to an underestimation of the glacierized area in highly debris-covered regions such as High Mountain Asia and the Caucasus. When fed into glaciological and/or hydrological models, users are thus advized to take into account these considerations."} {"chunk_id": "satellite_insitu-glaciers-extent_uncertainty_q01__e1b0611c2025", "report_id": "satellite_insitu-glaciers-extent_uncertainty_q01", "dataset_id": "insitu-glaciers-extent", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty_q01", "aspect_base": "uncertainty", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates > ℹ️ If you want to know more > Key resources", "title": "Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates", "chunk_index": 18, "token_count": 311, "text_raw": "- [Glaciers distribution data from the Randolph Glacier Inventory (RGI) for year 2000](https://cds.climate.copernicus.eu/datasets/insitu-glaciers-extent?tab=overview)\n- [Documentation on the CDS](https://cds.climate.copernicus.eu/datasets/insitu-glaciers-extent?tab=overview) and the [ECMWF Confluence Wiki](https://confluence.ecmwf.int/display/CKB/Glacier+Area)\n- [Copernicus climate change indicators: glaciers](https://climate.copernicus.eu/climate-indicators/glaciers)\n- [RGI website from GLIMS](https://www.glims.org/RGI/) (Global Land Ice Measurements from Space)\n- [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control` prepared by [B-Open](https://www.bopen.eu).", "text_with_prefix": "EQC Quality Assessment: \"Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates\"\nDataset: insitu-glaciers-extent [CDS]\nAspect: uncertainty_q01 | Category: Satellite_ECVs\nSection: Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates > ℹ️ If you want to know more > Key resources\n---\n- [Glaciers distribution data from the Randolph Glacier Inventory (RGI) for year 2000](https://cds.climate.copernicus.eu/datasets/insitu-glaciers-extent?tab=overview)\n- [Documentation on the CDS](https://cds.climate.copernicus.eu/datasets/insitu-glaciers-extent?tab=overview) and the [ECMWF Confluence Wiki](https://confluence.ecmwf.int/display/CKB/Glacier+Area)\n- [Copernicus climate change indicators: glaciers](https://climate.copernicus.eu/climate-indicators/glaciers)\n- [RGI website from GLIMS](https://www.glims.org/RGI/) (Global Land Ice Measurements from Space)\n- [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control` prepared by [B-Open](https://www.bopen.eu)."} {"chunk_id": "satellite_insitu-glaciers-extent_uncertainty_q01__971307d584f6", "report_id": "satellite_insitu-glaciers-extent_uncertainty_q01", "dataset_id": "insitu-glaciers-extent", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty_q01", "aspect_base": "uncertainty", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates > ℹ️ If you want to know more > References", "title": "Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates", "chunk_index": 19, "token_count": 1051, "text_raw": "- [[1](https://www.glims.org/RGI/randolph60.html)] RGI Consortium (2017). Randolph Glacier Inventory – A Dataset of Global Glacier Outlines: Version 6.0: Technical Report, Global Land Ice Measurements from Space, Colorado, USA. Digital Media. doi: 10.7265/N5-RGI-60.\n\n- [[2](https://doi.org/10.3189/2014JoG13J176)] Pfeffer, W. T., Arendt, A. A., Bliss, A., Bolch, T., Cogley, J. G., Gardner, A. S., Hagen, J. O., Hock, R., Kaser, G., Kienholz, C., Miles, E. S., Moholdt, G., Mölg, N., Paul, F., Radić, V., Rastner, P., Raup, B. H., Rich, J., Sharp, M. J., and Glasser, N. (2014). The Randolph Glacier Inventory: A globally complete inventory of glaciers, Journal of Glaciology, 60(221), 537-552. doi: 10.3189/2014JoG13J176.\n\n- [[3](https://doi.org/10.1038/s41561-019-0300-3)] Farinotti, D., Huss, M., Fürst, J. J., Landmann, J., Machguth, H., Maussion, F., and Pandit, A. (2019). A consensus estimate for the ice thickness distribution of all glaciers on Earth. Nature Geoscience, 12(3), 168-173. doi: 10.1038/s41561-019-0300-3.\n\n- [[4](https://doi.org/10.1029/2012JF002523)] Huss, M., and Farinotti, D. (2012). Distributed ice thickness and volume of all glaciers around the globe, Journal of Geophysical Research, 117, F04010. doi: 10.1029/2012JF002523.\n\n- [[5](https://doi.org/10.5194/tc-6-1295-2012)] Marzeion, B., Jarosch, A. H., and Hofer, M. (2012). Past and future sea-level change from the surface mass balance of glaciers, The Cryosphere, 6, 1295–1322. doi: 10.5194/tc-6-1295-2012.\n\n- [[6](https://doi.org/10.3189/2015JoG15J017)] Zemp, M., Frey, H., Gärtnew-Roer, I., Nussbaumer, S. U., Helzle, M., Paul, F., Haeberli, W., Denzinger, F., Ahlstrøm, A. P., Anderson, B., Bajracharya, S., Baroni, C., Braun, L. N., Cáceres, B. E., Casassa, G., Cobos, G., Dávila, L. R., Delgado Granados, H., Demuth, M. N., Espizua, L., Fischer, A., Fujita, K., Gadek, B., Ghazanfar, A., Hagen, J. O., Holmlund, P., Karimi, N., Li, Z., Pelto, M., Pitte, P., Popovnin, V. V., Portocarrero, C. A., Prinz, R., Sangewar, C. V., Severskiy, I., Sigurdsson, O., Soruco, A., Usubaliev, R., and Vincent, C. (2015). Historically unprecedented global glacier decline in the early 21st century, Journal of Glaciology, 61, 745-762. doi: 10.3189/2015JoG15J017.\n\n- [[7](https://doi.org/10.1016/j.accre.2020.03.003)] Li, Y. J., Ding, Y. J., Shangguan, D. H., and Wang, R. J. (2019). Regional differences in global glacier retreat from 1980 to 2015, Advances in Climate Change Research, 10(4), 203–213. doi: 10.1016/j.accre.2020.03.003.", "text_with_prefix": "EQC Quality Assessment: \"Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates\"\nDataset: insitu-glaciers-extent [CDS]\nAspect: uncertainty_q01 | Category: Satellite_ECVs\nSection: Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates > ℹ️ If you want to know more > References\n---\n- [[1](https://www.glims.org/RGI/randolph60.html)] RGI Consortium (2017). Randolph Glacier Inventory – A Dataset of Global Glacier Outlines: Version 6.0: Technical Report, Global Land Ice Measurements from Space, Colorado, USA. Digital Media. doi: 10.7265/N5-RGI-60.\n\n- [[2](https://doi.org/10.3189/2014JoG13J176)] Pfeffer, W. T., Arendt, A. A., Bliss, A., Bolch, T., Cogley, J. G., Gardner, A. S., Hagen, J. O., Hock, R., Kaser, G., Kienholz, C., Miles, E. S., Moholdt, G., Mölg, N., Paul, F., Radić, V., Rastner, P., Raup, B. H., Rich, J., Sharp, M. J., and Glasser, N. (2014). The Randolph Glacier Inventory: A globally complete inventory of glaciers, Journal of Glaciology, 60(221), 537-552. doi: 10.3189/2014JoG13J176.\n\n- [[3](https://doi.org/10.1038/s41561-019-0300-3)] Farinotti, D., Huss, M., Fürst, J. J., Landmann, J., Machguth, H., Maussion, F., and Pandit, A. (2019). A consensus estimate for the ice thickness distribution of all glaciers on Earth. Nature Geoscience, 12(3), 168-173. doi: 10.1038/s41561-019-0300-3.\n\n- [[4](https://doi.org/10.1029/2012JF002523)] Huss, M., and Farinotti, D. (2012). Distributed ice thickness and volume of all glaciers around the globe, Journal of Geophysical Research, 117, F04010. doi: 10.1029/2012JF002523.\n\n- [[5](https://doi.org/10.5194/tc-6-1295-2012)] Marzeion, B., Jarosch, A. H., and Hofer, M. (2012). Past and future sea-level change from the surface mass balance of glaciers, The Cryosphere, 6, 1295–1322. doi: 10.5194/tc-6-1295-2012.\n\n- [[6](https://doi.org/10.3189/2015JoG15J017)] Zemp, M., Frey, H., Gärtnew-Roer, I., Nussbaumer, S. U., Helzle, M., Paul, F., Haeberli, W., Denzinger, F., Ahlstrøm, A. P., Anderson, B., Bajracharya, S., Baroni, C., Braun, L. N., Cáceres, B. E., Casassa, G., Cobos, G., Dávila, L. R., Delgado Granados, H., Demuth, M. N., Espizua, L., Fischer, A., Fujita, K., Gadek, B., Ghazanfar, A., Hagen, J. O., Holmlund, P., Karimi, N., Li, Z., Pelto, M., Pitte, P., Popovnin, V. V., Portocarrero, C. A., Prinz, R., Sangewar, C. V., Severskiy, I., Sigurdsson, O., Soruco, A., Usubaliev, R., and Vincent, C. (2015). Historically unprecedented global glacier decline in the early 21st century, Journal of Glaciology, 61, 745-762. doi: 10.3189/2015JoG15J017.\n\n- [[7](https://doi.org/10.1016/j.accre.2020.03.003)] Li, Y. J., Ding, Y. J., Shangguan, D. H., and Wang, R. J. (2019). Regional differences in global glacier retreat from 1980 to 2015, Advances in Climate Change Research, 10(4), 203–213. doi: 10.1016/j.accre.2020.03.003."} {"chunk_id": "satellite_insitu-glaciers-extent_uncertainty_q01__3c19e203aed4", "report_id": "satellite_insitu-glaciers-extent_uncertainty_q01", "dataset_id": "insitu-glaciers-extent", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty_q01", "aspect_base": "uncertainty", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates > ℹ️ If you want to know more > References", "title": "Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates", "chunk_index": 20, "token_count": 877, "text_raw": "differences in global glacier retreat from 1980 to 2015, Advances in Climate Change Research, 10(4), 203–213. doi: 10.1016/j.accre.2020.03.003.\n\n- [[8](https://doi.org/10.5194/tc-8-659-2014)] Leclercq, P. W., Oerlemans, J., Basagic, H. J., Bushueva, I., Cook, A. J., and Le Bris, R. (2014). A data set of worldwide glacier length fluctuations, Cryosphere, 2014(8), 659–672. doi: 10.5194/tc-8-659-2014.\n\n- [[9](https://doi.org/10.5194/tc-7-877-2013)] Huss, M. (2013). Density assumptions for converting geodetic glacier volume change to mass change, The Cryosphere, 7, 877–887, doi: 10.5194/tc-7-877-2013.\n\n- [[10](https://doi.org/10.3189/2013AoG63A296)] Paul, F., Barrand, N. E., Baumann, S., Berthier, E., Bolch, T., Casey, K., Frey, H., Joshi, S. P., Konovalov, V., Le Bris, R., Mölg, N., Nosenko, G., Nuth, C., Pope, A., Racoviteanu, A., Rastner, P., Raup, B., Scharrer, K., Steffen, S., and Winsvold, S. (2013). On the Accuracy of Glacier Outlines Derived from Remote-Sensing Data. Annals of Glaciology, 54(63), 171–82. doi: 10.3189/2013AoG63A296.\n\n- [[11](https://doi.org/10.1017/jog.2023.1)] Hock, R., Maussion, F., Marzeion, B., and Nowicki, S. (2022). What is the global glacier ice volume outside the ice sheets? Journal of Glaciology, 69(273), 204–10. doi: 10.1017/jog.2023.1.\n\n- [[12](https://doi.org/10.1017/jog.2021.28)] Li, Y. J., Li, F., Shangguan, D. H., Ding, Y. J. (2021). A New Global Gridded Glacier Dataset Based on the Randolph Glacier Inventory Version 6.0. Journal of Glaciology, 67 (2021), 773–76. doi: 10.1017/jog.2021.28.\n\n- [[13](https://doi.org/10.5194/tc-6-1483-2012)] Rastner, P., Bolch, T., Mölg, N., Machguth, H., Le Bris, R., and Paul, F. (2012). The first complete inventory of the local glaciers and ice caps on Greenland, The Cryosphere, 6, 1483–1495, https://doi.org/10.5194/tc-6-1483-2012\n\n- [[14](https://doi.org/10.1038/s41586-024-08545-z)] The GlaMBIE Team (2025). Community estimate of global glacier mass changes from 2000 to 2023. Nature 639, 382–388. https://doi.org/10.1038/s41586-024-08545-z", "text_with_prefix": "EQC Quality Assessment: \"Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates\"\nDataset: insitu-glaciers-extent [CDS]\nAspect: uncertainty_q01 | Category: Satellite_ECVs\nSection: Utility of the Randolph Glacier Inventory (RGI) for regional and global glacier volume estimates > ℹ️ If you want to know more > References\n---\ndifferences in global glacier retreat from 1980 to 2015, Advances in Climate Change Research, 10(4), 203–213. doi: 10.1016/j.accre.2020.03.003.\n\n- [[8](https://doi.org/10.5194/tc-8-659-2014)] Leclercq, P. W., Oerlemans, J., Basagic, H. J., Bushueva, I., Cook, A. J., and Le Bris, R. (2014). A data set of worldwide glacier length fluctuations, Cryosphere, 2014(8), 659–672. doi: 10.5194/tc-8-659-2014.\n\n- [[9](https://doi.org/10.5194/tc-7-877-2013)] Huss, M. (2013). Density assumptions for converting geodetic glacier volume change to mass change, The Cryosphere, 7, 877–887, doi: 10.5194/tc-7-877-2013.\n\n- [[10](https://doi.org/10.3189/2013AoG63A296)] Paul, F., Barrand, N. E., Baumann, S., Berthier, E., Bolch, T., Casey, K., Frey, H., Joshi, S. P., Konovalov, V., Le Bris, R., Mölg, N., Nosenko, G., Nuth, C., Pope, A., Racoviteanu, A., Rastner, P., Raup, B., Scharrer, K., Steffen, S., and Winsvold, S. (2013). On the Accuracy of Glacier Outlines Derived from Remote-Sensing Data. Annals of Glaciology, 54(63), 171–82. doi: 10.3189/2013AoG63A296.\n\n- [[11](https://doi.org/10.1017/jog.2023.1)] Hock, R., Maussion, F., Marzeion, B., and Nowicki, S. (2022). What is the global glacier ice volume outside the ice sheets? Journal of Glaciology, 69(273), 204–10. doi: 10.1017/jog.2023.1.\n\n- [[12](https://doi.org/10.1017/jog.2021.28)] Li, Y. J., Li, F., Shangguan, D. H., Ding, Y. J. (2021). A New Global Gridded Glacier Dataset Based on the Randolph Glacier Inventory Version 6.0. Journal of Glaciology, 67 (2021), 773–76. doi: 10.1017/jog.2021.28.\n\n- [[13](https://doi.org/10.5194/tc-6-1483-2012)] Rastner, P., Bolch, T., Mölg, N., Machguth, H., Le Bris, R., and Paul, F. (2012). The first complete inventory of the local glaciers and ice caps on Greenland, The Cryosphere, 6, 1483–1495, https://doi.org/10.5194/tc-6-1483-2012\n\n- [[14](https://doi.org/10.1038/s41586-024-08545-z)] The GlaMBIE Team (2025). Community estimate of global glacier mass changes from 2000 to 2023. Nature 639, 382–388. https://doi.org/10.1038/s41586-024-08545-z"} {"chunk_id": "satellite_satellite-greenland-ice-sheet-velocity_trend-assessment_q01__a1bb39c948d4", "report_id": "satellite_satellite-greenland-ice-sheet-velocity_trend-assessment_q01", "dataset_id": "satellite-greenland-ice-sheet-velocity", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q01", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Ice sheet velocity data from satellite observations: temporal and spatial coverage and data completeness for trend analysis in glaciological applications", "title": "Ice sheet velocity data from satellite observations: temporal and spatial coverage and data completeness for trend analysis in glaciological applications", "chunk_index": 0, "token_count": 137, "text_raw": "Production date: 31-05-2025\n\nDataset version: 1.3 to 1.5\n\nProduced by: Yoni Verhaegen and Philippe Huybrechts (Vrije Universiteit Brussel)", "text_with_prefix": "EQC Quality Assessment: \"Ice sheet velocity data from satellite observations: temporal and spatial coverage and data completeness for trend analysis in glaciological applications\"\nDataset: satellite-greenland-ice-sheet-velocity [CDS]\nAspect: trend-assessment_q01 | Category: Satellite_ECVs\nSection: Ice sheet velocity data from satellite observations: temporal and spatial coverage and data completeness for trend analysis in glaciological applications\n---\nProduction date: 31-05-2025\n\nDataset version: 1.3 to 1.5\n\nProduced by: Yoni Verhaegen and Philippe Huybrechts (Vrije Universiteit Brussel)"} {"chunk_id": "satellite_satellite-greenland-ice-sheet-velocity_trend-assessment_q01__e15ecd848690", "report_id": "satellite_satellite-greenland-ice-sheet-velocity_trend-assessment_q01", "dataset_id": "satellite-greenland-ice-sheet-velocity", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q01", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Ice sheet velocity data from satellite observations: temporal and spatial coverage and data completeness for trend analysis in glaciological applications > Quality assessment question", "title": "Ice sheet velocity data from satellite observations: temporal and spatial coverage and data completeness for trend analysis in glaciological applications", "chunk_index": 1, "token_count": 518, "text_raw": "* **\"Is the dataset of sufficient spatial/temporal resolution, coverage and completeness to derive multi-year trends of horizontal surface flow velocities and the associated patterns of solid ice discharge?\"**\n\nIce sheets are not static, but instead flow. The total velocity magnitude is the sum of the velocity related to internal ice deformation, and is complemented by basal sliding and bed deformation components. Mapping ice sheet surface flow velocities and their temporal changes provides key information for investigating the dynamic response of the ice sheets to climate change. Remote sensing techniques, such as the use of satellites, are an important feature to derive and study these flow velocities. Satellites are able to inspect directly and repeatedly large areas of ice, and, as such are able to detect the movement of its surface. Remote sensing techniques that use satellite data are therefore considered the only feasible manner to derive accurate surface velocities of the ice sheets on a regular basis. In this dataset, they are derived by applying offset tracking techniques using Sentinel-1 synthetic aperture radar (SAR) satellite data. The main principle of the \"[Ice sheet velocity for Antarctica and Greenland derived from satellite observations](https://cds.climate.copernicus.eu/datasets/satellite-greenland-ice-sheet-velocity?tab=overview)\" dataset therefore relies on the preservation of surface features or other detectable patterns (e.g. speckle) in between multiple image acquisition periods. The extraction of these features/patterns from images acquired over different time periods is used to detect their displacements, and with further processing this information can be used to derive velocity fields [[1](https://doi.org/10.3390/rs70709371)].\n\nThis notebook investigates how well the dataset on the CDS (here we use versions 1.3 to 1.5) can be used to derive mean values, (inter/intra-)annual variability and multi-year trends of ice flow velocities and the associated patterns of solid ice discharge. More specifically, the notebook evaluates whether the dataset is of sufficient maturity and quality for that purpose in terms of its spatio-temporal coverage and resolution and its data completeness.", "text_with_prefix": "EQC Quality Assessment: \"Ice sheet velocity data from satellite observations: temporal and spatial coverage and data completeness for trend analysis in glaciological applications\"\nDataset: satellite-greenland-ice-sheet-velocity [CDS]\nAspect: trend-assessment_q01 | Category: Satellite_ECVs\nSection: Ice sheet velocity data from satellite observations: temporal and spatial coverage and data completeness for trend analysis in glaciological applications > Quality assessment question\n---\n* **\"Is the dataset of sufficient spatial/temporal resolution, coverage and completeness to derive multi-year trends of horizontal surface flow velocities and the associated patterns of solid ice discharge?\"**\n\nIce sheets are not static, but instead flow. The total velocity magnitude is the sum of the velocity related to internal ice deformation, and is complemented by basal sliding and bed deformation components. Mapping ice sheet surface flow velocities and their temporal changes provides key information for investigating the dynamic response of the ice sheets to climate change. Remote sensing techniques, such as the use of satellites, are an important feature to derive and study these flow velocities. Satellites are able to inspect directly and repeatedly large areas of ice, and, as such are able to detect the movement of its surface. Remote sensing techniques that use satellite data are therefore considered the only feasible manner to derive accurate surface velocities of the ice sheets on a regular basis. In this dataset, they are derived by applying offset tracking techniques using Sentinel-1 synthetic aperture radar (SAR) satellite data. The main principle of the \"[Ice sheet velocity for Antarctica and Greenland derived from satellite observations](https://cds.climate.copernicus.eu/datasets/satellite-greenland-ice-sheet-velocity?tab=overview)\" dataset therefore relies on the preservation of surface features or other detectable patterns (e.g. speckle) in between multiple image acquisition periods. The extraction of these features/patterns from images acquired over different time periods is used to detect their displacements, and with further processing this information can be used to derive velocity fields [[1](https://doi.org/10.3390/rs70709371)].\n\nThis notebook investigates how well the dataset on the CDS (here we use versions 1.3 to 1.5) can be used to derive mean values, (inter/intra-)annual variability and multi-year trends of ice flow velocities and the associated patterns of solid ice discharge. More specifically, the notebook evaluates whether the dataset is of sufficient maturity and quality for that purpose in terms of its spatio-temporal coverage and resolution and its data completeness."} {"chunk_id": "satellite_satellite-greenland-ice-sheet-velocity_trend-assessment_q01__4c77eae5dda7", "report_id": "satellite_satellite-greenland-ice-sheet-velocity_trend-assessment_q01", "dataset_id": "satellite-greenland-ice-sheet-velocity", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q01", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Ice sheet velocity data from satellite observations: temporal and spatial coverage and data completeness for trend analysis in glaciological applications > Quality assessment statement", "title": "Ice sheet velocity data from satellite observations: temporal and spatial coverage and data completeness for trend analysis in glaciological applications", "chunk_index": 2, "token_count": 444, "text_raw": "These are the key outcomes of this assessment\n\n- Surface ice flow velocity (“displacement”) observation by satellites is a useful tool to grasp the overall patterns of flow dynamics of the ice sheets. The C3S dataset exhibits a high spatial resolution, a consistent annual temporal resolution, and a nearly full ice sheet-wide coverage of the Greenland Ice Sheet (GrIS), which complies with international proposed standards. The majority of pixels also exhibit uncertainty values that fall within proposed error thresholds by GCOS. The data are therefore particularly well-suited for visualizing the spatial distribution of horizontal ice flow and for solid ice discharge calculations (i.e. dynamic ice mass loss), but additional datasets (i.e. the grounding line position, the ice thickness at those positions, and a conversion factor to derive vertically averaged horizontal velocities) are required for the latter. \n- When using the ice flow velocity data for the GrIS for certain applications, such as for solid ice discharge estimations, users should be aware of the typical strengths and limitations of the C3S SAR offset tracking-based velocity products. For example, the short temporal span (from 2014–2015 onwards) limits its use for detecting long-term trends or reliable climate change signals. Additionally, the ice sheet margins often suffer from higher absolute error values. These regions are nevertheless crucial for solid ice discharge estimates. Another limitation of this product is its yearly temporal resolution, which only captures interannual changes and misses important seasonal variations, making it less suitable for certain detailed studies. This especially holds for the Greenland Ice Sheet, where some outlet glaciers exhibit significant seasonal variations with respect to their dynamics. If desired, the C3S data can be complemented with external datasets to increase its temporal resolution and time span. \n```", "text_with_prefix": "EQC Quality Assessment: \"Ice sheet velocity data from satellite observations: temporal and spatial coverage and data completeness for trend analysis in glaciological applications\"\nDataset: satellite-greenland-ice-sheet-velocity [CDS]\nAspect: trend-assessment_q01 | Category: Satellite_ECVs\nSection: Ice sheet velocity data from satellite observations: temporal and spatial coverage and data completeness for trend analysis in glaciological applications > Quality assessment statement\n---\nThese are the key outcomes of this assessment\n\n- Surface ice flow velocity (“displacement”) observation by satellites is a useful tool to grasp the overall patterns of flow dynamics of the ice sheets. The C3S dataset exhibits a high spatial resolution, a consistent annual temporal resolution, and a nearly full ice sheet-wide coverage of the Greenland Ice Sheet (GrIS), which complies with international proposed standards. The majority of pixels also exhibit uncertainty values that fall within proposed error thresholds by GCOS. The data are therefore particularly well-suited for visualizing the spatial distribution of horizontal ice flow and for solid ice discharge calculations (i.e. dynamic ice mass loss), but additional datasets (i.e. the grounding line position, the ice thickness at those positions, and a conversion factor to derive vertically averaged horizontal velocities) are required for the latter. \n- When using the ice flow velocity data for the GrIS for certain applications, such as for solid ice discharge estimations, users should be aware of the typical strengths and limitations of the C3S SAR offset tracking-based velocity products. For example, the short temporal span (from 2014–2015 onwards) limits its use for detecting long-term trends or reliable climate change signals. Additionally, the ice sheet margins often suffer from higher absolute error values. These regions are nevertheless crucial for solid ice discharge estimates. Another limitation of this product is its yearly temporal resolution, which only captures interannual changes and misses important seasonal variations, making it less suitable for certain detailed studies. This especially holds for the Greenland Ice Sheet, where some outlet glaciers exhibit significant seasonal variations with respect to their dynamics. If desired, the C3S data can be complemented with external datasets to increase its temporal resolution and time span. \n```"} {"chunk_id": "satellite_satellite-greenland-ice-sheet-velocity_trend-assessment_q01__1f1473316cc9", "report_id": "satellite_satellite-greenland-ice-sheet-velocity_trend-assessment_q01", "dataset_id": "satellite-greenland-ice-sheet-velocity", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q01", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Ice sheet velocity data from satellite observations: temporal and spatial coverage and data completeness for trend analysis in glaciological applications > Methodology > Dataset description", "title": "Ice sheet velocity data from satellite observations: temporal and spatial coverage and data completeness for trend analysis in glaciological applications", "chunk_index": 3, "token_count": 328, "text_raw": "The 'Ice sheet velocity for Antarctica and Greenland derived from satellite observations' dataset, available on the Climate Data Store (CDS), offers annually averaged surface ice flow velocities and their easting and northing components for the Antarctic (AIS) and Greenland (GrIS) Ice Sheets. This data is provided on a spatial resolution grid of either 200 m, 250 m or 500 m, depending on the ice sheet and dataset version. The dataset includes horizontal velocity components in the east and north directions, as well as the total horizontal velocity magnitude and its uncertainty, expressed as 1-sigma precission errors. Surface flow velocities are treated as “displacements”, normalized to true meters per day, and are available for each glaciological year since 2014-2015 for the GrIS and since 2021-2022 for the AIS. The horizontal velocities and their components are derived using offset tracking techniques with Sentinel-1 synthetic aperture radar (SAR) satellite data. The data are provided in NetCDF format as gridded data fields in Polar Stereographic projection, covering the entire GrIS domain, including peripheral glaciers and ice caps.", "text_with_prefix": "EQC Quality Assessment: \"Ice sheet velocity data from satellite observations: temporal and spatial coverage and data completeness for trend analysis in glaciological applications\"\nDataset: satellite-greenland-ice-sheet-velocity [CDS]\nAspect: trend-assessment_q01 | Category: Satellite_ECVs\nSection: Ice sheet velocity data from satellite observations: temporal and spatial coverage and data completeness for trend analysis in glaciological applications > Methodology > Dataset description\n---\nThe 'Ice sheet velocity for Antarctica and Greenland derived from satellite observations' dataset, available on the Climate Data Store (CDS), offers annually averaged surface ice flow velocities and their easting and northing components for the Antarctic (AIS) and Greenland (GrIS) Ice Sheets. This data is provided on a spatial resolution grid of either 200 m, 250 m or 500 m, depending on the ice sheet and dataset version. The dataset includes horizontal velocity components in the east and north directions, as well as the total horizontal velocity magnitude and its uncertainty, expressed as 1-sigma precission errors. Surface flow velocities are treated as “displacements”, normalized to true meters per day, and are available for each glaciological year since 2014-2015 for the GrIS and since 2021-2022 for the AIS. The horizontal velocities and their components are derived using offset tracking techniques with Sentinel-1 synthetic aperture radar (SAR) satellite data. The data are provided in NetCDF format as gridded data fields in Polar Stereographic projection, covering the entire GrIS domain, including peripheral glaciers and ice caps."} {"chunk_id": "satellite_satellite-greenland-ice-sheet-velocity_trend-assessment_q01__51bd41103973", "report_id": "satellite_satellite-greenland-ice-sheet-velocity_trend-assessment_q01", "dataset_id": "satellite-greenland-ice-sheet-velocity", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q01", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Ice sheet velocity data from satellite observations: temporal and spatial coverage and data completeness for trend analysis in glaciological applications > Methodology > Structure and (sub)sections", "title": "Ice sheet velocity data from satellite observations: temporal and spatial coverage and data completeness for trend analysis in glaciological applications", "chunk_index": 4, "token_count": 212, "text_raw": "**[](section-1)**\n\n* [](section-1-1)\n* [](section-1-2)\n* [](section-1-3)\n* [](section-1-4)\n\n**[](section-2)**\n* [](section-2-1)\n* [](section-2-2)\n\n**[](section-3)**\n* [](section-3-1)\n* [](section-3-2)\n\n**[](section-4)**\n* [](section-4-1)\n* [](section-4-2)\n\n**[](section-5)**", "text_with_prefix": "EQC Quality Assessment: \"Ice sheet velocity data from satellite observations: temporal and spatial coverage and data completeness for trend analysis in glaciological applications\"\nDataset: satellite-greenland-ice-sheet-velocity [CDS]\nAspect: trend-assessment_q01 | Category: Satellite_ECVs\nSection: Ice sheet velocity data from satellite observations: temporal and spatial coverage and data completeness for trend analysis in glaciological applications > Methodology > Structure and (sub)sections\n---\n**[](section-1)**\n\n* [](section-1-1)\n* [](section-1-2)\n* [](section-1-3)\n* [](section-1-4)\n\n**[](section-2)**\n* [](section-2-1)\n* [](section-2-2)\n\n**[](section-3)**\n* [](section-3-1)\n* [](section-3-2)\n\n**[](section-4)**\n* [](section-4-1)\n* [](section-4-2)\n\n**[](section-5)**"} {"chunk_id": "satellite_satellite-greenland-ice-sheet-velocity_trend-assessment_q01__210d483c8de3", "report_id": "satellite_satellite-greenland-ice-sheet-velocity_trend-assessment_q01", "dataset_id": "satellite-greenland-ice-sheet-velocity", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q01", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Ice sheet velocity data from satellite observations: temporal and spatial coverage and data completeness for trend analysis in glaciological applications > Analysis and results > 1. Data preparation and processing > 1.2 Define request and download", "title": "Ice sheet velocity data from satellite observations: temporal and spatial coverage and data completeness for trend analysis in glaciological applications", "chunk_index": 5, "token_count": 446, "text_raw": "Then we define requests for download from the CDS and download the GrIS velocity data.\n\n🚨 **The files can be large! Since the data files to be downloaded have a considerable size due to the high spatial resolution, this may take a couple of minutes.**\n\n```text\nDownloading ice sheet velocity data, this may take a couple of minutes...\n```\n\n```text\n100%|██████████| 1/1 [00:00<00:00, 8.78it/s]\n100%|██████████| 1/1 [00:00<00:00, 86.43it/s]\n100%|██████████| 1/1 [00:00<00:00, 27.16it/s]\n100%|██████████| 1/1 [00:00<00:00, 98.45it/s]\n100%|██████████| 1/1 [00:00<00:00, 76.81it/s]\n100%|██████████| 1/1 [00:00<00:00, 108.28it/s]\n100%|██████████| 1/1 [00:00<00:00, 116.87it/s]\n100%|██████████| 1/1 [00:00<00:00, 108.75it/s]\n100%|██████████| 1/1 [00:00<00:00, 114.07it/s]\n```\n\n```text\nDownloading completed.\n```\n\n(section-1-3)=", "text_with_prefix": "EQC Quality Assessment: \"Ice sheet velocity data from satellite observations: temporal and spatial coverage and data completeness for trend analysis in glaciological applications\"\nDataset: satellite-greenland-ice-sheet-velocity [CDS]\nAspect: trend-assessment_q01 | Category: Satellite_ECVs\nSection: Ice sheet velocity data from satellite observations: temporal and spatial coverage and data completeness for trend analysis in glaciological applications > Analysis and results > 1. Data preparation and processing > 1.2 Define request and download\n---\nThen we define requests for download from the CDS and download the GrIS velocity data.\n\n🚨 **The files can be large! Since the data files to be downloaded have a considerable size due to the high spatial resolution, this may take a couple of minutes.**\n\n```text\nDownloading ice sheet velocity data, this may take a couple of minutes...\n```\n\n```text\n100%|██████████| 1/1 [00:00<00:00, 8.78it/s]\n100%|██████████| 1/1 [00:00<00:00, 86.43it/s]\n100%|██████████| 1/1 [00:00<00:00, 27.16it/s]\n100%|██████████| 1/1 [00:00<00:00, 98.45it/s]\n100%|██████████| 1/1 [00:00<00:00, 76.81it/s]\n100%|██████████| 1/1 [00:00<00:00, 108.28it/s]\n100%|██████████| 1/1 [00:00<00:00, 116.87it/s]\n100%|██████████| 1/1 [00:00<00:00, 108.75it/s]\n100%|██████████| 1/1 [00:00<00:00, 114.07it/s]\n```\n\n```text\nDownloading completed.\n```\n\n(section-1-3)="} {"chunk_id": "satellite_satellite-greenland-ice-sheet-velocity_trend-assessment_q01__7607b31399a3", "report_id": "satellite_satellite-greenland-ice-sheet-velocity_trend-assessment_q01", "dataset_id": "satellite-greenland-ice-sheet-velocity", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q01", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Ice sheet velocity data from satellite observations: temporal and spatial coverage and data completeness for trend analysis in glaciological applications > Analysis and results > 1. Data preparation and processing > 1.3 Display and inspect data structure", "title": "Ice sheet velocity data from satellite observations: temporal and spatial coverage and data completeness for trend analysis in glaciological applications", "chunk_index": 6, "token_count": 1105, "text_raw": "We can read and inspect the data. Let us print out the data to inspect its structure:\n\n```text\n Size: 19GB\nDimensions: (period: 9, y: 10801, x: 5984)\nCoordinates:\n * period (period) object 72B '2014_2015' ... ...\n * y (y) float64 86kB -6.556e+05 ... -3.3...\n * x (x) float64 48kB -6.399e+05 ... 8.55...\nData variables:\n crs (period) int32 36B -2147483647 ... -...\n land_ice_surface_easting_velocity (period, y, x) float32 2GB dask.array\n land_ice_surface_northing_velocity (period, y, x) float32 2GB dask.array\n land_ice_surface_vertical_velocity (period, y, x) float32 2GB dask.array\n land_ice_surface_velocity_magnitude (period, y, x) float32 2GB dask.array\n land_ice_surface_measurement_count (period, y, x) float64 5GB dask.array\n land_ice_surface_easting_stddev (period, y, x) float32 2GB dask.array\n land_ice_surface_northing_stddev (period, y, x) float32 2GB dask.array\nAttributes: (12/13)\n Conventions: CF-1.7\n title: Ice Velocity of the Greenland Ice Sheet\n reference: Main: Nagler, T.; Rott, H.; Hetzenecker, M.; Wuite, J.; P...\n source: Copernicus Sentinel-1A and Sentinel-1B\n institution: Copernicus Climate Change Service\n contact: copernicus-support@ecmwf.int\n ... ...\n creation_date: 2024-12-12\n comment: Ice velocity map of Greenland derived from Sentinel-1 SAR...\n history: product version 1.3\n summary: Ice velocity derived for Greenland Ice Sheet gridded at 2...\n keywords: EARTH SCIENCE CLIMATE INDICATORS CRYOSPHERIC INDICATORS G...\n license: C3S general license\n```\n\nThe versions 1.3 to 1.5 are a gridded dataset at a 250 m spatial resolution containing annually averaged values of the ice sheet horizontal surface flow velocity $V_s$ (in m/day) of a grid cell (`land_ice_surface_velocity_magnitude`) and its components (`land_ice_surface_easting_velocity` ($V_x$) and `land_ice_surface_northing_velocity` ($V_y$)) since the 2014-2015 hydrological year. The uncertainties (reported as precision errors or standard deviations) of the horizontal components are also given as `land_ice_surface_easting_stddev` ($\\sigma_{V_x}$) and `land_ice_surface_northing_stddev` ($\\sigma_{V_y}$). Also the vertical component is given as `land_ice_surface_vertical_velocity` ($V_z$), but without a corresponding uncertainty. Next to that, also a variable named `land_ice_surface_measurement_count` is present, which quantifies the total number of valid image pairs used in the annually averaged velocity estimate for a certain pixel.\n\nFor clarification, the total horizontal surface velocity magnitude $V_s$ and its uncertainty $\\sigma_{V_s}$ are calculated from its components as follows (with subscript 's' referring to the surface):\n\n$\nV_s\n$\n[m day⁻¹] \n$\n= \\sqrt {V_x^2 + V_y^2}\n$\n\n$\n\\sigma_{V_s}\n$\n[m day⁻¹] \n$\n= \\sqrt {(\\sigma_{V_x})^2 + (\\sigma_{V_y})^2}\n$\n\nwhere $V_x$ and $V_y$ are respectively the easting and northing components of the horizontal velocity vector.", "text_with_prefix": "EQC Quality Assessment: \"Ice sheet velocity data from satellite observations: temporal and spatial coverage and data completeness for trend analysis in glaciological applications\"\nDataset: satellite-greenland-ice-sheet-velocity [CDS]\nAspect: trend-assessment_q01 | Category: Satellite_ECVs\nSection: Ice sheet velocity data from satellite observations: temporal and spatial coverage and data completeness for trend analysis in glaciological applications > Analysis and results > 1. Data preparation and processing > 1.3 Display and inspect data structure\n---\nWe can read and inspect the data. Let us print out the data to inspect its structure:\n\n```text\n Size: 19GB\nDimensions: (period: 9, y: 10801, x: 5984)\nCoordinates:\n * period (period) object 72B '2014_2015' ... ...\n * y (y) float64 86kB -6.556e+05 ... -3.3...\n * x (x) float64 48kB -6.399e+05 ... 8.55...\nData variables:\n crs (period) int32 36B -2147483647 ... -...\n land_ice_surface_easting_velocity (period, y, x) float32 2GB dask.array\n land_ice_surface_northing_velocity (period, y, x) float32 2GB dask.array\n land_ice_surface_vertical_velocity (period, y, x) float32 2GB dask.array\n land_ice_surface_velocity_magnitude (period, y, x) float32 2GB dask.array\n land_ice_surface_measurement_count (period, y, x) float64 5GB dask.array\n land_ice_surface_easting_stddev (period, y, x) float32 2GB dask.array\n land_ice_surface_northing_stddev (period, y, x) float32 2GB dask.array\nAttributes: (12/13)\n Conventions: CF-1.7\n title: Ice Velocity of the Greenland Ice Sheet\n reference: Main: Nagler, T.; Rott, H.; Hetzenecker, M.; Wuite, J.; P...\n source: Copernicus Sentinel-1A and Sentinel-1B\n institution: Copernicus Climate Change Service\n contact: copernicus-support@ecmwf.int\n ... ...\n creation_date: 2024-12-12\n comment: Ice velocity map of Greenland derived from Sentinel-1 SAR...\n history: product version 1.3\n summary: Ice velocity derived for Greenland Ice Sheet gridded at 2...\n keywords: EARTH SCIENCE CLIMATE INDICATORS CRYOSPHERIC INDICATORS G...\n license: C3S general license\n```\n\nThe versions 1.3 to 1.5 are a gridded dataset at a 250 m spatial resolution containing annually averaged values of the ice sheet horizontal surface flow velocity $V_s$ (in m/day) of a grid cell (`land_ice_surface_velocity_magnitude`) and its components (`land_ice_surface_easting_velocity` ($V_x$) and `land_ice_surface_northing_velocity` ($V_y$)) since the 2014-2015 hydrological year. The uncertainties (reported as precision errors or standard deviations) of the horizontal components are also given as `land_ice_surface_easting_stddev` ($\\sigma_{V_x}$) and `land_ice_surface_northing_stddev` ($\\sigma_{V_y}$). Also the vertical component is given as `land_ice_surface_vertical_velocity` ($V_z$), but without a corresponding uncertainty. Next to that, also a variable named `land_ice_surface_measurement_count` is present, which quantifies the total number of valid image pairs used in the annually averaged velocity estimate for a certain pixel.\n\nFor clarification, the total horizontal surface velocity magnitude $V_s$ and its uncertainty $\\sigma_{V_s}$ are calculated from its components as follows (with subscript 's' referring to the surface):\n\n$\nV_s\n$\n[m day⁻¹] \n$\n= \\sqrt {V_x^2 + V_y^2}\n$\n\n$\n\\sigma_{V_s}\n$\n[m day⁻¹] \n$\n= \\sqrt {(\\sigma_{V_x})^2 + (\\sigma_{V_y})^2}\n$\n\nwhere $V_x$ and $V_y$ are respectively the easting and northing components of the horizontal velocity vector."} {"chunk_id": "satellite_satellite-greenland-ice-sheet-velocity_trend-assessment_q01__915410345846", "report_id": "satellite_satellite-greenland-ice-sheet-velocity_trend-assessment_q01", "dataset_id": "satellite-greenland-ice-sheet-velocity", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q01", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Ice sheet velocity data from satellite observations: temporal and spatial coverage and data completeness for trend analysis in glaciological applications > Analysis and results > 1. Data preparation and processing > 1.3 Display and inspect data structure", "title": "Ice sheet velocity data from satellite observations: temporal and spatial coverage and data completeness for trend analysis in glaciological applications", "chunk_index": 7, "token_count": 241, "text_raw": "$\n= \\sqrt {(\\sigma_{V_x})^2 + (\\sigma_{V_y})^2}\n$\n\nwhere $V_x$ and $V_y$ are respectively the easting and northing components of the horizontal velocity vector.\n\nThe flow direction of the ice can be calculated as follows:\n\n$\n\\theta_{\\text{V}_s} = \\left( 90^\\circ - \\left( \\arctan2(v_y, v_x) \\times \\left( \\frac{180}{\\pi} \\right) \\right) \\right) \\mod 360^\\circ\n$\n\n(section-1-4)=", "text_with_prefix": "EQC Quality Assessment: \"Ice sheet velocity data from satellite observations: temporal and spatial coverage and data completeness for trend analysis in glaciological applications\"\nDataset: satellite-greenland-ice-sheet-velocity [CDS]\nAspect: trend-assessment_q01 | Category: Satellite_ECVs\nSection: Ice sheet velocity data from satellite observations: temporal and spatial coverage and data completeness for trend analysis in glaciological applications > Analysis and results > 1. Data preparation and processing > 1.3 Display and inspect data structure\n---\n$\n= \\sqrt {(\\sigma_{V_x})^2 + (\\sigma_{V_y})^2}\n$\n\nwhere $V_x$ and $V_y$ are respectively the easting and northing components of the horizontal velocity vector.\n\nThe flow direction of the ice can be calculated as follows:\n\n$\n\\theta_{\\text{V}_s} = \\left( 90^\\circ - \\left( \\arctan2(v_y, v_x) \\times \\left( \\frac{180}{\\pi} \\right) \\right) \\right) \\mod 360^\\circ\n$\n\n(section-1-4)="} {"chunk_id": "satellite_satellite-greenland-ice-sheet-velocity_trend-assessment_q01__91f8cd32f355", "report_id": "satellite_satellite-greenland-ice-sheet-velocity_trend-assessment_q01", "dataset_id": "satellite-greenland-ice-sheet-velocity", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q01", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Ice sheet velocity data from satellite observations: temporal and spatial coverage and data completeness for trend analysis in glaciological applications > Analysis and results > 1. Data preparation and processing > 1.4 Data handling", "title": "Ice sheet velocity data from satellite observations: temporal and spatial coverage and data completeness for trend analysis in glaciological applications", "chunk_index": 8, "token_count": 335, "text_raw": "Let us calculate and add the total standard deviation to the data, as well as the flow direction of the ice, so we can work with it easily later on in the notebook. We also perform some data slight manipulations to the data array to improve data handling:\n\nCompute standard deviation\nFlow direction\nKeep only selected variables (coords like 'source' stay by default)\nApply transformation\nAdd attributes\n\nNow, our dataset array only holds the most important information: the total horizontal velocity magnitude (`land_ice_surface_velocity_magnitude`), the valid pixel count (`land_ice_surface_measurement_count`), the calculated ice flow direction (`land_ice_flow_direction`), and our calculated standard deviation (`land_ice_surface_stddev`) from the northing and easting components. Let us first make the plotting function before we begin with the analysis:\n\nCircular mean over time for the flow direction\nMain plotting function\nAuto-detect if this is directional data (degrees)\nTake mean over time (period)\nAssign cyclic colormap if direction\nCreate polar plot\nSet map extent and features\nColorbar with angular formatting if needed\n\n(section-2)=", "text_with_prefix": "EQC Quality Assessment: \"Ice sheet velocity data from satellite observations: temporal and spatial coverage and data completeness for trend analysis in glaciological applications\"\nDataset: satellite-greenland-ice-sheet-velocity [CDS]\nAspect: trend-assessment_q01 | Category: Satellite_ECVs\nSection: Ice sheet velocity data from satellite observations: temporal and spatial coverage and data completeness for trend analysis in glaciological applications > Analysis and results > 1. Data preparation and processing > 1.4 Data handling\n---\nLet us calculate and add the total standard deviation to the data, as well as the flow direction of the ice, so we can work with it easily later on in the notebook. We also perform some data slight manipulations to the data array to improve data handling:\n\nCompute standard deviation\nFlow direction\nKeep only selected variables (coords like 'source' stay by default)\nApply transformation\nAdd attributes\n\nNow, our dataset array only holds the most important information: the total horizontal velocity magnitude (`land_ice_surface_velocity_magnitude`), the valid pixel count (`land_ice_surface_measurement_count`), the calculated ice flow direction (`land_ice_flow_direction`), and our calculated standard deviation (`land_ice_surface_stddev`) from the northing and easting components. Let us first make the plotting function before we begin with the analysis:\n\nCircular mean over time for the flow direction\nMain plotting function\nAuto-detect if this is directional data (degrees)\nTake mean over time (period)\nAssign cyclic colormap if direction\nCreate polar plot\nSet map extent and features\nColorbar with angular formatting if needed\n\n(section-2)="} {"chunk_id": "satellite_satellite-greenland-ice-sheet-velocity_trend-assessment_q01__ecd8fb393d92", "report_id": "satellite_satellite-greenland-ice-sheet-velocity_trend-assessment_q01", "dataset_id": "satellite-greenland-ice-sheet-velocity", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q01", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Ice sheet velocity data from satellite observations: temporal and spatial coverage and data completeness for trend analysis in glaciological applications > Analysis and results > 2. Spatial and temporal patterns of ice sheet surface ice flow velocities > 2.1 Average horizontal ice flow velocities", "title": "Ice sheet velocity data from satellite observations: temporal and spatial coverage and data completeness for trend analysis in glaciological applications", "chunk_index": 9, "token_count": 469, "text_raw": "We begin by plotting the ice flow velocities averaged over every hydrological year between the beginning and end period that we selected for the velocity data with the defined plotting function.\n\n🚨 **The files can be large! Since the data files to be plotted have a considerable size, this may take a couple of minutes.**\n\n*Figure 1. Magnitude of the average horizontal surface flow velocities over Greenland over the defined time period.*\n\nThe corresponding data are annually averaged values, derived from all year-round observations during the glaciological balance year of the GrIS (1 October to 30 September). The horizontal velocity data effectively highlight the low-flow zones in the interior and the increased velocities near the ice sheet’s margins and outlet glaciers. This pattern aligns with findings from other GrIS velocity maps in the literature (e.g. [[2](https://doi.org/10.1017/jog.2017.73), [3](https://nsidc.org/data/nsidc-0777/versions/1)]). The annually averaged nature meets the minimum requirement set by the Global Climate Observing System (GCOS) [[4](https://library.wmo.int/records/item/58111-the-2022-gcos-ecvs-requirements-gcos-245)], providing a comprehensive overview of ice flow velocities. However, there is no information given related to the time of the year during which valid pixels for the velocity calculation were acquired. Also the spatial resolution of 250 m meets the threshold value proposed by GCOS. In other words, both the spatial and temporal resolution meet the minimum requirement to ensure that data are useful for use in climatological, hydrological and glaciological studies and applications.\n\n(section-2-2)=", "text_with_prefix": "EQC Quality Assessment: \"Ice sheet velocity data from satellite observations: temporal and spatial coverage and data completeness for trend analysis in glaciological applications\"\nDataset: satellite-greenland-ice-sheet-velocity [CDS]\nAspect: trend-assessment_q01 | Category: Satellite_ECVs\nSection: Ice sheet velocity data from satellite observations: temporal and spatial coverage and data completeness for trend analysis in glaciological applications > Analysis and results > 2. Spatial and temporal patterns of ice sheet surface ice flow velocities > 2.1 Average horizontal ice flow velocities\n---\nWe begin by plotting the ice flow velocities averaged over every hydrological year between the beginning and end period that we selected for the velocity data with the defined plotting function.\n\n🚨 **The files can be large! Since the data files to be plotted have a considerable size, this may take a couple of minutes.**\n\n*Figure 1. Magnitude of the average horizontal surface flow velocities over Greenland over the defined time period.*\n\nThe corresponding data are annually averaged values, derived from all year-round observations during the glaciological balance year of the GrIS (1 October to 30 September). The horizontal velocity data effectively highlight the low-flow zones in the interior and the increased velocities near the ice sheet’s margins and outlet glaciers. This pattern aligns with findings from other GrIS velocity maps in the literature (e.g. [[2](https://doi.org/10.1017/jog.2017.73), [3](https://nsidc.org/data/nsidc-0777/versions/1)]). The annually averaged nature meets the minimum requirement set by the Global Climate Observing System (GCOS) [[4](https://library.wmo.int/records/item/58111-the-2022-gcos-ecvs-requirements-gcos-245)], providing a comprehensive overview of ice flow velocities. However, there is no information given related to the time of the year during which valid pixels for the velocity calculation were acquired. Also the spatial resolution of 250 m meets the threshold value proposed by GCOS. In other words, both the spatial and temporal resolution meet the minimum requirement to ensure that data are useful for use in climatological, hydrological and glaciological studies and applications.\n\n(section-2-2)="} {"chunk_id": "satellite_satellite-greenland-ice-sheet-velocity_trend-assessment_q01__a221a2fad934", "report_id": "satellite_satellite-greenland-ice-sheet-velocity_trend-assessment_q01", "dataset_id": "satellite-greenland-ice-sheet-velocity", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q01", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Ice sheet velocity data from satellite observations: temporal and spatial coverage and data completeness for trend analysis in glaciological applications > Analysis and results > 2. Spatial and temporal patterns of ice sheet surface ice flow velocities > 2.2 Average horizontal ice flow direction", "title": "Ice sheet velocity data from satellite observations: temporal and spatial coverage and data completeness for trend analysis in glaciological applications", "chunk_index": 10, "token_count": 329, "text_raw": "We can also plot flow directions (i.e. the direction to which the ice flows measured clockwise from the north). For that, we calculate the circular mean:\n\nApply to flow direction\n\n*Figure 2. Average direction of the horizontal surface flow velocities over Greenland over the defined time period.*\n\nThe provided image displays the average ice flow direction for Greenland averaged over all available periods. The color scale indicates the direction of ice flow in degrees, measured clockwise from north (northwards = 0°/360°, eastwards = 90°, southwards = 180°, westwards = 270°). Several ice divides are visible, of which the most prominent one is situated meriodionally over central Greenland where flow directions diverge, leading to different drainage basins. The predominant direction of the ice flow is consistent with the topographical characteristics of the ice sheet (i.e. ice flows parallel to the direction of the steepest surface slope).\n\nLet us now analyze the spatial and temporal resolution and coverage of the dataset.\n\n(section-3)=", "text_with_prefix": "EQC Quality Assessment: \"Ice sheet velocity data from satellite observations: temporal and spatial coverage and data completeness for trend analysis in glaciological applications\"\nDataset: satellite-greenland-ice-sheet-velocity [CDS]\nAspect: trend-assessment_q01 | Category: Satellite_ECVs\nSection: Ice sheet velocity data from satellite observations: temporal and spatial coverage and data completeness for trend analysis in glaciological applications > Analysis and results > 2. Spatial and temporal patterns of ice sheet surface ice flow velocities > 2.2 Average horizontal ice flow direction\n---\nWe can also plot flow directions (i.e. the direction to which the ice flows measured clockwise from the north). For that, we calculate the circular mean:\n\nApply to flow direction\n\n*Figure 2. Average direction of the horizontal surface flow velocities over Greenland over the defined time period.*\n\nThe provided image displays the average ice flow direction for Greenland averaged over all available periods. The color scale indicates the direction of ice flow in degrees, measured clockwise from north (northwards = 0°/360°, eastwards = 90°, southwards = 180°, westwards = 270°). Several ice divides are visible, of which the most prominent one is situated meriodionally over central Greenland where flow directions diverge, leading to different drainage basins. The predominant direction of the ice flow is consistent with the topographical characteristics of the ice sheet (i.e. ice flows parallel to the direction of the steepest surface slope).\n\nLet us now analyze the spatial and temporal resolution and coverage of the dataset.\n\n(section-3)="} {"chunk_id": "satellite_satellite-greenland-ice-sheet-velocity_trend-assessment_q01__ada910324221", "report_id": "satellite_satellite-greenland-ice-sheet-velocity_trend-assessment_q01", "dataset_id": "satellite-greenland-ice-sheet-velocity", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q01", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Ice sheet velocity data from satellite observations: temporal and spatial coverage and data completeness for trend analysis in glaciological applications > Analysis and results > 3. Analysis of spatio-temporal resolution and extent of ice sheet velocities > 3.1 Temporal extent of the ice sheet velocity dataset", "title": "Ice sheet velocity data from satellite observations: temporal and spatial coverage and data completeness for trend analysis in glaciological applications", "chunk_index": 11, "token_count": 367, "text_raw": "We begin our analysis by examining the temporal extent of the dataset:\n\n```text\nThe temporal extent of the ice sheet velocity dataset for the GrIS is 9 years.\n```\n\nTo determine whether the temporal extent of the ice sheet velocity dataset is sufficient to capture reliable temporal trends in horizontal surface velocities and to use these trends as indicators of climatic changes, we turn to the literature. The Intergovernmental Panel on Climate Change (IPCC) often uses 30 years as a standard period for climate normals and trend analysis to ensure that the analysis captures meaningful climatic changes rather than short-term (intra/interannual) fluctuations. We therefore consider these guidelines to be likewise applicative for ice flow velocities.\n\nWhen measured over a long period (> 30 years), significant trends in surface flow velocities can be considered a clear indicator of a disequilibrium of the ice sheet with the environment, for example due to climate change. Longer periods provide even more robust trend estimates and reduce the influence of short-term variability. However, the dataset (as of versions 1.3 to 1.5 with 9 years of data) is clearly not of sufficient temporal extent for this purpose.\n\n(section-3-2)=", "text_with_prefix": "EQC Quality Assessment: \"Ice sheet velocity data from satellite observations: temporal and spatial coverage and data completeness for trend analysis in glaciological applications\"\nDataset: satellite-greenland-ice-sheet-velocity [CDS]\nAspect: trend-assessment_q01 | Category: Satellite_ECVs\nSection: Ice sheet velocity data from satellite observations: temporal and spatial coverage and data completeness for trend analysis in glaciological applications > Analysis and results > 3. Analysis of spatio-temporal resolution and extent of ice sheet velocities > 3.1 Temporal extent of the ice sheet velocity dataset\n---\nWe begin our analysis by examining the temporal extent of the dataset:\n\n```text\nThe temporal extent of the ice sheet velocity dataset for the GrIS is 9 years.\n```\n\nTo determine whether the temporal extent of the ice sheet velocity dataset is sufficient to capture reliable temporal trends in horizontal surface velocities and to use these trends as indicators of climatic changes, we turn to the literature. The Intergovernmental Panel on Climate Change (IPCC) often uses 30 years as a standard period for climate normals and trend analysis to ensure that the analysis captures meaningful climatic changes rather than short-term (intra/interannual) fluctuations. We therefore consider these guidelines to be likewise applicative for ice flow velocities.\n\nWhen measured over a long period (> 30 years), significant trends in surface flow velocities can be considered a clear indicator of a disequilibrium of the ice sheet with the environment, for example due to climate change. Longer periods provide even more robust trend estimates and reduce the influence of short-term variability. However, the dataset (as of versions 1.3 to 1.5 with 9 years of data) is clearly not of sufficient temporal extent for this purpose.\n\n(section-3-2)="} {"chunk_id": "satellite_satellite-greenland-ice-sheet-velocity_trend-assessment_q01__fa354f9c2385", "report_id": "satellite_satellite-greenland-ice-sheet-velocity_trend-assessment_q01", "dataset_id": "satellite-greenland-ice-sheet-velocity", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q01", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Ice sheet velocity data from satellite observations: temporal and spatial coverage and data completeness for trend analysis in glaciological applications > Analysis and results > 3. Analysis of spatio-temporal resolution and extent of ice sheet velocities > 3.2 Spatial coverage of the ice sheet velocity dataset", "title": "Ice sheet velocity data from satellite observations: temporal and spatial coverage and data completeness for trend analysis in glaciological applications", "chunk_index": 12, "token_count": 1027, "text_raw": "To have a more extended idea of the temporal and spatial coverage of the gridded dataset, we can plot the total amount of pixels that hold non-NaN velocity data for each year. Let us have this quantified:\n\nApply the function to the velocity counts\nPrint values\n\n```text\nThe total amount of pixels holding non-NaN values for the year 2014-2015 CE is 33048867 counts, which equals a surface area of 2.06555 million km².\nThe total amount of pixels holding non-NaN values for the year 2015-2016 CE is 34139948 counts, which equals a surface area of 2.13375 million km².\nThe total amount of pixels holding non-NaN values for the year 2016-2017 CE is 34195665 counts, which equals a surface area of 2.13723 million km².\nThe total amount of pixels holding non-NaN values for the year 2017-2018 CE is 34146643 counts, which equals a surface area of 2.13417 million km².\nThe total amount of pixels holding non-NaN values for the year 2018-2019 CE is 34166241 counts, which equals a surface area of 2.13539 million km².\nThe total amount of pixels holding non-NaN values for the year 2019-2020 CE is 34181540 counts, which equals a surface area of 2.13635 million km².\nThe total amount of pixels holding non-NaN values for the year 2020-2021 CE is 34210311 counts, which equals a surface area of 2.13814 million km².\nThe total amount of pixels holding non-NaN values for the year 2021-2022 CE is 34203801 counts, which equals a surface area of 2.13774 million km².\nThe total amount of pixels holding non-NaN values for the year 2022-2023 CE is 34070897 counts, which equals a surface area of 2.12943 million km².\n```\n\nGiven that the surface area of the entire continent of Greenland roughly equals 2.165 million km² (of which ca. 80% covered by ice), it is clear that almost the entire continent is covered by the product, including ice-free areas. Let us determine what percentage of the pixels hold a time series of data with a length that is equal to the magnitude of the `time` dimension of the dataset:\n\n```text\nThe number of pixels that have a time series of valid velocity data of 9 years, which is the total number of years in the dataset, is 32911734 pixels or 96.18%.\n```\n\nThe amount of missing data can therefore be considered limited.\n\nLet us now check where the missing pixels are. In the plot below, pixels with a time series of valid data that are lower than the time dimension of the data are colored red, else they are colored green:\n\nCreate subplots with Polar Stereographic projection\nPlot the data\nSet extent and plot features\nAdd colorbar\nApply the function to the velocity data\nDefine bounds and colormap\n\n*Figure 3. Presence of missing data of the horizontal surface flow velocities over Greenland over the defined time period.*\n\nThe dataset thus offers GrIS velocity data at regular and consistently spaced temporal intervals (e.g. annually averaged data), and there are (almost) no spatial/temporal gaps in the data that could affect a (local or ice sheet-wide) trend analysis. However, the total number of consecutive years in the dataset is too low to perform a reliable temporal trend analysis. In other words, from the above analysis it becomes clear that the dataset exhibits a consistently complete temporal and spatial coverage, with practically no data gaps, which would allow for reliable quantifications of velocity means, variability and trends if the temporal extent would have been sufficient. Moreover, an ice mask is not included and it is thus not possible to exclude non ice-covered pixels from the data. Pixels over ice-free terrain hold non-NaN data and have not been removed, implying that the data exhibits false low velocity measurements and noise over ice-free terrain. These non-ice values are, however, intentionally retained because the velocity retrievals over stable land are used for the quality assessment of the algorithm, as detailed in the PQAR.\n\n(section-4)=", "text_with_prefix": "EQC Quality Assessment: \"Ice sheet velocity data from satellite observations: temporal and spatial coverage and data completeness for trend analysis in glaciological applications\"\nDataset: satellite-greenland-ice-sheet-velocity [CDS]\nAspect: trend-assessment_q01 | Category: Satellite_ECVs\nSection: Ice sheet velocity data from satellite observations: temporal and spatial coverage and data completeness for trend analysis in glaciological applications > Analysis and results > 3. Analysis of spatio-temporal resolution and extent of ice sheet velocities > 3.2 Spatial coverage of the ice sheet velocity dataset\n---\nTo have a more extended idea of the temporal and spatial coverage of the gridded dataset, we can plot the total amount of pixels that hold non-NaN velocity data for each year. Let us have this quantified:\n\nApply the function to the velocity counts\nPrint values\n\n```text\nThe total amount of pixels holding non-NaN values for the year 2014-2015 CE is 33048867 counts, which equals a surface area of 2.06555 million km².\nThe total amount of pixels holding non-NaN values for the year 2015-2016 CE is 34139948 counts, which equals a surface area of 2.13375 million km².\nThe total amount of pixels holding non-NaN values for the year 2016-2017 CE is 34195665 counts, which equals a surface area of 2.13723 million km².\nThe total amount of pixels holding non-NaN values for the year 2017-2018 CE is 34146643 counts, which equals a surface area of 2.13417 million km².\nThe total amount of pixels holding non-NaN values for the year 2018-2019 CE is 34166241 counts, which equals a surface area of 2.13539 million km².\nThe total amount of pixels holding non-NaN values for the year 2019-2020 CE is 34181540 counts, which equals a surface area of 2.13635 million km².\nThe total amount of pixels holding non-NaN values for the year 2020-2021 CE is 34210311 counts, which equals a surface area of 2.13814 million km².\nThe total amount of pixels holding non-NaN values for the year 2021-2022 CE is 34203801 counts, which equals a surface area of 2.13774 million km².\nThe total amount of pixels holding non-NaN values for the year 2022-2023 CE is 34070897 counts, which equals a surface area of 2.12943 million km².\n```\n\nGiven that the surface area of the entire continent of Greenland roughly equals 2.165 million km² (of which ca. 80% covered by ice), it is clear that almost the entire continent is covered by the product, including ice-free areas. Let us determine what percentage of the pixels hold a time series of data with a length that is equal to the magnitude of the `time` dimension of the dataset:\n\n```text\nThe number of pixels that have a time series of valid velocity data of 9 years, which is the total number of years in the dataset, is 32911734 pixels or 96.18%.\n```\n\nThe amount of missing data can therefore be considered limited.\n\nLet us now check where the missing pixels are. In the plot below, pixels with a time series of valid data that are lower than the time dimension of the data are colored red, else they are colored green:\n\nCreate subplots with Polar Stereographic projection\nPlot the data\nSet extent and plot features\nAdd colorbar\nApply the function to the velocity data\nDefine bounds and colormap\n\n*Figure 3. Presence of missing data of the horizontal surface flow velocities over Greenland over the defined time period.*\n\nThe dataset thus offers GrIS velocity data at regular and consistently spaced temporal intervals (e.g. annually averaged data), and there are (almost) no spatial/temporal gaps in the data that could affect a (local or ice sheet-wide) trend analysis. However, the total number of consecutive years in the dataset is too low to perform a reliable temporal trend analysis. In other words, from the above analysis it becomes clear that the dataset exhibits a consistently complete temporal and spatial coverage, with practically no data gaps, which would allow for reliable quantifications of velocity means, variability and trends if the temporal extent would have been sufficient. Moreover, an ice mask is not included and it is thus not possible to exclude non ice-covered pixels from the data. Pixels over ice-free terrain hold non-NaN data and have not been removed, implying that the data exhibits false low velocity measurements and noise over ice-free terrain. These non-ice values are, however, intentionally retained because the velocity retrievals over stable land are used for the quality assessment of the algorithm, as detailed in the PQAR.\n\n(section-4)="} {"chunk_id": "satellite_satellite-greenland-ice-sheet-velocity_trend-assessment_q01__3f09c6836865", "report_id": "satellite_satellite-greenland-ice-sheet-velocity_trend-assessment_q01", "dataset_id": "satellite-greenland-ice-sheet-velocity", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q01", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Ice sheet velocity data from satellite observations: temporal and spatial coverage and data completeness for trend analysis in glaciological applications > Analysis and results > 4. Trends of annually averaged horizontal surface ice flow velocities > 4.1 Linear trends of horizontal ice flow velocities", "title": "Ice sheet velocity data from satellite observations: temporal and spatial coverage and data completeness for trend analysis in glaciological applications", "chunk_index": 13, "token_count": 1108, "text_raw": "Although the temporal extent is not sufficient to derive reliable values for the mean, variability and trends (as discussed above), we can still plot a time series of annually averaged ice sheet-wide velocity data to get an idea of their magnitudes and their change over time in the dataset. Let us print the average values over the entire ice sheet for each hydrological year:\n\nPrint values\n\n```text\nThe annually averaged ice sheet-wide horizontal surface ice flow velocity for the year 2014-2015 CE is 0.1124 m/day.\nThe annually averaged ice sheet-wide horizontal surface ice flow velocity for the year 2015-2016 CE is 0.1113 m/day.\nThe annually averaged ice sheet-wide horizontal surface ice flow velocity for the year 2016-2017 CE is 0.1134 m/day.\nThe annually averaged ice sheet-wide horizontal surface ice flow velocity for the year 2017-2018 CE is 0.1172 m/day.\nThe annually averaged ice sheet-wide horizontal surface ice flow velocity for the year 2018-2019 CE is 0.1164 m/day.\nThe annually averaged ice sheet-wide horizontal surface ice flow velocity for the year 2019-2020 CE is 0.1154 m/day.\nThe annually averaged ice sheet-wide horizontal surface ice flow velocity for the year 2020-2021 CE is 0.1131 m/day.\nThe annually averaged ice sheet-wide horizontal surface ice flow velocity for the year 2021-2022 CE is 0.1132 m/day.\nThe annually averaged ice sheet-wide horizontal surface ice flow velocity for the year 2022-2023 CE is 0.1156 m/day.\n```\n\nLet us now have this plotted on a graph:\n\nExtract time and velocity data\nConvert period strings to numerical values (midpoint of each period)\nCalculate linear trend\nPlot the data\n\n*Figure 4. Linear trend of the ice sheet-wide average horizontal surface flow velocities over Greenland over the defined time period.*\n\nWe can quantify the linear trend as follows:\n\nQuantify the linear trend (coeffs[0] is the slope, coeffs[1] is the intercept)\n\n```text\nThe linear trend of the ice sheet-wide horizontal surface ice flow velocity between 2014-2015 and 2022-2023 is 0.0003 m day⁻¹ yr⁻¹.\n```\n\nThe provided plot shows the trend of ice sheet-wide horizontal surface ice flow velocities since the period 2014-2015. The linear trend shows an increase in the ice sheet-wide horizontal surface ice flow velocity over time. This means that the ice flow has generally been speeding up over this period. This generally agrees with what can be expected from a warming climate, although the impact of climate change on the GrIS dynamic patterns is complex in both space and time. The outlet glaciers of the GrIS are currently retreating, calving more icebergs, and also flowing faster. Many major outlet glaciers in Greenland (e.g. Jakobshavn Isbræ, Helheim, and Kangerlussuaq) experienced substantial speed-ups during the last several decades. This acceleration was mainly linked to oceanic and atmospheric warming, which led to increased basal sliding and reduced backstress (buttressing), hereby enhancing flow speed (e.g. [[5](https://doi.org/10.5194/tc-6-923-2012), [6](https://doi.org/10.1038/s41586-022-05301-z), [7](https://doi.org/10.1038/s43247-020-0001-2)]). However, studies indicate that this ongoing acceleration cannot simply be extrapolated throughout the 21st century and beyond. In the future, the ice sheet's thinning is expected to slow down the ice due to the inverse relationship between ice thickness and ice velocity, as expected from physical theory (e.g. [[8](https://doi.org/10.5194/tc-9-1039-2015)]).\n\nHowever, it must be said that the temporal extent of the velocity data is relatively short. Longer-term observations (> 30 years) are necessary to identify persistent trends and separate short-term variability from long-term changes. Moreover, the trend in ice sheet-wide velocity data for Greenland may be influenced by the inclusion of unrealistic velocity data over ice-free terrain, whereas certain zones, such as the glacier margins, may be subject to lower data quality as well [[9](https://doi.org/10.1016/j.rse.2017.08.038)]. Therefore, further investigation and long-term monitoring of velocities over ice-covered areas are essential to accurately and reliably assess and interpret trends of ice flow velocities and their implications for future ice sheet behavior and sea-level rise.\n\n(section-4-2)=", "text_with_prefix": "EQC Quality Assessment: \"Ice sheet velocity data from satellite observations: temporal and spatial coverage and data completeness for trend analysis in glaciological applications\"\nDataset: satellite-greenland-ice-sheet-velocity [CDS]\nAspect: trend-assessment_q01 | Category: Satellite_ECVs\nSection: Ice sheet velocity data from satellite observations: temporal and spatial coverage and data completeness for trend analysis in glaciological applications > Analysis and results > 4. Trends of annually averaged horizontal surface ice flow velocities > 4.1 Linear trends of horizontal ice flow velocities\n---\nAlthough the temporal extent is not sufficient to derive reliable values for the mean, variability and trends (as discussed above), we can still plot a time series of annually averaged ice sheet-wide velocity data to get an idea of their magnitudes and their change over time in the dataset. Let us print the average values over the entire ice sheet for each hydrological year:\n\nPrint values\n\n```text\nThe annually averaged ice sheet-wide horizontal surface ice flow velocity for the year 2014-2015 CE is 0.1124 m/day.\nThe annually averaged ice sheet-wide horizontal surface ice flow velocity for the year 2015-2016 CE is 0.1113 m/day.\nThe annually averaged ice sheet-wide horizontal surface ice flow velocity for the year 2016-2017 CE is 0.1134 m/day.\nThe annually averaged ice sheet-wide horizontal surface ice flow velocity for the year 2017-2018 CE is 0.1172 m/day.\nThe annually averaged ice sheet-wide horizontal surface ice flow velocity for the year 2018-2019 CE is 0.1164 m/day.\nThe annually averaged ice sheet-wide horizontal surface ice flow velocity for the year 2019-2020 CE is 0.1154 m/day.\nThe annually averaged ice sheet-wide horizontal surface ice flow velocity for the year 2020-2021 CE is 0.1131 m/day.\nThe annually averaged ice sheet-wide horizontal surface ice flow velocity for the year 2021-2022 CE is 0.1132 m/day.\nThe annually averaged ice sheet-wide horizontal surface ice flow velocity for the year 2022-2023 CE is 0.1156 m/day.\n```\n\nLet us now have this plotted on a graph:\n\nExtract time and velocity data\nConvert period strings to numerical values (midpoint of each period)\nCalculate linear trend\nPlot the data\n\n*Figure 4. Linear trend of the ice sheet-wide average horizontal surface flow velocities over Greenland over the defined time period.*\n\nWe can quantify the linear trend as follows:\n\nQuantify the linear trend (coeffs[0] is the slope, coeffs[1] is the intercept)\n\n```text\nThe linear trend of the ice sheet-wide horizontal surface ice flow velocity between 2014-2015 and 2022-2023 is 0.0003 m day⁻¹ yr⁻¹.\n```\n\nThe provided plot shows the trend of ice sheet-wide horizontal surface ice flow velocities since the period 2014-2015. The linear trend shows an increase in the ice sheet-wide horizontal surface ice flow velocity over time. This means that the ice flow has generally been speeding up over this period. This generally agrees with what can be expected from a warming climate, although the impact of climate change on the GrIS dynamic patterns is complex in both space and time. The outlet glaciers of the GrIS are currently retreating, calving more icebergs, and also flowing faster. Many major outlet glaciers in Greenland (e.g. Jakobshavn Isbræ, Helheim, and Kangerlussuaq) experienced substantial speed-ups during the last several decades. This acceleration was mainly linked to oceanic and atmospheric warming, which led to increased basal sliding and reduced backstress (buttressing), hereby enhancing flow speed (e.g. [[5](https://doi.org/10.5194/tc-6-923-2012), [6](https://doi.org/10.1038/s41586-022-05301-z), [7](https://doi.org/10.1038/s43247-020-0001-2)]). However, studies indicate that this ongoing acceleration cannot simply be extrapolated throughout the 21st century and beyond. In the future, the ice sheet's thinning is expected to slow down the ice due to the inverse relationship between ice thickness and ice velocity, as expected from physical theory (e.g. [[8](https://doi.org/10.5194/tc-9-1039-2015)]).\n\nHowever, it must be said that the temporal extent of the velocity data is relatively short. Longer-term observations (> 30 years) are necessary to identify persistent trends and separate short-term variability from long-term changes. Moreover, the trend in ice sheet-wide velocity data for Greenland may be influenced by the inclusion of unrealistic velocity data over ice-free terrain, whereas certain zones, such as the glacier margins, may be subject to lower data quality as well [[9](https://doi.org/10.1016/j.rse.2017.08.038)]. Therefore, further investigation and long-term monitoring of velocities over ice-covered areas are essential to accurately and reliably assess and interpret trends of ice flow velocities and their implications for future ice sheet behavior and sea-level rise.\n\n(section-4-2)="} {"chunk_id": "satellite_satellite-greenland-ice-sheet-velocity_trend-assessment_q01__1a945edb6c2e", "report_id": "satellite_satellite-greenland-ice-sheet-velocity_trend-assessment_q01", "dataset_id": "satellite-greenland-ice-sheet-velocity", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q01", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Ice sheet velocity data from satellite observations: temporal and spatial coverage and data completeness for trend analysis in glaciological applications > Analysis and results > 4. Trends of annually averaged horizontal surface ice flow velocities > 4.2 Implications for deriving multi-year trends of surface flow velocities and associated patterns of solid ice discharge", "title": "Ice sheet velocity data from satellite observations: temporal and spatial coverage and data completeness for trend analysis in glaciological applications", "chunk_index": 14, "token_count": 1071, "text_raw": "Reliable and long-term trends in ice sheet flow velocities are essential for understanding the complex dynamics of ice sheets and the impact of climate change. One of such processes for which ice flow velocities are important is the quantification of solid ice discharge.\n\nSolid ice discharge is the transport of solid ice across the grounding line of an ice sheet. This process represents the loss of grounded ice from the ice sheet to the ocean due to ice flow. The grounding line is the boundary between the grounded ice sheet, where ice rests on bedrock, and the floating ice, which is buoyant. As ice flows from the grounded portion of the ice sheet into the ocean, it contributes to sea-level rise. Solid ice discharge $D$ therefore is a critical component of the total mass balance of an ice sheet [[10](https://essd.copernicus.org/articles/12/1367/2020/)]:\n\n$\\Delta M = SMB - D$\n\nwhere:\n- $\\Delta M$ is the total mass balance or the total ice mass loss or gain [Gt yr$^{-1}$],\n- $SMB$ is surface mass balance (often also supplemented by the basal and internal mass balance) [Gt yr$^{-1}$],\n- $D$ is the solid ice discharge (zero for land-terminating glaciers/ice sheets) [Gt yr$^{-1}$].\n\nThe solid discharge $D$ can be further expressed as a mass flux as follows:\n\n$D = \\rho_i \\cdot \\overline{V_P} \\cdot A = \\rho_i \\cdot \\overline{V_P} \\cdot H \\cdot w \n$\n\nwhere\n- $\\rho_i$ is the ice density [kg m$^{-3}$]\n- $\\overline{V_P}$ the gate-perpendicular vertically averaged horizontal ice flow velocity at the grounding line [m yr$^{-1}$]\n\n≈ $\\overline{V}\\cdot cos(\\theta)$ with $\\theta$ the difference between ice flow and gate-perpendicular azimuths and where $\\overline{V}$ approximately equals the surface velocity $V_s$\n- $H$ the ice thickness at the grounding line [m]\n- $A$ the considered cross-sectional area [m]\n- $w$ the width of the cross section across the groundling line [m].\n\nIce flow velocities are thus an important component to assess the solid ice discharge and its trends over time. The dataset provided for monitoring GriS surface flow velocity patterns on the CDS is a relatively mature dataset that can serve as input for solid ice discharge calculations, as it has a suitable spatial resolution for capturing detailed flow patterns, particularly near ice margins and outlet glaciers. The spatial coverage is furthermore comprehensive, encompassing the entire ice sheet, including peripheral glaciers and ice caps, with almost no gaps present. The dataset's completeness thus appears high, with minimal missing values.\n\nHowever, the temporal coverage only extends from 2014-2015 onwards, which is too short to perform a reliable long-term trend analysis. The dataset's short temporal extent therefore poses a limitation to deduce reliable long-term trends of horizontal surface flow velocities and solid ice discharge. Furthermore, to precisely quantify the solid ice discharge, also other additional and external datasets of other variables are needed, of which the most important ones are the position of the grounding line and the ice thickness at those locations [[10](https://doi.org/10.5194/essd-12-1367-2020)]. Moreover, the dataset provides horizontal surface velocities, while solid ice discharge calculations require vertically averaged horizontal velocities. However, because basal sliding is generally the dominant process that defines the ice velocity at the grounding line, velocities generally do not vary significantly with depth at these locations [[10](https://doi.org/10.5194/essd-12-1367-2020)]. Next, with only yearly maps, this product can only calculate interannual changes in ice discharge, which may not be sufficient for certain users. This may be particularly true for the GrIS, where many outlet glaciers exhibit significantly distinct seasonal variations in ice velocity. The lack of information related to the time of the year of the valid data acquisitions, as well as higher standard deviations near the margins and a missing ice mask that can be used to distinguish between ice-covered and non ice-covered grid cells, further complicate the use of the data. Hence, for comprehensive glaciological studies and solid ice discharge modeling, supplementary datasets with a longer temporal extent (preferably > 30 years) and higher temporal resolution (subannual) should be considered to fully capture the changing patterns of ice flow dynamics.\n\n(section-5)=", "text_with_prefix": "EQC Quality Assessment: \"Ice sheet velocity data from satellite observations: temporal and spatial coverage and data completeness for trend analysis in glaciological applications\"\nDataset: satellite-greenland-ice-sheet-velocity [CDS]\nAspect: trend-assessment_q01 | Category: Satellite_ECVs\nSection: Ice sheet velocity data from satellite observations: temporal and spatial coverage and data completeness for trend analysis in glaciological applications > Analysis and results > 4. Trends of annually averaged horizontal surface ice flow velocities > 4.2 Implications for deriving multi-year trends of surface flow velocities and associated patterns of solid ice discharge\n---\nReliable and long-term trends in ice sheet flow velocities are essential for understanding the complex dynamics of ice sheets and the impact of climate change. One of such processes for which ice flow velocities are important is the quantification of solid ice discharge.\n\nSolid ice discharge is the transport of solid ice across the grounding line of an ice sheet. This process represents the loss of grounded ice from the ice sheet to the ocean due to ice flow. The grounding line is the boundary between the grounded ice sheet, where ice rests on bedrock, and the floating ice, which is buoyant. As ice flows from the grounded portion of the ice sheet into the ocean, it contributes to sea-level rise. Solid ice discharge $D$ therefore is a critical component of the total mass balance of an ice sheet [[10](https://essd.copernicus.org/articles/12/1367/2020/)]:\n\n$\\Delta M = SMB - D$\n\nwhere:\n- $\\Delta M$ is the total mass balance or the total ice mass loss or gain [Gt yr$^{-1}$],\n- $SMB$ is surface mass balance (often also supplemented by the basal and internal mass balance) [Gt yr$^{-1}$],\n- $D$ is the solid ice discharge (zero for land-terminating glaciers/ice sheets) [Gt yr$^{-1}$].\n\nThe solid discharge $D$ can be further expressed as a mass flux as follows:\n\n$D = \\rho_i \\cdot \\overline{V_P} \\cdot A = \\rho_i \\cdot \\overline{V_P} \\cdot H \\cdot w \n$\n\nwhere\n- $\\rho_i$ is the ice density [kg m$^{-3}$]\n- $\\overline{V_P}$ the gate-perpendicular vertically averaged horizontal ice flow velocity at the grounding line [m yr$^{-1}$]\n\n≈ $\\overline{V}\\cdot cos(\\theta)$ with $\\theta$ the difference between ice flow and gate-perpendicular azimuths and where $\\overline{V}$ approximately equals the surface velocity $V_s$\n- $H$ the ice thickness at the grounding line [m]\n- $A$ the considered cross-sectional area [m]\n- $w$ the width of the cross section across the groundling line [m].\n\nIce flow velocities are thus an important component to assess the solid ice discharge and its trends over time. The dataset provided for monitoring GriS surface flow velocity patterns on the CDS is a relatively mature dataset that can serve as input for solid ice discharge calculations, as it has a suitable spatial resolution for capturing detailed flow patterns, particularly near ice margins and outlet glaciers. The spatial coverage is furthermore comprehensive, encompassing the entire ice sheet, including peripheral glaciers and ice caps, with almost no gaps present. The dataset's completeness thus appears high, with minimal missing values.\n\nHowever, the temporal coverage only extends from 2014-2015 onwards, which is too short to perform a reliable long-term trend analysis. The dataset's short temporal extent therefore poses a limitation to deduce reliable long-term trends of horizontal surface flow velocities and solid ice discharge. Furthermore, to precisely quantify the solid ice discharge, also other additional and external datasets of other variables are needed, of which the most important ones are the position of the grounding line and the ice thickness at those locations [[10](https://doi.org/10.5194/essd-12-1367-2020)]. Moreover, the dataset provides horizontal surface velocities, while solid ice discharge calculations require vertically averaged horizontal velocities. However, because basal sliding is generally the dominant process that defines the ice velocity at the grounding line, velocities generally do not vary significantly with depth at these locations [[10](https://doi.org/10.5194/essd-12-1367-2020)]. Next, with only yearly maps, this product can only calculate interannual changes in ice discharge, which may not be sufficient for certain users. This may be particularly true for the GrIS, where many outlet glaciers exhibit significantly distinct seasonal variations in ice velocity. The lack of information related to the time of the year of the valid data acquisitions, as well as higher standard deviations near the margins and a missing ice mask that can be used to distinguish between ice-covered and non ice-covered grid cells, further complicate the use of the data. Hence, for comprehensive glaciological studies and solid ice discharge modeling, supplementary datasets with a longer temporal extent (preferably > 30 years) and higher temporal resolution (subannual) should be considered to fully capture the changing patterns of ice flow dynamics.\n\n(section-5)="} {"chunk_id": "satellite_satellite-greenland-ice-sheet-velocity_trend-assessment_q01__e67367e26d1d", "report_id": "satellite_satellite-greenland-ice-sheet-velocity_trend-assessment_q01", "dataset_id": "satellite-greenland-ice-sheet-velocity", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q01", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Ice sheet velocity data from satellite observations: temporal and spatial coverage and data completeness for trend analysis in glaciological applications > Analysis and results > 5. Short summary and take-home messages", "title": "Ice sheet velocity data from satellite observations: temporal and spatial coverage and data completeness for trend analysis in glaciological applications", "chunk_index": 15, "token_count": 767, "text_raw": "When measured over long periods and extensive geographical scales, trends in GrIS ice sheet flow velocities and solid ice discharge are clear indicators of global climate change [[9](https://doi.org/10.1016/j.rse.2017.08.038), [10](https://doi.org/10.5194/essd-12-1367-2020)]. To be able to derive such trends of ice sheet velocities and for climate change analysis/monitoring to become reliable and possible, the velocity dataset should at least exhibit a comprehensive spatial coverage (i.e. ice sheet-wide), a long and continuous temporal coverage (> 30 years), quantified and transparent pixel-by-pixel uncertainty estimates that meet international proposed thresholds [[4](https://library.wmo.int/records/item/58111-the-2022-gcos-ecvs-requirements-gcos-245)], a validation effort or a comparison to theoretical models, and an adequate spatio-temporal resolution (cfr. the \"Maturity Matrix\" [[11](https://doi.org/10.1175/BAMS-D-21-0109.1)]).\n\nThe [ice sheet velocity dataset on the CDS](https://cds.climate.copernicus.eu/datasets/satellite-greenland-ice-sheet-velocity?tab=overview) is found to exhibit a consistent suitable spatial (250 m) and temporal (annual) resolution, but a less extensive temporal coverage (< 30 years), to conduct a meaningful analysis of linear (and quadratic) trends in ice sheet surface velocities and solid ice discharge. Nevertheless, the spatial and temporal resolution/extent of the data align with international standards such as those proposed by GCOS [[4](https://library.wmo.int/records/item/58111-the-2022-gcos-ecvs-requirements-gcos-245)] and data gaps are practically non-existant in this dataset. An even finer temporal resolution would, however, be beneficial to better capture subannual variations in flow speeds and solid ice discharge estimates. Validation efforts also show encouraging results, as shown in the [PQAR](https://confluence.ecmwf.int/pages/viewpage.action?pageId=355348656). The resulting trends from the C3S dataset show a general speedup of the GrIS over the last few years.\n\nTaking all of the above into account, the SAR offset tracking-based GrIS velocity dataset is generally considered to be relatively well-suited for use in the context of visualizing the spatial patterns of ice flow and for solid ice discharge quantifications (if it is assumed that vertically averaged horizontal velocities approximate the surface velocities near the grounding line) in terms of its spatial/temporal resolution, spatial coverage and data completeness. However, the GrIS velocity data are at this stage found to be less suitable to derive the corresponding long-term trends (climate change signals) and inter/intra-annual variability. Moreover, noise over ice-free terrain and the lack of imformation regarding the time of valid data acquisitions may impact the results of the analysis. Users should thus acknowledge the above-mentioned limitations for this dataset. The derivation of (long-term) surface ice flow velocity trends from C3S data, as well as the associated patterns of solid ice discharge, should therefore be handled with care.", "text_with_prefix": "EQC Quality Assessment: \"Ice sheet velocity data from satellite observations: temporal and spatial coverage and data completeness for trend analysis in glaciological applications\"\nDataset: satellite-greenland-ice-sheet-velocity [CDS]\nAspect: trend-assessment_q01 | Category: Satellite_ECVs\nSection: Ice sheet velocity data from satellite observations: temporal and spatial coverage and data completeness for trend analysis in glaciological applications > Analysis and results > 5. Short summary and take-home messages\n---\nWhen measured over long periods and extensive geographical scales, trends in GrIS ice sheet flow velocities and solid ice discharge are clear indicators of global climate change [[9](https://doi.org/10.1016/j.rse.2017.08.038), [10](https://doi.org/10.5194/essd-12-1367-2020)]. To be able to derive such trends of ice sheet velocities and for climate change analysis/monitoring to become reliable and possible, the velocity dataset should at least exhibit a comprehensive spatial coverage (i.e. ice sheet-wide), a long and continuous temporal coverage (> 30 years), quantified and transparent pixel-by-pixel uncertainty estimates that meet international proposed thresholds [[4](https://library.wmo.int/records/item/58111-the-2022-gcos-ecvs-requirements-gcos-245)], a validation effort or a comparison to theoretical models, and an adequate spatio-temporal resolution (cfr. the \"Maturity Matrix\" [[11](https://doi.org/10.1175/BAMS-D-21-0109.1)]).\n\nThe [ice sheet velocity dataset on the CDS](https://cds.climate.copernicus.eu/datasets/satellite-greenland-ice-sheet-velocity?tab=overview) is found to exhibit a consistent suitable spatial (250 m) and temporal (annual) resolution, but a less extensive temporal coverage (< 30 years), to conduct a meaningful analysis of linear (and quadratic) trends in ice sheet surface velocities and solid ice discharge. Nevertheless, the spatial and temporal resolution/extent of the data align with international standards such as those proposed by GCOS [[4](https://library.wmo.int/records/item/58111-the-2022-gcos-ecvs-requirements-gcos-245)] and data gaps are practically non-existant in this dataset. An even finer temporal resolution would, however, be beneficial to better capture subannual variations in flow speeds and solid ice discharge estimates. Validation efforts also show encouraging results, as shown in the [PQAR](https://confluence.ecmwf.int/pages/viewpage.action?pageId=355348656). The resulting trends from the C3S dataset show a general speedup of the GrIS over the last few years.\n\nTaking all of the above into account, the SAR offset tracking-based GrIS velocity dataset is generally considered to be relatively well-suited for use in the context of visualizing the spatial patterns of ice flow and for solid ice discharge quantifications (if it is assumed that vertically averaged horizontal velocities approximate the surface velocities near the grounding line) in terms of its spatial/temporal resolution, spatial coverage and data completeness. However, the GrIS velocity data are at this stage found to be less suitable to derive the corresponding long-term trends (climate change signals) and inter/intra-annual variability. Moreover, noise over ice-free terrain and the lack of imformation regarding the time of valid data acquisitions may impact the results of the analysis. Users should thus acknowledge the above-mentioned limitations for this dataset. The derivation of (long-term) surface ice flow velocity trends from C3S data, as well as the associated patterns of solid ice discharge, should therefore be handled with care."} {"chunk_id": "satellite_satellite-greenland-ice-sheet-velocity_trend-assessment_q01__ed5c95ffd872", "report_id": "satellite_satellite-greenland-ice-sheet-velocity_trend-assessment_q01", "dataset_id": "satellite-greenland-ice-sheet-velocity", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q01", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Ice sheet velocity data from satellite observations: temporal and spatial coverage and data completeness for trend analysis in glaciological applications > ℹ️ If you want to know more > Key resources", "title": "Ice sheet velocity data from satellite observations: temporal and spatial coverage and data completeness for trend analysis in glaciological applications", "chunk_index": 16, "token_count": 310, "text_raw": "- \"[Ice sheet velocity for Antarctica and Greenland derived from satellite observations](https://cds.climate.copernicus.eu/datasets/satellite-greenland-ice-sheet-velocity?tab=overview)\" on the CDS\n- [Documentation on the CDS](https://cds.climate.copernicus.eu/datasets/satellite-greenland-ice-sheet-velocity?tab=documentation) and the [ECMWF Confluence Wiki](https://confluence.ecmwf.int/pages/viewpage.action?pageId=355348656) (Copernicus Knowledge Base).\n- [Copernicus climate change indicators: ice sheets](https://climate.copernicus.eu/climate-indicators/ice-sheets)\n- [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control` prepared by [B-Open](https://www.bopen.eu).", "text_with_prefix": "EQC Quality Assessment: \"Ice sheet velocity data from satellite observations: temporal and spatial coverage and data completeness for trend analysis in glaciological applications\"\nDataset: satellite-greenland-ice-sheet-velocity [CDS]\nAspect: trend-assessment_q01 | Category: Satellite_ECVs\nSection: Ice sheet velocity data from satellite observations: temporal and spatial coverage and data completeness for trend analysis in glaciological applications > ℹ️ If you want to know more > Key resources\n---\n- \"[Ice sheet velocity for Antarctica and Greenland derived from satellite observations](https://cds.climate.copernicus.eu/datasets/satellite-greenland-ice-sheet-velocity?tab=overview)\" on the CDS\n- [Documentation on the CDS](https://cds.climate.copernicus.eu/datasets/satellite-greenland-ice-sheet-velocity?tab=documentation) and the [ECMWF Confluence Wiki](https://confluence.ecmwf.int/pages/viewpage.action?pageId=355348656) (Copernicus Knowledge Base).\n- [Copernicus climate change indicators: ice sheets](https://climate.copernicus.eu/climate-indicators/ice-sheets)\n- [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control` prepared by [B-Open](https://www.bopen.eu)."} {"chunk_id": "satellite_satellite-greenland-ice-sheet-velocity_trend-assessment_q01__5cee18be9e57", "report_id": "satellite_satellite-greenland-ice-sheet-velocity_trend-assessment_q01", "dataset_id": "satellite-greenland-ice-sheet-velocity", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q01", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Ice sheet velocity data from satellite observations: temporal and spatial coverage and data completeness for trend analysis in glaciological applications > ℹ️ If you want to know more > References", "title": "Ice sheet velocity data from satellite observations: temporal and spatial coverage and data completeness for trend analysis in glaciological applications", "chunk_index": 17, "token_count": 1074, "text_raw": "- [[1](https://doi.org/10.3390/rs70709371)] Nagler, T., Rott, H., Hetzenecker, M., Wuite, J. and Potin, P. (2015). The Sentinel-1 Mission: New Opportunities for Ice Sheet Observations. Remote Sensing. 7(7):9371-9389. https://doi.org/10.3390/rs70709371\n\n- [[2](https://doi.org/10.1017/jog.2017.73)] Joughin, I., Smith, B. E., and Howat, I. M. (2018). A complete map of Greenland ice velocity derived from satellite data collected over 20 years, J. Glaciol., 64, 1–11, https://doi.org/10.1017/jog.2017.73\n\n- [[3](https://nsidc.org/data/nsidc-0777/versions/1)] Howat, I., Chudley, T., and Noh, M. (2022). MEaSUREs Greenland Ice Velocity: Selected Glacier Site Single-Pair Velocity Maps from Optical Images, https://doi.org/10.5067/b28fm2qvvywy\n\n- [[4](https://library.wmo.int/records/item/58111-the-2022-gcos-ecvs-requirements-gcos-245)] GCOS (Global Climate Observing System) (2022). The 2022 GCOS ECVs Requirements (GCOS-245). World Meteorological Organization: Geneva, Switzerland. doi: https://library.wmo.int/idurl/4/58111\n\n- [[5](https://doi.org/10.5194/tc-6-923-2012)] Bevan, S. L., Luckman, A. J., and Murray, T. (2012). Glacier dynamics over the last quarter of a century at Helheim, Kangerdlugssuaq and 14 other major Greenland outlet glaciers, The Cryosphere, 6, 923–937, https://doi.org/10.5194/tc-6-923-2012\n\n- [[6](https://doi.org/10.1038/s41586-022-05301-z)] Khan, S. A., Choi, Y., Morlighem, M., Rignot, E., Helm, V., Humbert, A., Mouginot, J., Millan, R., Kjær, K. H., and Bjørk, A. A. (2022). Extensive inland thinning and speed-up of Northeast Greenland Ice Stream. Nature, 611, 727–732, https://doi.org/10.1038/s41586-022-05301-z\n\n- [[7](https://doi.org/10.1038/s43247-020-0001-2)] King, M. D., Howat, I. M., Candela, S. G., Noh, M. J., Jeong, S., Noël, B. P. Y., van den Broeke, M. R., Wouters, B., and Negrete, A. (2020). Dynamic ice loss from the Greenland Ice Sheet driven by sustained glacier retreat. Communications Earth & Environment, 1, Article 1. https://doi.org/10.1038/s43247-020-0001-2\n\n- [[8](https://doi.org/10.5194/tc-9-1039-2015)] Fürst, J. J., Goelzer, H., and Huybrechts, P. (2015). Ice-dynamic projections of the Greenland ice sheet in response to atmospheric and oceanic warming, The Cryosphere, 9, 1039–1062, https://doi.org/10.5194/tc-9-1039-2015\n\n- [[9](https://doi.org/10.1016/j.rse.2017.08.038)] Paul, F., Bolch, T., Briggs, K., Kääb, A., McMillan, M., McNabb, R., Nagler, T., Nuth, C., Rastner, P., Strozzi, T., and Wuite, J. (2017). Error sources and guidelines for quality assessment of glacier area, elevation change, and velocity products derived from satellite data in the Glaciers_cci project, Remote Sensing of Environment, 203, 256-275. https://doi.org/10.1016/j.rse.2017.08.038", "text_with_prefix": "EQC Quality Assessment: \"Ice sheet velocity data from satellite observations: temporal and spatial coverage and data completeness for trend analysis in glaciological applications\"\nDataset: satellite-greenland-ice-sheet-velocity [CDS]\nAspect: trend-assessment_q01 | Category: Satellite_ECVs\nSection: Ice sheet velocity data from satellite observations: temporal and spatial coverage and data completeness for trend analysis in glaciological applications > ℹ️ If you want to know more > References\n---\n- [[1](https://doi.org/10.3390/rs70709371)] Nagler, T., Rott, H., Hetzenecker, M., Wuite, J. and Potin, P. (2015). The Sentinel-1 Mission: New Opportunities for Ice Sheet Observations. Remote Sensing. 7(7):9371-9389. https://doi.org/10.3390/rs70709371\n\n- [[2](https://doi.org/10.1017/jog.2017.73)] Joughin, I., Smith, B. E., and Howat, I. M. (2018). A complete map of Greenland ice velocity derived from satellite data collected over 20 years, J. Glaciol., 64, 1–11, https://doi.org/10.1017/jog.2017.73\n\n- [[3](https://nsidc.org/data/nsidc-0777/versions/1)] Howat, I., Chudley, T., and Noh, M. (2022). MEaSUREs Greenland Ice Velocity: Selected Glacier Site Single-Pair Velocity Maps from Optical Images, https://doi.org/10.5067/b28fm2qvvywy\n\n- [[4](https://library.wmo.int/records/item/58111-the-2022-gcos-ecvs-requirements-gcos-245)] GCOS (Global Climate Observing System) (2022). The 2022 GCOS ECVs Requirements (GCOS-245). World Meteorological Organization: Geneva, Switzerland. doi: https://library.wmo.int/idurl/4/58111\n\n- [[5](https://doi.org/10.5194/tc-6-923-2012)] Bevan, S. L., Luckman, A. J., and Murray, T. (2012). Glacier dynamics over the last quarter of a century at Helheim, Kangerdlugssuaq and 14 other major Greenland outlet glaciers, The Cryosphere, 6, 923–937, https://doi.org/10.5194/tc-6-923-2012\n\n- [[6](https://doi.org/10.1038/s41586-022-05301-z)] Khan, S. A., Choi, Y., Morlighem, M., Rignot, E., Helm, V., Humbert, A., Mouginot, J., Millan, R., Kjær, K. H., and Bjørk, A. A. (2022). Extensive inland thinning and speed-up of Northeast Greenland Ice Stream. Nature, 611, 727–732, https://doi.org/10.1038/s41586-022-05301-z\n\n- [[7](https://doi.org/10.1038/s43247-020-0001-2)] King, M. D., Howat, I. M., Candela, S. G., Noh, M. J., Jeong, S., Noël, B. P. Y., van den Broeke, M. R., Wouters, B., and Negrete, A. (2020). Dynamic ice loss from the Greenland Ice Sheet driven by sustained glacier retreat. Communications Earth & Environment, 1, Article 1. https://doi.org/10.1038/s43247-020-0001-2\n\n- [[8](https://doi.org/10.5194/tc-9-1039-2015)] Fürst, J. J., Goelzer, H., and Huybrechts, P. (2015). Ice-dynamic projections of the Greenland ice sheet in response to atmospheric and oceanic warming, The Cryosphere, 9, 1039–1062, https://doi.org/10.5194/tc-9-1039-2015\n\n- [[9](https://doi.org/10.1016/j.rse.2017.08.038)] Paul, F., Bolch, T., Briggs, K., Kääb, A., McMillan, M., McNabb, R., Nagler, T., Nuth, C., Rastner, P., Strozzi, T., and Wuite, J. (2017). Error sources and guidelines for quality assessment of glacier area, elevation change, and velocity products derived from satellite data in the Glaciers_cci project, Remote Sensing of Environment, 203, 256-275. https://doi.org/10.1016/j.rse.2017.08.038"} {"chunk_id": "satellite_satellite-greenland-ice-sheet-velocity_trend-assessment_q01__7bd90d5ff855", "report_id": "satellite_satellite-greenland-ice-sheet-velocity_trend-assessment_q01", "dataset_id": "satellite-greenland-ice-sheet-velocity", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q01", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Ice sheet velocity data from satellite observations: temporal and spatial coverage and data completeness for trend analysis in glaciological applications > ℹ️ If you want to know more > References", "title": "Ice sheet velocity data from satellite observations: temporal and spatial coverage and data completeness for trend analysis in glaciological applications", "chunk_index": 18, "token_count": 612, "text_raw": "velocity products derived from satellite data in the Glaciers_cci project, Remote Sensing of Environment, 203, 256-275. https://doi.org/10.1016/j.rse.2017.08.038\n\n- [[10](https://doi.org/10.5194/essd-12-1367-2020)] Mankoff, K.D., Solgaard, A., Colgan, W., Ahlstrøm, A.P., Khan, S.A., and Fausto, R.S. (2020). Greenland Ice Sheet solid ice discharge from 1986 through March 2020, Earth System Science Data, 12, 1367–1383. https://doi.org/10.5194/essd-12-1367-2020\n\n- [[11](https://doi.org/10.1175/BAMS-D-21-0109.1)] Yang, C. X., Cagnazzo, C., Artale, V., Nardelli, B. B., Buontempo, C., Busatto, J., Caporaso, L., Cesarini, C., Cionni, I., Coll, J., Crezee, B., Cristofanelli, P., de Toma, V., Essa, Y. H., Eyring, V., Fierli, F., Grant, L., Hassler, B., Hirschi, M., Huybrechts, P., Le Merle, E., Leonelli, F. E., Lin, X., Madonna, F., Mason, E., Massonnet, F., Marcos, M., Marullo, S., Muller, B., Obregon, A., Organelli, E., Palacz, A., Pascual, A., Pisano, A., Putero, D., Rana, A., Sanchez-Roman, A., Seneviratne, S. I., Serva, F., Storto, A., Thiery, W., Throne, P., Van Tricht, L., Verhaegen, Y., Volpe, G., and Santoleri, R. (2022). Independent Quality Assessment of Essential Climate Variables: Lessons Learned from the Copernicus Climate Change Service, B. Am. Meteorol. Soc., 103, E2032–E2049, https://doi.org/10.1175/Bams-D-21-0109.1", "text_with_prefix": "EQC Quality Assessment: \"Ice sheet velocity data from satellite observations: temporal and spatial coverage and data completeness for trend analysis in glaciological applications\"\nDataset: satellite-greenland-ice-sheet-velocity [CDS]\nAspect: trend-assessment_q01 | Category: Satellite_ECVs\nSection: Ice sheet velocity data from satellite observations: temporal and spatial coverage and data completeness for trend analysis in glaciological applications > ℹ️ If you want to know more > References\n---\nvelocity products derived from satellite data in the Glaciers_cci project, Remote Sensing of Environment, 203, 256-275. https://doi.org/10.1016/j.rse.2017.08.038\n\n- [[10](https://doi.org/10.5194/essd-12-1367-2020)] Mankoff, K.D., Solgaard, A., Colgan, W., Ahlstrøm, A.P., Khan, S.A., and Fausto, R.S. (2020). Greenland Ice Sheet solid ice discharge from 1986 through March 2020, Earth System Science Data, 12, 1367–1383. https://doi.org/10.5194/essd-12-1367-2020\n\n- [[11](https://doi.org/10.1175/BAMS-D-21-0109.1)] Yang, C. X., Cagnazzo, C., Artale, V., Nardelli, B. B., Buontempo, C., Busatto, J., Caporaso, L., Cesarini, C., Cionni, I., Coll, J., Crezee, B., Cristofanelli, P., de Toma, V., Essa, Y. H., Eyring, V., Fierli, F., Grant, L., Hassler, B., Hirschi, M., Huybrechts, P., Le Merle, E., Leonelli, F. E., Lin, X., Madonna, F., Mason, E., Massonnet, F., Marcos, M., Marullo, S., Muller, B., Obregon, A., Organelli, E., Palacz, A., Pascual, A., Pisano, A., Putero, D., Rana, A., Sanchez-Roman, A., Seneviratne, S. I., Serva, F., Storto, A., Thiery, W., Throne, P., Van Tricht, L., Verhaegen, Y., Volpe, G., and Santoleri, R. (2022). Independent Quality Assessment of Essential Climate Variables: Lessons Learned from the Copernicus Climate Change Service, B. Am. Meteorol. Soc., 103, E2032–E2049, https://doi.org/10.1175/Bams-D-21-0109.1"} {"chunk_id": "satellite_satellite-greenland-ice-sheet-velocity_uncertainty_q02__a1bb39c948d4", "report_id": "satellite_satellite-greenland-ice-sheet-velocity_uncertainty_q02", "dataset_id": "satellite-greenland-ice-sheet-velocity", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty_q02", "aspect_base": "uncertainty", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications", "title": "Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications", "chunk_index": 0, "token_count": 131, "text_raw": "Production date: 31-05-2025\n\nDataset version: 1.3 to 1.5\n\nProduced by: Yoni Verhaegen and Philippe Huybrechts (Vrije Universiteit Brussel)", "text_with_prefix": "EQC Quality Assessment: \"Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications\"\nDataset: satellite-greenland-ice-sheet-velocity [CDS]\nAspect: uncertainty_q02 | Category: Satellite_ECVs\nSection: Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications\n---\nProduction date: 31-05-2025\n\nDataset version: 1.3 to 1.5\n\nProduced by: Yoni Verhaegen and Philippe Huybrechts (Vrije Universiteit Brussel)"} {"chunk_id": "satellite_satellite-greenland-ice-sheet-velocity_uncertainty_q02__88e47e60f42d", "report_id": "satellite_satellite-greenland-ice-sheet-velocity_uncertainty_q02", "dataset_id": "satellite-greenland-ice-sheet-velocity", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty_q02", "aspect_base": "uncertainty", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications > Quality assessment question", "title": "Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications", "chunk_index": 1, "token_count": 518, "text_raw": "* **\"How does the uncertainty of the horizontal ice sheet surface flow velocity vary in space and time and is the dataset sufficiently accurate and precise to be used in an ice sheet modeling framework?\"**\n\nIce sheets are not static, but instead flow. The total velocity magnitude is the sum of the velocity related to internal ice deformation, and is complemented by basal sliding and bed deformation components. Mapping ice sheet surface flow velocities and their temporal changes provides key information for investigating the dynamic response of the ice sheets to climate change. Remote sensing techniques, such as the use of satellites, are an important feature to derive and study these flow velocities. Satellites are able to inspect directly and repeatedly large areas of ice, and, as such are able to detect the movement of its surface. Remote sensing techniques that use satellite data are therefore considered the only feasible manner to derive accurate surface velocities of the ice sheets on a regular basis. In this dataset, they are derived by applying offset tracking techniques using Sentinel-1 synthetic aperture radar (SAR) satellite data. The main principle of the \"[Ice sheet velocity for Antarctica and Greenland derived from satellite observations](https://cds.climate.copernicus.eu/datasets/satellite-greenland-ice-sheet-velocity?tab=overview)\" dataset therefore relies on the preservation of surface features or other detectable patterns (e.g. speckle) in between multiple image acquisition periods. The extraction of these features/patterns from images acquired over different time periods is used to detect their displacements, and with further processing this information can be used to derive velocity fields [[1](https://doi.org/10.3390/rs70709371), [11](https://doi.org/10.1016/j.rse.2025.115092)].\n\nThis notebook investigates how well the dataset on the CDS (here we use versions 1.3 to 1.5) can be used in the context of an ice sheet modelling framework for Greenland and Antarctica. More specifically, the notebook evaluates whether the dataset is of sufficient maturity and quality for that purpose in terms of its spatio-temporal coverage and uncertainty.", "text_with_prefix": "EQC Quality Assessment: \"Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications\"\nDataset: satellite-greenland-ice-sheet-velocity [CDS]\nAspect: uncertainty_q02 | Category: Satellite_ECVs\nSection: Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications > Quality assessment question\n---\n* **\"How does the uncertainty of the horizontal ice sheet surface flow velocity vary in space and time and is the dataset sufficiently accurate and precise to be used in an ice sheet modeling framework?\"**\n\nIce sheets are not static, but instead flow. The total velocity magnitude is the sum of the velocity related to internal ice deformation, and is complemented by basal sliding and bed deformation components. Mapping ice sheet surface flow velocities and their temporal changes provides key information for investigating the dynamic response of the ice sheets to climate change. Remote sensing techniques, such as the use of satellites, are an important feature to derive and study these flow velocities. Satellites are able to inspect directly and repeatedly large areas of ice, and, as such are able to detect the movement of its surface. Remote sensing techniques that use satellite data are therefore considered the only feasible manner to derive accurate surface velocities of the ice sheets on a regular basis. In this dataset, they are derived by applying offset tracking techniques using Sentinel-1 synthetic aperture radar (SAR) satellite data. The main principle of the \"[Ice sheet velocity for Antarctica and Greenland derived from satellite observations](https://cds.climate.copernicus.eu/datasets/satellite-greenland-ice-sheet-velocity?tab=overview)\" dataset therefore relies on the preservation of surface features or other detectable patterns (e.g. speckle) in between multiple image acquisition periods. The extraction of these features/patterns from images acquired over different time periods is used to detect their displacements, and with further processing this information can be used to derive velocity fields [[1](https://doi.org/10.3390/rs70709371), [11](https://doi.org/10.1016/j.rse.2025.115092)].\n\nThis notebook investigates how well the dataset on the CDS (here we use versions 1.3 to 1.5) can be used in the context of an ice sheet modelling framework for Greenland and Antarctica. More specifically, the notebook evaluates whether the dataset is of sufficient maturity and quality for that purpose in terms of its spatio-temporal coverage and uncertainty."} {"chunk_id": "satellite_satellite-greenland-ice-sheet-velocity_uncertainty_q02__2ff724245903", "report_id": "satellite_satellite-greenland-ice-sheet-velocity_uncertainty_q02", "dataset_id": "satellite-greenland-ice-sheet-velocity", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty_q02", "aspect_base": "uncertainty", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications > Quality assessment statement", "title": "Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications", "chunk_index": 2, "token_count": 364, "text_raw": "These are the key outcomes of this assessment\n\n- Data quality of the C3S ice sheet velocity dataset in terms of error and uncertainty can be assessed by two variables: the standard deviation (as a precision error estimate) and the valid pixel count maps (as the total number of valid measurements used in the annually averaged velocity estimate over 1 year of data acquisition). Consulting the accompanying standard deviation and the valid measurement count maps is therefore recommended to assess the error and uncertainty characterization of the product. The majority of pixels meet the proposed GCOS thresholds in terms of uncertainty. However, data quality can be slightly lower in some regions (due to 'striping', higher error values and/or lower valid measurements). Errors are furthermore provided as absolute values, while for flow fields, the relative error is generally considered to be more useful.\n- Due to the high spatial resolution/coverage and the quality-rich uncertainty characterization, the C3S ice sheet velocity dataset is particularly well-suited for the use as a calibration or validation tool within an ice sheet modelling framework. Validation efforts furthermore show good agreement with independent data. The user can, if desired, prioritize high-quality data (i.e. pixels with low absolute or relative errors and/or high valid pixel counts) when adjusting tuning parameters in an ice sheet model, such that the modelled ice thickness or surface velocity matches the observed one as close as possible.\n```", "text_with_prefix": "EQC Quality Assessment: \"Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications\"\nDataset: satellite-greenland-ice-sheet-velocity [CDS]\nAspect: uncertainty_q02 | Category: Satellite_ECVs\nSection: Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications > Quality assessment statement\n---\nThese are the key outcomes of this assessment\n\n- Data quality of the C3S ice sheet velocity dataset in terms of error and uncertainty can be assessed by two variables: the standard deviation (as a precision error estimate) and the valid pixel count maps (as the total number of valid measurements used in the annually averaged velocity estimate over 1 year of data acquisition). Consulting the accompanying standard deviation and the valid measurement count maps is therefore recommended to assess the error and uncertainty characterization of the product. The majority of pixels meet the proposed GCOS thresholds in terms of uncertainty. However, data quality can be slightly lower in some regions (due to 'striping', higher error values and/or lower valid measurements). Errors are furthermore provided as absolute values, while for flow fields, the relative error is generally considered to be more useful.\n- Due to the high spatial resolution/coverage and the quality-rich uncertainty characterization, the C3S ice sheet velocity dataset is particularly well-suited for the use as a calibration or validation tool within an ice sheet modelling framework. Validation efforts furthermore show good agreement with independent data. The user can, if desired, prioritize high-quality data (i.e. pixels with low absolute or relative errors and/or high valid pixel counts) when adjusting tuning parameters in an ice sheet model, such that the modelled ice thickness or surface velocity matches the observed one as close as possible.\n```"} {"chunk_id": "satellite_satellite-greenland-ice-sheet-velocity_uncertainty_q02__1f1473316cc9", "report_id": "satellite_satellite-greenland-ice-sheet-velocity_uncertainty_q02", "dataset_id": "satellite-greenland-ice-sheet-velocity", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty_q02", "aspect_base": "uncertainty", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications > Methodology > Dataset description", "title": "Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications", "chunk_index": 3, "token_count": 322, "text_raw": "The 'Ice sheet velocity for Antarctica and Greenland derived from satellite observations' dataset, available on the Climate Data Store (CDS), offers annually averaged surface ice flow velocities and their easting and northing components for the Antarctic (AIS) and Greenland (GrIS) Ice Sheets. This data is provided on a spatial resolution grid of either 200 m, 250 m or 500 m, depending on the ice sheet and dataset version. The dataset includes horizontal velocity components in the east and north directions, as well as the total horizontal velocity magnitude and its uncertainty, expressed as 1-sigma precission errors. Surface flow velocities are treated as “displacements”, normalized to true meters per day, and are available for each glaciological year since 2014-2015 for the GrIS and since 2021-2022 for the AIS. The horizontal velocities and their components are derived using offset tracking techniques with Sentinel-1 synthetic aperture radar (SAR) satellite data. The data are provided in NetCDF format as gridded data fields in Polar Stereographic projection, covering the entire GrIS domain, including peripheral glaciers and ice caps.", "text_with_prefix": "EQC Quality Assessment: \"Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications\"\nDataset: satellite-greenland-ice-sheet-velocity [CDS]\nAspect: uncertainty_q02 | Category: Satellite_ECVs\nSection: Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications > Methodology > Dataset description\n---\nThe 'Ice sheet velocity for Antarctica and Greenland derived from satellite observations' dataset, available on the Climate Data Store (CDS), offers annually averaged surface ice flow velocities and their easting and northing components for the Antarctic (AIS) and Greenland (GrIS) Ice Sheets. This data is provided on a spatial resolution grid of either 200 m, 250 m or 500 m, depending on the ice sheet and dataset version. The dataset includes horizontal velocity components in the east and north directions, as well as the total horizontal velocity magnitude and its uncertainty, expressed as 1-sigma precission errors. Surface flow velocities are treated as “displacements”, normalized to true meters per day, and are available for each glaciological year since 2014-2015 for the GrIS and since 2021-2022 for the AIS. The horizontal velocities and their components are derived using offset tracking techniques with Sentinel-1 synthetic aperture radar (SAR) satellite data. The data are provided in NetCDF format as gridded data fields in Polar Stereographic projection, covering the entire GrIS domain, including peripheral glaciers and ice caps."} {"chunk_id": "satellite_satellite-greenland-ice-sheet-velocity_uncertainty_q02__55ba056fced7", "report_id": "satellite_satellite-greenland-ice-sheet-velocity_uncertainty_q02", "dataset_id": "satellite-greenland-ice-sheet-velocity", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty_q02", "aspect_base": "uncertainty", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications > Methodology > Structure and (sub)sections", "title": "Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications", "chunk_index": 4, "token_count": 214, "text_raw": "**[](section-1)**\n\n* [](section-1-1)\n* [](section-1-2)\n* [](section-1-3)\n\n**[](section-2)**\n* [](section-2-1)\n* [](section-2-2)\n* [](section-2-3)\n* [](section-2-4)\n\n**[](section-3)**\n* [](section-3-1)\n* [](section-3-2)\n\n**[](section-4)**\n* [](section-4-1)\n* [](section-4-2)\n\n**[](section-5)**", "text_with_prefix": "EQC Quality Assessment: \"Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications\"\nDataset: satellite-greenland-ice-sheet-velocity [CDS]\nAspect: uncertainty_q02 | Category: Satellite_ECVs\nSection: Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications > Methodology > Structure and (sub)sections\n---\n**[](section-1)**\n\n* [](section-1-1)\n* [](section-1-2)\n* [](section-1-3)\n\n**[](section-2)**\n* [](section-2-1)\n* [](section-2-2)\n* [](section-2-3)\n* [](section-2-4)\n\n**[](section-3)**\n* [](section-3-1)\n* [](section-3-2)\n\n**[](section-4)**\n* [](section-4-1)\n* [](section-4-2)\n\n**[](section-5)**"} {"chunk_id": "satellite_satellite-greenland-ice-sheet-velocity_uncertainty_q02__53b3bdf7b73d", "report_id": "satellite_satellite-greenland-ice-sheet-velocity_uncertainty_q02", "dataset_id": "satellite-greenland-ice-sheet-velocity", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty_q02", "aspect_base": "uncertainty", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications > Analysis and results > 1. Data preparation and processing > 1.2 Define request and download", "title": "Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications", "chunk_index": 5, "token_count": 314, "text_raw": "Then we define requests for download from the CDS and download the GrIS and AIS velocity data and immediately perform some data handling to make the dataset more readable.\n\n🚨 **The files can be large! Since the data files to be downloaded are of considerable size due to the high spatial resolution, this may take a couple of minutes.**\n\nAdd period dim\nCompute standard deviation\nFlow direction\nKeep only selected variables\n\n```text\nDownloading ice sheet velocity data, this make take a couple of minutes...\ndomain = 'greenland_ice_sheet'\n```\n\n```text\n100%|██████████| 9/9 [00:05<00:00, 1.57it/s]\n```\n\n```text\ndomain = 'antarctic_ice_sheet'\n```\n\n```text\n100%|██████████| 2/2 [00:00<00:00, 5.84it/s]\n```\n\n```text\nDownloading ice sheet velocity data complete.\n```\n\n(section-1-3)=", "text_with_prefix": "EQC Quality Assessment: \"Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications\"\nDataset: satellite-greenland-ice-sheet-velocity [CDS]\nAspect: uncertainty_q02 | Category: Satellite_ECVs\nSection: Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications > Analysis and results > 1. Data preparation and processing > 1.2 Define request and download\n---\nThen we define requests for download from the CDS and download the GrIS and AIS velocity data and immediately perform some data handling to make the dataset more readable.\n\n🚨 **The files can be large! Since the data files to be downloaded are of considerable size due to the high spatial resolution, this may take a couple of minutes.**\n\nAdd period dim\nCompute standard deviation\nFlow direction\nKeep only selected variables\n\n```text\nDownloading ice sheet velocity data, this make take a couple of minutes...\ndomain = 'greenland_ice_sheet'\n```\n\n```text\n100%|██████████| 9/9 [00:05<00:00, 1.57it/s]\n```\n\n```text\ndomain = 'antarctic_ice_sheet'\n```\n\n```text\n100%|██████████| 2/2 [00:00<00:00, 5.84it/s]\n```\n\n```text\nDownloading ice sheet velocity data complete.\n```\n\n(section-1-3)="} {"chunk_id": "satellite_satellite-greenland-ice-sheet-velocity_uncertainty_q02__02fb14cfbcdb", "report_id": "satellite_satellite-greenland-ice-sheet-velocity_uncertainty_q02", "dataset_id": "satellite-greenland-ice-sheet-velocity", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty_q02", "aspect_base": "uncertainty", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications > Analysis and results > 1. Data preparation and processing > 1.3 Display and inspect data structure", "title": "Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications", "chunk_index": 6, "token_count": 1105, "text_raw": ":\n\n```text\n{'greenland_ice_sheet': Size: 12GB\n Dimensions: (period: 9, y: 10801, x: 5984)\n Coordinates:\n * period (period) \n land_ice_surface_measurement_count (period, y, x) float64 5GB dask.array\n land_ice_surface_stddev (period, y, x) float32 2GB dask.array\n land_ice_flow_direction (period, y, x) float32 2GB dask.array\n Attributes: (12/13)\n Conventions: CF-1.7\n title: Ice Velocity of the Greenland Ice Sheet\n reference: Main: Nagler, T.; Rott, H.; Hetzenecker, M.; Wuite, J.; P...\n source: Copernicus Sentinel-1A and Sentinel-1B\n institution: Copernicus Climate Change Service\n contact: copernicus-support@ecmwf.int\n ... ...\n creation_date: 2024-12-12\n comment: Ice velocity map of Greenland derived from Sentinel-1 SAR...\n history: product version 1.3\n summary: Ice velocity derived for Greenland Ice Sheet gridded at 2...\n keywords: EARTH SCIENCE CLIMATE INDICATORS CRYOSPHERIC INDICATORS G...\n license: C3S general license,\n 'antarctic_ice_sheet': Size: 28GB\n Dimensions: (period: 2, y: 24580, x: 28680)\n Coordinates:\n * period (period) \n land_ice_surface_measurement_count (period, y, x) float64 11GB dask.array\n land_ice_surface_stddev (period, y, x) float32 6GB dask.array\n land_ice_flow_direction (period, y, x) float32 6GB dask.array\n Attributes: (12/13)\n Conventions: CF-1.7\n title: Ice Velocity of the Antarctic Ice Sheet\n reference: Main: Nagler, T.; Rott, H.; Hetzenecker, M.; Wuite, J.; P...\n source: Copernicus Sentinel-1A and Sentinel-1B\n institution: Copernicus Climate Change Service\n contact: copernicus-support@ecmwf.int\n ... ...\n creation_date: 2023-12-15\n comment: Ice velocity map of Antarctica derived from Sentinel-1 SA...\n history: product version 1.5\n summary: Ice velocity derived for Antarctic Ice Sheet gridded at 2...\n keywords: EARTH SCIENCE CLIMATE INDICATORS CRYOSPHERIC INDICATORS G...\n license", "text_with_prefix": "EQC Quality Assessment: \"Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications\"\nDataset: satellite-greenland-ice-sheet-velocity [CDS]\nAspect: uncertainty_q02 | Category: Satellite_ECVs\nSection: Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications > Analysis and results > 1. Data preparation and processing > 1.3 Display and inspect data structure\n---\n:\n\n```text\n{'greenland_ice_sheet': Size: 12GB\n Dimensions: (period: 9, y: 10801, x: 5984)\n Coordinates:\n * period (period) \n land_ice_surface_measurement_count (period, y, x) float64 5GB dask.array\n land_ice_surface_stddev (period, y, x) float32 2GB dask.array\n land_ice_flow_direction (period, y, x) float32 2GB dask.array\n Attributes: (12/13)\n Conventions: CF-1.7\n title: Ice Velocity of the Greenland Ice Sheet\n reference: Main: Nagler, T.; Rott, H.; Hetzenecker, M.; Wuite, J.; P...\n source: Copernicus Sentinel-1A and Sentinel-1B\n institution: Copernicus Climate Change Service\n contact: copernicus-support@ecmwf.int\n ... ...\n creation_date: 2024-12-12\n comment: Ice velocity map of Greenland derived from Sentinel-1 SAR...\n history: product version 1.3\n summary: Ice velocity derived for Greenland Ice Sheet gridded at 2...\n keywords: EARTH SCIENCE CLIMATE INDICATORS CRYOSPHERIC INDICATORS G...\n license: C3S general license,\n 'antarctic_ice_sheet': Size: 28GB\n Dimensions: (period: 2, y: 24580, x: 28680)\n Coordinates:\n * period (period) \n land_ice_surface_measurement_count (period, y, x) float64 11GB dask.array\n land_ice_surface_stddev (period, y, x) float32 6GB dask.array\n land_ice_flow_direction (period, y, x) float32 6GB dask.array\n Attributes: (12/13)\n Conventions: CF-1.7\n title: Ice Velocity of the Antarctic Ice Sheet\n reference: Main: Nagler, T.; Rott, H.; Hetzenecker, M.; Wuite, J.; P...\n source: Copernicus Sentinel-1A and Sentinel-1B\n institution: Copernicus Climate Change Service\n contact: copernicus-support@ecmwf.int\n ... ...\n creation_date: 2023-12-15\n comment: Ice velocity map of Antarctica derived from Sentinel-1 SA...\n history: product version 1.5\n summary: Ice velocity derived for Antarctic Ice Sheet gridded at 2...\n keywords: EARTH SCIENCE CLIMATE INDICATORS CRYOSPHERIC INDICATORS G...\n license"} {"chunk_id": "satellite_satellite-greenland-ice-sheet-velocity_uncertainty_q02__3c1be00ad089", "report_id": "satellite_satellite-greenland-ice-sheet-velocity_uncertainty_q02", "dataset_id": "satellite-greenland-ice-sheet-velocity", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty_q02", "aspect_base": "uncertainty", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications > Analysis and results > 1. Data preparation and processing > 1.3 Display and inspect data structure", "title": "Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications", "chunk_index": 7, "token_count": 164, "text_raw": "product version 1.5\n summary: Ice velocity derived for Antarctic Ice Sheet gridded at 2...\n keywords: EARTH SCIENCE CLIMATE INDICATORS CRYOSPHERIC INDICATORS G...\n license\n\n: C3S general license}\n```", "text_with_prefix": "EQC Quality Assessment: \"Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications\"\nDataset: satellite-greenland-ice-sheet-velocity [CDS]\nAspect: uncertainty_q02 | Category: Satellite_ECVs\nSection: Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications > Analysis and results > 1. Data preparation and processing > 1.3 Display and inspect data structure\n---\nproduct version 1.5\n summary: Ice velocity derived for Antarctic Ice Sheet gridded at 2...\n keywords: EARTH SCIENCE CLIMATE INDICATORS CRYOSPHERIC INDICATORS G...\n license\n\n: C3S general license}\n```"} {"chunk_id": "satellite_satellite-greenland-ice-sheet-velocity_uncertainty_q02__67a48417c0d2", "report_id": "satellite_satellite-greenland-ice-sheet-velocity_uncertainty_q02", "dataset_id": "satellite-greenland-ice-sheet-velocity", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty_q02", "aspect_base": "uncertainty", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications > Analysis and results > 1. Data preparation and processing > 1.3 Display and inspect data structure", "title": "Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications", "chunk_index": 8, "token_count": 668, "text_raw": "```\n\nFor the GrIS, the versions 1.3 to 1.5 are a gridded dataset at a 250 m spatial resolution containing annually averaged values of the ice sheet horizontal surface flow velocity $V_s$ (in m/day) of a grid cell (`land_ice_surface_velocity_magnitude`) and its components (`land_ice_surface_easting_velocity` ($V_x$) and `land_ice_surface_northing_velocity` ($V_y$)) since the 2014-2015 hydrological year. The uncertainties (reported as precision errors or standard deviations) of the horizontal components are also given as `land_ice_surface_easting_stddev` ($\\sigma_{V_x}$) and `land_ice_surface_northing_stddev` ($\\sigma_{V_y}$). Also the vertical component is given as `land_ice_surface_vertical_velocity` ($V_z$), but without a corresponding uncertainty. Next to that, also a variable named `land_ice_surface_measurement_count` is present, which quantifies the total number of valid image pairs used in the annually averaged velocity estimate for a certain pixel.\n\nFor the AIS, only version 1.5 is available at a spatial resolution of 200 m since the 2021-2022 period. After the data handling we performed, our dataset array now only holds the most important information: the total horizontal velocity magnitude (`land_ice_surface_velocity_magnitude`), the valid pixel count (`land_ice_surface_measurement_count`), the calculated ice flow direction (`land_ice_flow_direction`), and our calculated standard deviation (`land_ice_surface_stddev`) from the northing and easting components.\n\nFor clarification, the total horizontal surface velocity magnitude $V_s$ and its uncertainty $\\sigma_{V_s}$ are calculated from its components as follows (with subscript 's' referring to the surface):\n\n$\nV_s\n$\n[m day⁻¹] \n$\n= \\sqrt {V_x^2 + V_y^2}\n$\n\n$\n\\sigma_{V_s}\n$\n[m day⁻¹] \n$\n= \\sqrt {(\\sigma_{V_x})^2 + (\\sigma_{V_y})^2}\n$\n\nwhere $V_x$ and $V_y$ are respectively the easting and northing components of the horizontal velocity vector.\n\nThe flow direction of the ice can be calculated as follows:\n\n$\n\\theta_{\\text{V}_s} = \\left( 90^\\circ - \\left( \\arctan2(v_y, v_x) \\times \\left( \\frac{180}{\\pi} \\right) \\right) \\right) \\mod 360^\\circ\n$\n\nLet us begin with the analysis.\n\n(section-2)=", "text_with_prefix": "EQC Quality Assessment: \"Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications\"\nDataset: satellite-greenland-ice-sheet-velocity [CDS]\nAspect: uncertainty_q02 | Category: Satellite_ECVs\nSection: Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications > Analysis and results > 1. Data preparation and processing > 1.3 Display and inspect data structure\n---\n```\n\nFor the GrIS, the versions 1.3 to 1.5 are a gridded dataset at a 250 m spatial resolution containing annually averaged values of the ice sheet horizontal surface flow velocity $V_s$ (in m/day) of a grid cell (`land_ice_surface_velocity_magnitude`) and its components (`land_ice_surface_easting_velocity` ($V_x$) and `land_ice_surface_northing_velocity` ($V_y$)) since the 2014-2015 hydrological year. The uncertainties (reported as precision errors or standard deviations) of the horizontal components are also given as `land_ice_surface_easting_stddev` ($\\sigma_{V_x}$) and `land_ice_surface_northing_stddev` ($\\sigma_{V_y}$). Also the vertical component is given as `land_ice_surface_vertical_velocity` ($V_z$), but without a corresponding uncertainty. Next to that, also a variable named `land_ice_surface_measurement_count` is present, which quantifies the total number of valid image pairs used in the annually averaged velocity estimate for a certain pixel.\n\nFor the AIS, only version 1.5 is available at a spatial resolution of 200 m since the 2021-2022 period. After the data handling we performed, our dataset array now only holds the most important information: the total horizontal velocity magnitude (`land_ice_surface_velocity_magnitude`), the valid pixel count (`land_ice_surface_measurement_count`), the calculated ice flow direction (`land_ice_flow_direction`), and our calculated standard deviation (`land_ice_surface_stddev`) from the northing and easting components.\n\nFor clarification, the total horizontal surface velocity magnitude $V_s$ and its uncertainty $\\sigma_{V_s}$ are calculated from its components as follows (with subscript 's' referring to the surface):\n\n$\nV_s\n$\n[m day⁻¹] \n$\n= \\sqrt {V_x^2 + V_y^2}\n$\n\n$\n\\sigma_{V_s}\n$\n[m day⁻¹] \n$\n= \\sqrt {(\\sigma_{V_x})^2 + (\\sigma_{V_y})^2}\n$\n\nwhere $V_x$ and $V_y$ are respectively the easting and northing components of the horizontal velocity vector.\n\nThe flow direction of the ice can be calculated as follows:\n\n$\n\\theta_{\\text{V}_s} = \\left( 90^\\circ - \\left( \\arctan2(v_y, v_x) \\times \\left( \\frac{180}{\\pi} \\right) \\right) \\right) \\mod 360^\\circ\n$\n\nLet us begin with the analysis.\n\n(section-2)="} {"chunk_id": "satellite_satellite-greenland-ice-sheet-velocity_uncertainty_q02__be8755d02128", "report_id": "satellite_satellite-greenland-ice-sheet-velocity_uncertainty_q02", "dataset_id": "satellite-greenland-ice-sheet-velocity", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty_q02", "aspect_base": "uncertainty", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications > Analysis and results > 2. Ice sheet flow velocities and their uncertainty estimates in space and time > 2.1 Average horizontal ice flow velocities", "title": "Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications", "chunk_index": 9, "token_count": 551, "text_raw": "We begin by plotting the average ice flow velocities for each hydrological year between the beginning and end period that we selected for the velocity data with a defined plotting function. Let us begin with the GrIS:\n\n🚨 **The files can be large! Since the data files to be plotted are of considerable size, this may take a couple of minutes.**\n\nGreenland setup\nExtract and preprocess data\nPlot Greenland\nColorbar with formatting\n\n*Figure 1. Magnitude of the average horizontal surface flow velocities over Greenland over the defined time period.*\n\nAnd now also the AIS:\n\nAntarctica setup\nExtract and preprocess data\nPlot Antarctica\nColorbar with formatting\n\n*Figure 2. Magnitude of the average horizontal surface flow velocities over Antarctica over the defined time period.*\n\nThe corresponding data are annually averaged values, and are hence derived from all-year round observations during the glaciological balance year. The annually averaged nature of the data meets the minimum requirement for GCOS (Global Climate Observing System) [[2](https://library.wmo.int/records/item/58111-the-2022-gcos-ecvs-requirements-gcos-245)].\n\nFor the GrIS, the dataset shows almost no temporal or spatial (as most have been filled up) data gaps and is presented as gridded values at a spatial resolution of 250 meter (for versions 1.3 to 1.5 as used here), even though spatially filled gaps are not flagged and hence not detectable. The spatial coverage of the data thus approximates 100% and encompasses the entire ice sheet, including peripheral glaciers and ice caps. However, an ice mask is not included and it is thus not possible to exclude non ice-covered pixels from the data. Pixels over ice-free terrain hold non-NaN data and have not been removed, implying that the data exhibits false low velocity measurements and noise over ice-free terrain.\n\nFor the AIS, most of the margin is covered by the C3S product, but no values are present over the interior of the ice sheet due to the Sentinel-1 acquisition mask (data is only collected at the margins).\n\n(section-2-2)=", "text_with_prefix": "EQC Quality Assessment: \"Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications\"\nDataset: satellite-greenland-ice-sheet-velocity [CDS]\nAspect: uncertainty_q02 | Category: Satellite_ECVs\nSection: Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications > Analysis and results > 2. Ice sheet flow velocities and their uncertainty estimates in space and time > 2.1 Average horizontal ice flow velocities\n---\nWe begin by plotting the average ice flow velocities for each hydrological year between the beginning and end period that we selected for the velocity data with a defined plotting function. Let us begin with the GrIS:\n\n🚨 **The files can be large! Since the data files to be plotted are of considerable size, this may take a couple of minutes.**\n\nGreenland setup\nExtract and preprocess data\nPlot Greenland\nColorbar with formatting\n\n*Figure 1. Magnitude of the average horizontal surface flow velocities over Greenland over the defined time period.*\n\nAnd now also the AIS:\n\nAntarctica setup\nExtract and preprocess data\nPlot Antarctica\nColorbar with formatting\n\n*Figure 2. Magnitude of the average horizontal surface flow velocities over Antarctica over the defined time period.*\n\nThe corresponding data are annually averaged values, and are hence derived from all-year round observations during the glaciological balance year. The annually averaged nature of the data meets the minimum requirement for GCOS (Global Climate Observing System) [[2](https://library.wmo.int/records/item/58111-the-2022-gcos-ecvs-requirements-gcos-245)].\n\nFor the GrIS, the dataset shows almost no temporal or spatial (as most have been filled up) data gaps and is presented as gridded values at a spatial resolution of 250 meter (for versions 1.3 to 1.5 as used here), even though spatially filled gaps are not flagged and hence not detectable. The spatial coverage of the data thus approximates 100% and encompasses the entire ice sheet, including peripheral glaciers and ice caps. However, an ice mask is not included and it is thus not possible to exclude non ice-covered pixels from the data. Pixels over ice-free terrain hold non-NaN data and have not been removed, implying that the data exhibits false low velocity measurements and noise over ice-free terrain.\n\nFor the AIS, most of the margin is covered by the C3S product, but no values are present over the interior of the ice sheet due to the Sentinel-1 acquisition mask (data is only collected at the margins).\n\n(section-2-2)="} {"chunk_id": "satellite_satellite-greenland-ice-sheet-velocity_uncertainty_q02__7aa4ca1d3497", "report_id": "satellite_satellite-greenland-ice-sheet-velocity_uncertainty_q02", "dataset_id": "satellite-greenland-ice-sheet-velocity", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty_q02", "aspect_base": "uncertainty", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications > Analysis and results > 2. Ice sheet flow velocities and their uncertainty estimates in space and time > 2.2 Ice flow velocity uncertainty: accuracy and precision", "title": "Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications", "chunk_index": 10, "token_count": 731, "text_raw": "Let us now consider the unceratinty of the dataset. The total error of an annual ice flow velocity estimate is theoretically given by the sum of the precision (random) and the accuracy (systematic) error:\n\n$\n\\varepsilon = a\\sigma + \\delta\n$\nwhere $\\sigma$ is the random error (i.e. standard deviation), $a$ the critical z-score related to a certain confidence interval, and $\\delta$ the systematic error.\n\nIn the ice flow velocity dataset, precision errors are reported as the standard deviation (i.e. the 68% confidence interval) and the accuracy error is not provided. Therefore, in our case, $\\delta$ = 0.\n\nAccuracy errors are thus not directly included in the total error estimate of the dataset. To get an idea of the accuracy, the [PQAR](https://confluence.ecmwf.int/pages/viewpage.action?pageId=355348656) can be consulted, where the results of a validation study are summarized. The validation process involved comparing the velocity maps with in-situ GPS data from various sites, as well as with NASA's MEaSUREs velocity product [[3](https://nsidc.org/data/nsidc-0777/versions/1)]. Quality assurance furthermore included an algorithm performance assessment in stable terrain (where velocities should equal 0) [[6](https://doi.org/10.1016/j.rse.2017.08.038)]. The results demonstrate a relatively good agreement between the dataset and independent external data.\n\nQuantitative pixel-by-pixel error estimates are, on the other hand, available for the dataset in the form of precision errors and valid measurement counts. The uncertainty characterization is hence quantified by two variables: the standard deviation (i.e. a precision error from a 5x5 pixel neighbourhood) of the velocity estimate with units of m/day, and the valid pixel count (the number of non-NaN observations used in the production of each annual velocity estimate for a certain pixel, or differently stated, the amount of individual displacement estimates (i.e. image pairs) that were used to calculate the annually averaged velocity at that pixel). Let us express the valid pixel count variable with symbol $N_{VP}$.\n\nWe will proceed by changing the units of the error $\\varepsilon$ to m yr⁻¹ because the GCOS Implementation Plan report advises uncertainty values to be provided in these units [[2](https://library.wmo.int/records/item/58111-the-2022-gcos-ecvs-requirements-gcos-245)]. We will also multiply the standard deviation by 2 because the GCOS assesses precision errors in the form of 2 standard deviations:\n\n$\n\\varepsilon_{{V_s}}\n$\n[m yr⁻¹]\n$\n= 365.25 \\cdot (2 \\cdot \\sigma_{V_{s}})\n$\n\nWe can now perform some statistics on the error term over all pixels and all years to inspect its overall value for both ice sheets:\n\nShow measurements count", "text_with_prefix": "EQC Quality Assessment: \"Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications\"\nDataset: satellite-greenland-ice-sheet-velocity [CDS]\nAspect: uncertainty_q02 | Category: Satellite_ECVs\nSection: Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications > Analysis and results > 2. Ice sheet flow velocities and their uncertainty estimates in space and time > 2.2 Ice flow velocity uncertainty: accuracy and precision\n---\nLet us now consider the unceratinty of the dataset. The total error of an annual ice flow velocity estimate is theoretically given by the sum of the precision (random) and the accuracy (systematic) error:\n\n$\n\\varepsilon = a\\sigma + \\delta\n$\nwhere $\\sigma$ is the random error (i.e. standard deviation), $a$ the critical z-score related to a certain confidence interval, and $\\delta$ the systematic error.\n\nIn the ice flow velocity dataset, precision errors are reported as the standard deviation (i.e. the 68% confidence interval) and the accuracy error is not provided. Therefore, in our case, $\\delta$ = 0.\n\nAccuracy errors are thus not directly included in the total error estimate of the dataset. To get an idea of the accuracy, the [PQAR](https://confluence.ecmwf.int/pages/viewpage.action?pageId=355348656) can be consulted, where the results of a validation study are summarized. The validation process involved comparing the velocity maps with in-situ GPS data from various sites, as well as with NASA's MEaSUREs velocity product [[3](https://nsidc.org/data/nsidc-0777/versions/1)]. Quality assurance furthermore included an algorithm performance assessment in stable terrain (where velocities should equal 0) [[6](https://doi.org/10.1016/j.rse.2017.08.038)]. The results demonstrate a relatively good agreement between the dataset and independent external data.\n\nQuantitative pixel-by-pixel error estimates are, on the other hand, available for the dataset in the form of precision errors and valid measurement counts. The uncertainty characterization is hence quantified by two variables: the standard deviation (i.e. a precision error from a 5x5 pixel neighbourhood) of the velocity estimate with units of m/day, and the valid pixel count (the number of non-NaN observations used in the production of each annual velocity estimate for a certain pixel, or differently stated, the amount of individual displacement estimates (i.e. image pairs) that were used to calculate the annually averaged velocity at that pixel). Let us express the valid pixel count variable with symbol $N_{VP}$.\n\nWe will proceed by changing the units of the error $\\varepsilon$ to m yr⁻¹ because the GCOS Implementation Plan report advises uncertainty values to be provided in these units [[2](https://library.wmo.int/records/item/58111-the-2022-gcos-ecvs-requirements-gcos-245)]. We will also multiply the standard deviation by 2 because the GCOS assesses precision errors in the form of 2 standard deviations:\n\n$\n\\varepsilon_{{V_s}}\n$\n[m yr⁻¹]\n$\n= 365.25 \\cdot (2 \\cdot \\sigma_{V_{s}})\n$\n\nWe can now perform some statistics on the error term over all pixels and all years to inspect its overall value for both ice sheets:\n\nShow measurements count"} {"chunk_id": "satellite_satellite-greenland-ice-sheet-velocity_uncertainty_q02__79efc7be1b37", "report_id": "satellite_satellite-greenland-ice-sheet-velocity_uncertainty_q02", "dataset_id": "satellite-greenland-ice-sheet-velocity", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty_q02", "aspect_base": "uncertainty", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications > Analysis and results > 2. Ice sheet flow velocities and their uncertainty estimates in space and time > 2.2 Ice flow velocity uncertainty: accuracy and precision", "title": "Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications", "chunk_index": 11, "token_count": 814, "text_raw": "$\n\nWe can now perform some statistics on the error term over all pixels and all years to inspect its overall value for both ice sheets:\n\nShow measurements count\n\n```text\ndomain = 'greenland_ice_sheet'\nMean velocity error (2σ) for 2014-2015 CE: 20.42 m/yr\nMean velocity error (2σ) for 2015-2016 CE: 10.93 m/yr\nMean velocity error (2σ) for 2016-2017 CE: 9.92 m/yr\nMean velocity error (2σ) for 2017-2018 CE: 9.19 m/yr\nMean velocity error (2σ) for 2018-2019 CE: 7.90 m/yr\nMean velocity error (2σ) for 2019-2020 CE: 7.14 m/yr\nMean velocity error (2σ) for 2020-2021 CE: 328.67 m/yr\nMean velocity error (2σ) for 2021-2022 CE: 7.69 m/yr\nMean velocity error (2σ) for 2022-2023 CE: 12.84 m/yr\nMean measurement count for 2014-2015 CE: 19.75 estimates/pixel\nMean measurement count for 2015-2016 CE: 61.04 estimates/pixel\nMean measurement count for 2016-2017 CE: 108.80 estimates/pixel\nMean measurement count for 2017-2018 CE: 133.65 estimates/pixel\nMean measurement count for 2018-2019 CE: 138.09 estimates/pixel\nMean measurement count for 2019-2020 CE: 136.80 estimates/pixel\nMean measurement count for 2020-2021 CE: 122.56 estimates/pixel\nMean measurement count for 2021-2022 CE: 78.79 estimates/pixel\nMean measurement count for 2022-2023 CE: 51.34 estimates/pixel\n\ndomain = 'antarctic_ice_sheet'\nMean velocity error (2σ) for 2021-2022 CE: 14.09 m/yr\nMean velocity error (2σ) for 2022-2023 CE: 19.45 m/yr\nMean measurement count for 2021-2022 CE: 71.04 estimates/pixel\nMean measurement count for 2022-2023 CE: 45.61 estimates/pixel\n```\n\nThe threshold (i.e. the minimum requirement to be met to ensure that data are useful) for surface ice flow velocity uncertainty (expressed in terms of 2$\\sigma$) proposed by the GCOS is 100 m yr⁻¹, while the \"goal\" value (i.e. an ideal requirement above which further improvements are not necessary) would be 10 m yr⁻¹ per grid point [[2](https://library.wmo.int/records/item/58111-the-2022-gcos-ecvs-requirements-gcos-245)]. The overall arithmetic mean error over both space and time shows that the minimum threshold is reached for all years but the 2020-2021 period for the GrIS (which will be discussed later).\n\n(section-2-3)=", "text_with_prefix": "EQC Quality Assessment: \"Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications\"\nDataset: satellite-greenland-ice-sheet-velocity [CDS]\nAspect: uncertainty_q02 | Category: Satellite_ECVs\nSection: Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications > Analysis and results > 2. Ice sheet flow velocities and their uncertainty estimates in space and time > 2.2 Ice flow velocity uncertainty: accuracy and precision\n---\n$\n\nWe can now perform some statistics on the error term over all pixels and all years to inspect its overall value for both ice sheets:\n\nShow measurements count\n\n```text\ndomain = 'greenland_ice_sheet'\nMean velocity error (2σ) for 2014-2015 CE: 20.42 m/yr\nMean velocity error (2σ) for 2015-2016 CE: 10.93 m/yr\nMean velocity error (2σ) for 2016-2017 CE: 9.92 m/yr\nMean velocity error (2σ) for 2017-2018 CE: 9.19 m/yr\nMean velocity error (2σ) for 2018-2019 CE: 7.90 m/yr\nMean velocity error (2σ) for 2019-2020 CE: 7.14 m/yr\nMean velocity error (2σ) for 2020-2021 CE: 328.67 m/yr\nMean velocity error (2σ) for 2021-2022 CE: 7.69 m/yr\nMean velocity error (2σ) for 2022-2023 CE: 12.84 m/yr\nMean measurement count for 2014-2015 CE: 19.75 estimates/pixel\nMean measurement count for 2015-2016 CE: 61.04 estimates/pixel\nMean measurement count for 2016-2017 CE: 108.80 estimates/pixel\nMean measurement count for 2017-2018 CE: 133.65 estimates/pixel\nMean measurement count for 2018-2019 CE: 138.09 estimates/pixel\nMean measurement count for 2019-2020 CE: 136.80 estimates/pixel\nMean measurement count for 2020-2021 CE: 122.56 estimates/pixel\nMean measurement count for 2021-2022 CE: 78.79 estimates/pixel\nMean measurement count for 2022-2023 CE: 51.34 estimates/pixel\n\ndomain = 'antarctic_ice_sheet'\nMean velocity error (2σ) for 2021-2022 CE: 14.09 m/yr\nMean velocity error (2σ) for 2022-2023 CE: 19.45 m/yr\nMean measurement count for 2021-2022 CE: 71.04 estimates/pixel\nMean measurement count for 2022-2023 CE: 45.61 estimates/pixel\n```\n\nThe threshold (i.e. the minimum requirement to be met to ensure that data are useful) for surface ice flow velocity uncertainty (expressed in terms of 2$\\sigma$) proposed by the GCOS is 100 m yr⁻¹, while the \"goal\" value (i.e. an ideal requirement above which further improvements are not necessary) would be 10 m yr⁻¹ per grid point [[2](https://library.wmo.int/records/item/58111-the-2022-gcos-ecvs-requirements-gcos-245)]. The overall arithmetic mean error over both space and time shows that the minimum threshold is reached for all years but the 2020-2021 period for the GrIS (which will be discussed later).\n\n(section-2-3)="} {"chunk_id": "satellite_satellite-greenland-ice-sheet-velocity_uncertainty_q02__d4a0908e8e71", "report_id": "satellite_satellite-greenland-ice-sheet-velocity_uncertainty_q02", "dataset_id": "satellite-greenland-ice-sheet-velocity", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty_q02", "aspect_base": "uncertainty", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications > Analysis and results > 2. Ice sheet flow velocities and their uncertainty estimates in space and time > 2.3 Spatio-temporal distribution of ice flow velocity uncertainty for the GrIS", "title": "Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications", "chunk_index": 12, "token_count": 794, "text_raw": "Let us now plot the spatial distribution of the errors (here thus visualized as 2 times the standard deviation with units in meters per year):\n\n🚨 **The files can be large! Since the data files to be plotted are of considerable size, this may therefore take a couple of minutes.**\n\nGreenland setup\nExtract and preprocess data\nPlot Greenland\nColorbar with formatting\n\n*Figure 3. Magnitude of the average horizontal surface flow velocity error over Greenland over the defined time period.*\n\nHere, high standard deviation areas suggest lower precision and higher uncertainty. From the images above, it can immediately be seen why the horizontal velocity error for the 2020-2021 period is relatively high when compared to the earlier years: the data for this particular year have not been masked. High errors prevail over the ocean, which is of course unrealistic, and data even extend further into the Canadian Arctic. We can solve this by masking the data (i.e. by excluding the data outside of the Greenland continent land borders) and recalculating the error:\n\n```text\nThe masked arithmetic mean precision error (2σ) for the 2020-2021 period is 6.51 m yr⁻¹.\n```\n\nNevertheless, from the figure above, the precision errors seem to be especially higher near the margins of the ice sheet. In these regions, the threshold value of 100 m yr⁻¹ is locally not reached. Let us make this plot a bit clearer by grouping the errors (here averaged over all time periods) with respect to the threshold values proposed by the GCOS:\n\nCreate subplots with Polar Stereographic projection\nPlot the data\nSet extent and plot features\nAdd colorbar\nApply the function to the velocity standard deviation data\nDefine bounds and colormap\n\n*Figure 4. Classification of the average horizontal surface flow velocity error over Greenland over the defined time period according to proposed GCOS thresholds.*\n\nWhen plotted, the error (in the figure above plotted as averaged over all time periods) of the ice velocity product exhibits a significant spatial variability, with higher values generally observed near the ice sheet margins. Here, the terrain complexity, rapidly changing surface conditions (such as high ice melt or any other changes in the terrain that can alter the scattering characteristics of the radar pulse), as well as the relatively high speed of the outlet glaciers, have a direct impact on the absolute error value [[6](https://doi.org/10.1016/j.rse.2017.08.038)]. Conversely, towards the interior of the ice sheet, where there is little or no melt and ice flow speeds are relatively low and stable, the standard deviation is relatively low as well.\n\nWe can check in what percentage of the pixels the GCOS requirements have been reached or not reached:\n\n```text\nThe percentage of data points with a velocity precision error value (2σ) less than 10 m yr⁻¹ is 71.03%.\nThe percentage of data points with a velocity precision error value (2σ) more than 100 m yr⁻¹ is 1.54%.\n```\n\nThe majority of data therefore meet the uncertainty threshold proposed by the GCOS in the case of the GrIS, which adds credibility to the C3S ice sheet velocity product.\n\n(section-2-4)=", "text_with_prefix": "EQC Quality Assessment: \"Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications\"\nDataset: satellite-greenland-ice-sheet-velocity [CDS]\nAspect: uncertainty_q02 | Category: Satellite_ECVs\nSection: Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications > Analysis and results > 2. Ice sheet flow velocities and their uncertainty estimates in space and time > 2.3 Spatio-temporal distribution of ice flow velocity uncertainty for the GrIS\n---\nLet us now plot the spatial distribution of the errors (here thus visualized as 2 times the standard deviation with units in meters per year):\n\n🚨 **The files can be large! Since the data files to be plotted are of considerable size, this may therefore take a couple of minutes.**\n\nGreenland setup\nExtract and preprocess data\nPlot Greenland\nColorbar with formatting\n\n*Figure 3. Magnitude of the average horizontal surface flow velocity error over Greenland over the defined time period.*\n\nHere, high standard deviation areas suggest lower precision and higher uncertainty. From the images above, it can immediately be seen why the horizontal velocity error for the 2020-2021 period is relatively high when compared to the earlier years: the data for this particular year have not been masked. High errors prevail over the ocean, which is of course unrealistic, and data even extend further into the Canadian Arctic. We can solve this by masking the data (i.e. by excluding the data outside of the Greenland continent land borders) and recalculating the error:\n\n```text\nThe masked arithmetic mean precision error (2σ) for the 2020-2021 period is 6.51 m yr⁻¹.\n```\n\nNevertheless, from the figure above, the precision errors seem to be especially higher near the margins of the ice sheet. In these regions, the threshold value of 100 m yr⁻¹ is locally not reached. Let us make this plot a bit clearer by grouping the errors (here averaged over all time periods) with respect to the threshold values proposed by the GCOS:\n\nCreate subplots with Polar Stereographic projection\nPlot the data\nSet extent and plot features\nAdd colorbar\nApply the function to the velocity standard deviation data\nDefine bounds and colormap\n\n*Figure 4. Classification of the average horizontal surface flow velocity error over Greenland over the defined time period according to proposed GCOS thresholds.*\n\nWhen plotted, the error (in the figure above plotted as averaged over all time periods) of the ice velocity product exhibits a significant spatial variability, with higher values generally observed near the ice sheet margins. Here, the terrain complexity, rapidly changing surface conditions (such as high ice melt or any other changes in the terrain that can alter the scattering characteristics of the radar pulse), as well as the relatively high speed of the outlet glaciers, have a direct impact on the absolute error value [[6](https://doi.org/10.1016/j.rse.2017.08.038)]. Conversely, towards the interior of the ice sheet, where there is little or no melt and ice flow speeds are relatively low and stable, the standard deviation is relatively low as well.\n\nWe can check in what percentage of the pixels the GCOS requirements have been reached or not reached:\n\n```text\nThe percentage of data points with a velocity precision error value (2σ) less than 10 m yr⁻¹ is 71.03%.\nThe percentage of data points with a velocity precision error value (2σ) more than 100 m yr⁻¹ is 1.54%.\n```\n\nThe majority of data therefore meet the uncertainty threshold proposed by the GCOS in the case of the GrIS, which adds credibility to the C3S ice sheet velocity product.\n\n(section-2-4)="} {"chunk_id": "satellite_satellite-greenland-ice-sheet-velocity_uncertainty_q02__2aadf5868f22", "report_id": "satellite_satellite-greenland-ice-sheet-velocity_uncertainty_q02", "dataset_id": "satellite-greenland-ice-sheet-velocity", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty_q02", "aspect_base": "uncertainty", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications > Analysis and results > 2. Ice sheet flow velocities and their uncertainty estimates in space and time > 2.4 Spatio-temporal distribution of ice flow velocity uncertainty for the AIS", "title": "Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications", "chunk_index": 13, "token_count": 470, "text_raw": "Now let us repeat the same procedure for Antarctica. We first plot the average velocity errors:\n\nAntarctica setup\nExtract and preprocess data\nPlot Antarctica\nColorbar with formatting\n\n*Figure 5. Magnitude of the average horizontal surface flow velocity error over Antarctica over the defined time period.*\n\nIn the above figure, we note a very similar pattern as with the GrIS: the flow velocity errors are relatively high near the margins, but are considered small over the interior low-flow zones of the ice sheet. Let us classify them according to the proposed GCOS thresholds:\n\nCreate subplots with Polar Stereographic projection\nPlot the data\nSet extent and plot features\nAdd colorbar\nApply the function to the velocity standard deviation data\nDefine bounds and colormap\n\n*Figure 6. Classification of the average horizontal surface flow velocity error over Antarctica over the defined time period according to proposed GCOS thresholds.*\n\nThe spatio-temporal pattern of ice sheet velocity errors is therefore similar to that of the GrIS. Let us check in what percentage of the pixels the GCOS requirements have been reached or not reached:\n\n```text\nThe percentage of data points with a velocity precision error value (2σ) less than 10 m yr⁻¹ is 55.56%.\nThe percentage of data points with a velocity precision error value (2σ) more than 100 m yr⁻¹ is 1.59%.\n```\n\nAs with the GrIS, the majority of the pixels exhibit uncertainty values that fall within the thresholds proposed by GCOS. This ensures that the C3S AIS velocity product holds credible data. Next, let us deal with the second measure of error and uncertainty: the valid pixel count.\n\n(section-3)=", "text_with_prefix": "EQC Quality Assessment: \"Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications\"\nDataset: satellite-greenland-ice-sheet-velocity [CDS]\nAspect: uncertainty_q02 | Category: Satellite_ECVs\nSection: Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications > Analysis and results > 2. Ice sheet flow velocities and their uncertainty estimates in space and time > 2.4 Spatio-temporal distribution of ice flow velocity uncertainty for the AIS\n---\nNow let us repeat the same procedure for Antarctica. We first plot the average velocity errors:\n\nAntarctica setup\nExtract and preprocess data\nPlot Antarctica\nColorbar with formatting\n\n*Figure 5. Magnitude of the average horizontal surface flow velocity error over Antarctica over the defined time period.*\n\nIn the above figure, we note a very similar pattern as with the GrIS: the flow velocity errors are relatively high near the margins, but are considered small over the interior low-flow zones of the ice sheet. Let us classify them according to the proposed GCOS thresholds:\n\nCreate subplots with Polar Stereographic projection\nPlot the data\nSet extent and plot features\nAdd colorbar\nApply the function to the velocity standard deviation data\nDefine bounds and colormap\n\n*Figure 6. Classification of the average horizontal surface flow velocity error over Antarctica over the defined time period according to proposed GCOS thresholds.*\n\nThe spatio-temporal pattern of ice sheet velocity errors is therefore similar to that of the GrIS. Let us check in what percentage of the pixels the GCOS requirements have been reached or not reached:\n\n```text\nThe percentage of data points with a velocity precision error value (2σ) less than 10 m yr⁻¹ is 55.56%.\nThe percentage of data points with a velocity precision error value (2σ) more than 100 m yr⁻¹ is 1.59%.\n```\n\nAs with the GrIS, the majority of the pixels exhibit uncertainty values that fall within the thresholds proposed by GCOS. This ensures that the C3S AIS velocity product holds credible data. Next, let us deal with the second measure of error and uncertainty: the valid pixel count.\n\n(section-3)="} {"chunk_id": "satellite_satellite-greenland-ice-sheet-velocity_uncertainty_q02__3bc833502eb8", "report_id": "satellite_satellite-greenland-ice-sheet-velocity_uncertainty_q02", "dataset_id": "satellite-greenland-ice-sheet-velocity", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty_q02", "aspect_base": "uncertainty", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications > Analysis and results > 3. Valid measurement counts for the surface ice flow velocity > 3.1 Spatio-temporal distribution of ice flow velocity valid pixel counts for the GrIS", "title": "Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications", "chunk_index": 14, "token_count": 410, "text_raw": "In the figure below, we plot the averaged pixel-by-pixel valid measurement counts of the horizontal surface velocity for the GrIS:\n\nGreenland setup\nExtract and preprocess data\nPlot Greenland\nColorbar with formatting\n\n*Figure 7. Average annual valid pixel count of the horizontal surface flow velocity for Greenland over the defined time period.*\n\nIn the figure above, the straight edges of different valid pixel counts and the apparent paths of data acquisition are an indicator that the differences in valid pixel count are partly due to the satellite acquisition. However, higher valid pixel counts are clearly observed along the margins of the ice sheet and major outlet glaciers, consistent with areas of fast dynamics that require a more detailed monitoring.\n\nThis reflects the fact that generating high-density velocity estimates using SAR-based offset tracking is generally more challenging in the interior of the ice sheet. This difficulty mainly arises because of the relatively slow movement of ice in the interior compared to the faster-flowing edges, which results in smaller displacements between image pairs. It may also be because of less traceable surface features in the interior as compared to the margins, where for example, crevasses are more frequent. This makes it harder to detect and measure them accurately. A drawback to the data is the fact that there is no information given related to the time of the year during which valid measurements were acquired and that no data at a finer temporal resolution than 1 year are present.\n\n(section-3-2)=", "text_with_prefix": "EQC Quality Assessment: \"Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications\"\nDataset: satellite-greenland-ice-sheet-velocity [CDS]\nAspect: uncertainty_q02 | Category: Satellite_ECVs\nSection: Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications > Analysis and results > 3. Valid measurement counts for the surface ice flow velocity > 3.1 Spatio-temporal distribution of ice flow velocity valid pixel counts for the GrIS\n---\nIn the figure below, we plot the averaged pixel-by-pixel valid measurement counts of the horizontal surface velocity for the GrIS:\n\nGreenland setup\nExtract and preprocess data\nPlot Greenland\nColorbar with formatting\n\n*Figure 7. Average annual valid pixel count of the horizontal surface flow velocity for Greenland over the defined time period.*\n\nIn the figure above, the straight edges of different valid pixel counts and the apparent paths of data acquisition are an indicator that the differences in valid pixel count are partly due to the satellite acquisition. However, higher valid pixel counts are clearly observed along the margins of the ice sheet and major outlet glaciers, consistent with areas of fast dynamics that require a more detailed monitoring.\n\nThis reflects the fact that generating high-density velocity estimates using SAR-based offset tracking is generally more challenging in the interior of the ice sheet. This difficulty mainly arises because of the relatively slow movement of ice in the interior compared to the faster-flowing edges, which results in smaller displacements between image pairs. It may also be because of less traceable surface features in the interior as compared to the margins, where for example, crevasses are more frequent. This makes it harder to detect and measure them accurately. A drawback to the data is the fact that there is no information given related to the time of the year during which valid measurements were acquired and that no data at a finer temporal resolution than 1 year are present.\n\n(section-3-2)="} {"chunk_id": "satellite_satellite-greenland-ice-sheet-velocity_uncertainty_q02__fd202d088d5a", "report_id": "satellite_satellite-greenland-ice-sheet-velocity_uncertainty_q02", "dataset_id": "satellite-greenland-ice-sheet-velocity", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty_q02", "aspect_base": "uncertainty", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications > Analysis and results > 3. Valid measurement counts for the surface ice flow velocity > 3.2 Spatio-temporal distribution of ice flow velocity valid pixel counts for the AIS", "title": "Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications", "chunk_index": 15, "token_count": 316, "text_raw": "Let us plot the same for the AIS:\n\nAntarctica setup\nExtract and preprocess data\nPlot Antarctica\nColorbar with formatting\n\n*Figure 8. Average annual valid pixel count of the horizontal surface flow velocity for Antarctica over the defined time period.*\n\nAs with the GrIS, a higher count in the figure above typically indicates a more reliable estimate due to the inclusion of more observations into the velocity estimate. Paths of data acquisition are, however, also clearly visible in the pattern. This suggests that valid pixel counts are also dependent on the satellite acquisition and do not always represent glaciological signals. As with Greenland, a drawback to the data is the fact that there is no information given related to the time of the year during which valid measurements were acquired. Again, no thresholds are proposed by GCOS with respect to valid pixel counts, so no comparison is possible. However, the higher the valid pixel count, the more reliable a velocity estimate can be considered.\n\n(section-4)=", "text_with_prefix": "EQC Quality Assessment: \"Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications\"\nDataset: satellite-greenland-ice-sheet-velocity [CDS]\nAspect: uncertainty_q02 | Category: Satellite_ECVs\nSection: Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications > Analysis and results > 3. Valid measurement counts for the surface ice flow velocity > 3.2 Spatio-temporal distribution of ice flow velocity valid pixel counts for the AIS\n---\nLet us plot the same for the AIS:\n\nAntarctica setup\nExtract and preprocess data\nPlot Antarctica\nColorbar with formatting\n\n*Figure 8. Average annual valid pixel count of the horizontal surface flow velocity for Antarctica over the defined time period.*\n\nAs with the GrIS, a higher count in the figure above typically indicates a more reliable estimate due to the inclusion of more observations into the velocity estimate. Paths of data acquisition are, however, also clearly visible in the pattern. This suggests that valid pixel counts are also dependent on the satellite acquisition and do not always represent glaciological signals. As with Greenland, a drawback to the data is the fact that there is no information given related to the time of the year during which valid measurements were acquired. Again, no thresholds are proposed by GCOS with respect to valid pixel counts, so no comparison is possible. However, the higher the valid pixel count, the more reliable a velocity estimate can be considered.\n\n(section-4)="} {"chunk_id": "satellite_satellite-greenland-ice-sheet-velocity_uncertainty_q02__f2b1ddbc1852", "report_id": "satellite_satellite-greenland-ice-sheet-velocity_uncertainty_q02", "dataset_id": "satellite-greenland-ice-sheet-velocity", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty_q02", "aspect_base": "uncertainty", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications > Analysis and results > 4. Implications for use of horizontal ice flow velocity data in an ice sheet modelling framework > 4.1 Ice sheet model initialisation and tuning of the surface flow velocity field", "title": "Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications", "chunk_index": 16, "token_count": 920, "text_raw": "In the following section, we integrate all information derived above (with respect to spatio-temporal coverage and uncertainty) to assess the suitability of the ice sheet surface ice flow velocity dataset to use the velocity product in an ice sheet modelling framework.\n\nAt the heart of an ice sheet model is the solution of the time-dependent continuity equation for ice thickness $H$:\n\n$\n\\dfrac{\\partial H}{\\partial t} = -\\nabla \\cdot (\\vec{\\overline{v}}H) + M - S\n$\n\nwhere $H$ is the ice thickness, $\\vec{\\overline{v}}$ is the vertically averaged horizontal velocity vector, $M$ is the surface mass balance, and $S$ is the basal mass balance. A calibration procedure involves adjusting, for example, the basal sliding coefficient in Weertman’s basal sliding law in areas with basal temperatures at the pressure melting point and/or to adjust the rate factor in Glen’s flow law for interior regions frozen to bedrock (which impact $\\vec{\\overline{v}}$). This allows the modelled ice thickness or surface velocity to match the observed one (e.g. [[4](https://doi.org/10.5194/gmd-12-2481-2019), [5](https://doi.org/10.3390/rs11121407)]). This approach is valid as the present-day velocities of the ice sheets are supposedly close to the balance velocities.\n\nIn that regard, comparing modelled surface velocity fields to the observed ones provides a tool for the calibration/validation of an ice sheet model. The C3S ice sheet velocity product is particularly well-suited for this purpose. Even though large spatial gaps remain for the AIS product, the most prominent and dynamic regions across the margins are covered. Moreover, the GrIS product practically entirely covers the Greenland continent. Furthermore, the C3S velocity product includes quantitative pixel-by-pixel error estimates, for which the uncertainty characterization is represented by two variables: the standard deviation (i.e. a precision error) and the valid pixel count (the number of observations used to estimate the eventual average pixel value). The inclusion of both variables enhances the product quality and can serve as an important factor to consider when using the velocity product in an ice sheet modelling framework. When tuning an ice sheet model, it can, for example, be beneficial to prioritize high-quality velocity data (i.e. with a low absolute or relative error and a high valid measurement count). This ensures that the model solution is based on the most reliable observations and higher quality data have a greater influence on the eventual solution. One strategy could be to weigh the data using weights based on both the velocity standard deviation and the valid measurement count. This approach involves combining the two variables to form a single weight $w_i$ for each data point $i$:\n\n$w_i = \\alpha \\cdot (1-\\varepsilon_{V_{s},i}^{'}) + \\beta \\cdot N_{VP,i}^{'}\n$\n\nwhere:\n- $\\varepsilon_{V_{s},i}^{'}$ are the normalized (relative) errors, \n- $N_{VP,i}^{'}$ are the normalized valid measurement counts, \n- weights $\\alpha$ and $\\beta$ for which $\\alpha + \\beta = 1$.\n\nWhen using these weights in model validation, each data point's contribution to the model can be weighed accordingly. For instance, the weighted sum of squared residuals can be compared:\n\n$\n\\sum_i w_i (y_i - f(x_i))^2\n$\n\nwhere:\n- $y_i$ is the observed data (i.e. here the velocities from the dataset)\n- $f(x_i)$ is the model prediction for the i$^{th}$ data point\n- $w_i$ is the weight for the i$^{th}$ data point\n\n(section-4-2)=", "text_with_prefix": "EQC Quality Assessment: \"Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications\"\nDataset: satellite-greenland-ice-sheet-velocity [CDS]\nAspect: uncertainty_q02 | Category: Satellite_ECVs\nSection: Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications > Analysis and results > 4. Implications for use of horizontal ice flow velocity data in an ice sheet modelling framework > 4.1 Ice sheet model initialisation and tuning of the surface flow velocity field\n---\nIn the following section, we integrate all information derived above (with respect to spatio-temporal coverage and uncertainty) to assess the suitability of the ice sheet surface ice flow velocity dataset to use the velocity product in an ice sheet modelling framework.\n\nAt the heart of an ice sheet model is the solution of the time-dependent continuity equation for ice thickness $H$:\n\n$\n\\dfrac{\\partial H}{\\partial t} = -\\nabla \\cdot (\\vec{\\overline{v}}H) + M - S\n$\n\nwhere $H$ is the ice thickness, $\\vec{\\overline{v}}$ is the vertically averaged horizontal velocity vector, $M$ is the surface mass balance, and $S$ is the basal mass balance. A calibration procedure involves adjusting, for example, the basal sliding coefficient in Weertman’s basal sliding law in areas with basal temperatures at the pressure melting point and/or to adjust the rate factor in Glen’s flow law for interior regions frozen to bedrock (which impact $\\vec{\\overline{v}}$). This allows the modelled ice thickness or surface velocity to match the observed one (e.g. [[4](https://doi.org/10.5194/gmd-12-2481-2019), [5](https://doi.org/10.3390/rs11121407)]). This approach is valid as the present-day velocities of the ice sheets are supposedly close to the balance velocities.\n\nIn that regard, comparing modelled surface velocity fields to the observed ones provides a tool for the calibration/validation of an ice sheet model. The C3S ice sheet velocity product is particularly well-suited for this purpose. Even though large spatial gaps remain for the AIS product, the most prominent and dynamic regions across the margins are covered. Moreover, the GrIS product practically entirely covers the Greenland continent. Furthermore, the C3S velocity product includes quantitative pixel-by-pixel error estimates, for which the uncertainty characterization is represented by two variables: the standard deviation (i.e. a precision error) and the valid pixel count (the number of observations used to estimate the eventual average pixel value). The inclusion of both variables enhances the product quality and can serve as an important factor to consider when using the velocity product in an ice sheet modelling framework. When tuning an ice sheet model, it can, for example, be beneficial to prioritize high-quality velocity data (i.e. with a low absolute or relative error and a high valid measurement count). This ensures that the model solution is based on the most reliable observations and higher quality data have a greater influence on the eventual solution. One strategy could be to weigh the data using weights based on both the velocity standard deviation and the valid measurement count. This approach involves combining the two variables to form a single weight $w_i$ for each data point $i$:\n\n$w_i = \\alpha \\cdot (1-\\varepsilon_{V_{s},i}^{'}) + \\beta \\cdot N_{VP,i}^{'}\n$\n\nwhere:\n- $\\varepsilon_{V_{s},i}^{'}$ are the normalized (relative) errors, \n- $N_{VP,i}^{'}$ are the normalized valid measurement counts, \n- weights $\\alpha$ and $\\beta$ for which $\\alpha + \\beta = 1$.\n\nWhen using these weights in model validation, each data point's contribution to the model can be weighed accordingly. For instance, the weighted sum of squared residuals can be compared:\n\n$\n\\sum_i w_i (y_i - f(x_i))^2\n$\n\nwhere:\n- $y_i$ is the observed data (i.e. here the velocities from the dataset)\n- $f(x_i)$ is the model prediction for the i$^{th}$ data point\n- $w_i$ is the weight for the i$^{th}$ data point\n\n(section-4-2)="} {"chunk_id": "satellite_satellite-greenland-ice-sheet-velocity_uncertainty_q02__f865438cdaeb", "report_id": "satellite_satellite-greenland-ice-sheet-velocity_uncertainty_q02", "dataset_id": "satellite-greenland-ice-sheet-velocity", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty_q02", "aspect_base": "uncertainty", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications > Analysis and results > 4. Implications for use of horizontal ice flow velocity data in an ice sheet modelling framework > 4.2 Presence of arteficial 'striping'", "title": "Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications", "chunk_index": 17, "token_count": 460, "text_raw": "Apart from quantitive error estimates, another phenomenom deserves special attention. Namely, the presence of 'striping' (i.e. banding or noise in the velocity data that become especially visible in low-flow zones) in SAR offset tracking-based ice velocity estimates can have several significant impacts when using the C3S product for the validation of ice sheet models, especially in interior low-flow zones. In short, these stripes originate from certain specific atmospheric disturbances that interfer with the SAR signal, producing spatially variable unwanted displacements (e.g. [[12](https://doi.org/10.1109/IGARSS.2006.956)]).\n\nThe presence of striping reduces the overall quality of the velocity data in low-flow zones and may lead to the misinterpretation of physical processes driving ice flow in these areas. This unrealistic pattern, especially visible in the interior of the ice sheet, may therefore limit the use of the product for slowly moving ice and is hence important to consider when using the data. Mitigating these artefacts involves, for example, noise reduction methods such as spatial averaging or filtering, or the integration of velocities from SAR interferometry, which are generally very accurate for slow-moving ice [[1](https://doi.org/10.3390/rs70709371), [7](https://doi.org/10.3390/rs9101062), [9](https://doi.org/10.1029/2019GL083826)]. Another possible solution is replacing the distorted velocities in these areas by balance velocities or to only use velocities above a certain threshold (i.e. >5 m yr⁻¹) for the relevant procedure.\n\n(section-5)=", "text_with_prefix": "EQC Quality Assessment: \"Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications\"\nDataset: satellite-greenland-ice-sheet-velocity [CDS]\nAspect: uncertainty_q02 | Category: Satellite_ECVs\nSection: Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications > Analysis and results > 4. Implications for use of horizontal ice flow velocity data in an ice sheet modelling framework > 4.2 Presence of arteficial 'striping'\n---\nApart from quantitive error estimates, another phenomenom deserves special attention. Namely, the presence of 'striping' (i.e. banding or noise in the velocity data that become especially visible in low-flow zones) in SAR offset tracking-based ice velocity estimates can have several significant impacts when using the C3S product for the validation of ice sheet models, especially in interior low-flow zones. In short, these stripes originate from certain specific atmospheric disturbances that interfer with the SAR signal, producing spatially variable unwanted displacements (e.g. [[12](https://doi.org/10.1109/IGARSS.2006.956)]).\n\nThe presence of striping reduces the overall quality of the velocity data in low-flow zones and may lead to the misinterpretation of physical processes driving ice flow in these areas. This unrealistic pattern, especially visible in the interior of the ice sheet, may therefore limit the use of the product for slowly moving ice and is hence important to consider when using the data. Mitigating these artefacts involves, for example, noise reduction methods such as spatial averaging or filtering, or the integration of velocities from SAR interferometry, which are generally very accurate for slow-moving ice [[1](https://doi.org/10.3390/rs70709371), [7](https://doi.org/10.3390/rs9101062), [9](https://doi.org/10.1029/2019GL083826)]. Another possible solution is replacing the distorted velocities in these areas by balance velocities or to only use velocities above a certain threshold (i.e. >5 m yr⁻¹) for the relevant procedure.\n\n(section-5)="} {"chunk_id": "satellite_satellite-greenland-ice-sheet-velocity_uncertainty_q02__61480651e1b5", "report_id": "satellite_satellite-greenland-ice-sheet-velocity_uncertainty_q02", "dataset_id": "satellite-greenland-ice-sheet-velocity", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty_q02", "aspect_base": "uncertainty", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications > Analysis and results > 5. Short summary and take-home messages", "title": "Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications", "chunk_index": 18, "token_count": 594, "text_raw": "Satellite-derived ice velocity data provide valuable insight into the flow dynamics of the ice sheets [[11](https://doi.org/10.1016/j.rse.2025.115092)], but users should be aware of specific limitations in the SAR offset tracking-based velocity products. It should be noted that data quality varies across the ice sheet and is best assessed using the provided standard deviation maps (indicating precision errors) and valid measurement count maps (reflecting the number of successful matches used in annual velocity averaging). Challenging regions include the fast-flowing margins, where high absolute errors occur, as well as the interior low-velocity zones, where lower valid pixel counts and artefacts such as striping may occur. Nevertheless, the majority of pixels of the C3S ice sheet velocity dataset meets the GCOS uncertainty target. The product also meets the spatial (250 m) and temporal (annual) resolution thresholds set by GCOS, which further enhances the credibility of the C3S ice sheet velocity product (cfr. the \"Maturity Matrix\" [[8](https://doi.org/10.1175/BAMS-D-21-0109.1)]). However, significant data gaps remain over the interior of the AIS.\n\nThe dataset thus offers a strong spatial/temporal coverage and uncertainty pattern for use in ice sheet model calibration or validation (e.g. [[4](https://doi.org/10.5194/gmd-12-2481-2019), [5](https://doi.org/10.3390/rs11121407), [10](https://doi.org/10.5194/tc-6-1561-2012)]). Validation efforts show a good agreement with independent data, further adding credibility to the C3S dataset. The user can, if desired, prioritize high-quality data (i.e. pixels with low absolute/relative errors and high valid pixel counts) for this purpose. Errors are, however, provided as absolute values, while for flow fields, the relative error is generally considered to be more useful. Other potential limitations include the absence of information on acquisition timing, no flagging of spatially filled pixels or missing data, the lack of an ice mask to exclude non-ice-covered cells with velocity values, and the relatively short temporal span of the dataset. To address these issues, users can, for example, apply custom ice masks and complement the product with external datasets for a more robust velocity estimation.", "text_with_prefix": "EQC Quality Assessment: \"Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications\"\nDataset: satellite-greenland-ice-sheet-velocity [CDS]\nAspect: uncertainty_q02 | Category: Satellite_ECVs\nSection: Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications > Analysis and results > 5. Short summary and take-home messages\n---\nSatellite-derived ice velocity data provide valuable insight into the flow dynamics of the ice sheets [[11](https://doi.org/10.1016/j.rse.2025.115092)], but users should be aware of specific limitations in the SAR offset tracking-based velocity products. It should be noted that data quality varies across the ice sheet and is best assessed using the provided standard deviation maps (indicating precision errors) and valid measurement count maps (reflecting the number of successful matches used in annual velocity averaging). Challenging regions include the fast-flowing margins, where high absolute errors occur, as well as the interior low-velocity zones, where lower valid pixel counts and artefacts such as striping may occur. Nevertheless, the majority of pixels of the C3S ice sheet velocity dataset meets the GCOS uncertainty target. The product also meets the spatial (250 m) and temporal (annual) resolution thresholds set by GCOS, which further enhances the credibility of the C3S ice sheet velocity product (cfr. the \"Maturity Matrix\" [[8](https://doi.org/10.1175/BAMS-D-21-0109.1)]). However, significant data gaps remain over the interior of the AIS.\n\nThe dataset thus offers a strong spatial/temporal coverage and uncertainty pattern for use in ice sheet model calibration or validation (e.g. [[4](https://doi.org/10.5194/gmd-12-2481-2019), [5](https://doi.org/10.3390/rs11121407), [10](https://doi.org/10.5194/tc-6-1561-2012)]). Validation efforts show a good agreement with independent data, further adding credibility to the C3S dataset. The user can, if desired, prioritize high-quality data (i.e. pixels with low absolute/relative errors and high valid pixel counts) for this purpose. Errors are, however, provided as absolute values, while for flow fields, the relative error is generally considered to be more useful. Other potential limitations include the absence of information on acquisition timing, no flagging of spatially filled pixels or missing data, the lack of an ice mask to exclude non-ice-covered cells with velocity values, and the relatively short temporal span of the dataset. To address these issues, users can, for example, apply custom ice masks and complement the product with external datasets for a more robust velocity estimation."} {"chunk_id": "satellite_satellite-greenland-ice-sheet-velocity_uncertainty_q02__750647b89921", "report_id": "satellite_satellite-greenland-ice-sheet-velocity_uncertainty_q02", "dataset_id": "satellite-greenland-ice-sheet-velocity", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty_q02", "aspect_base": "uncertainty", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications > ℹ️ If you want to know more > Key resources", "title": "Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications", "chunk_index": 19, "token_count": 360, "text_raw": "- \"[Ice sheet velocity for Antarctica and Greenland derived from satellite observations](https://cds.climate.copernicus.eu/datasets/satellite-greenland-ice-sheet-velocity?tab=overview)\" on the CDS\n- [Documentation on the CDS](https://cds.climate.copernicus.eu/datasets/satellite-greenland-ice-sheet-velocity?tab=documentation) and the [ECMWF Confluence Wiki](https://confluence.ecmwf.int/pages/viewpage.action?pageId=355348656) (Copernicus Knowledge Base).\n- [Copernicus climate change indicators: ice sheets](https://climate.copernicus.eu/climate-indicators/ice-sheets)\n- [A web article about remote sensing-derived ice velocity acquisition](https://www.esa.int/Applications/Observing_the_Earth/Copernicus/Sentinel-1/Sentinel-1_s_decade_of_essential_data_over_shifting_ice_sheets)\n- [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control` prepared by [B-Open](https://www.bopen.eu).", "text_with_prefix": "EQC Quality Assessment: \"Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications\"\nDataset: satellite-greenland-ice-sheet-velocity [CDS]\nAspect: uncertainty_q02 | Category: Satellite_ECVs\nSection: Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications > ℹ️ If you want to know more > Key resources\n---\n- \"[Ice sheet velocity for Antarctica and Greenland derived from satellite observations](https://cds.climate.copernicus.eu/datasets/satellite-greenland-ice-sheet-velocity?tab=overview)\" on the CDS\n- [Documentation on the CDS](https://cds.climate.copernicus.eu/datasets/satellite-greenland-ice-sheet-velocity?tab=documentation) and the [ECMWF Confluence Wiki](https://confluence.ecmwf.int/pages/viewpage.action?pageId=355348656) (Copernicus Knowledge Base).\n- [Copernicus climate change indicators: ice sheets](https://climate.copernicus.eu/climate-indicators/ice-sheets)\n- [A web article about remote sensing-derived ice velocity acquisition](https://www.esa.int/Applications/Observing_the_Earth/Copernicus/Sentinel-1/Sentinel-1_s_decade_of_essential_data_over_shifting_ice_sheets)\n- [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control` prepared by [B-Open](https://www.bopen.eu)."} {"chunk_id": "satellite_satellite-greenland-ice-sheet-velocity_uncertainty_q02__a727e6b8566b", "report_id": "satellite_satellite-greenland-ice-sheet-velocity_uncertainty_q02", "dataset_id": "satellite-greenland-ice-sheet-velocity", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty_q02", "aspect_base": "uncertainty", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications > ℹ️ If you want to know more > References", "title": "Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications", "chunk_index": 20, "token_count": 858, "text_raw": "- [[1](https://doi.org/10.3390/rs70709371)] Nagler, T., Rott, H., Hetzenecker, M., Wuite, J. and Potin, P. (2015). The Sentinel-1 Mission: New Opportunities for Ice Sheet Observations. Remote Sensing. 7(7):9371-9389. https://doi.org/10.3390/rs70709371\n\n- [[2](https://library.wmo.int/records/item/58111-the-2022-gcos-ecvs-requirements-gcos-245)] GCOS (Global Climate Observing System) (2022). The 2022 GCOS ECVs Requirements (GCOS-245). World Meteorological Organization: Geneva, Switzerland. doi: https://library.wmo.int/idurl/4/58111\n\n- [[3](https://nsidc.org/data/nsidc-0777/versions/1)] Howat, I., Chudley, T., and Noh, M. (2022). MEaSUREs Greenland Ice Velocity: Selected Glacier Site Single-Pair Velocity Maps from Optical Images, https://doi.org/10.5067/b28fm2qvvywy\n\n- [[4](https://doi.org/10.5194/gmd-12-2481-2019)] Le clec'h, S., Quiquet, A., Charbit, S., Dumas, C., Kageyama, M., and Ritz, C. (2019). A rapidly converging initialisation method to simulate the present-day Greenland ice sheet using the GRISLI ice sheet model (version 1.3), Geosci. Model Dev., 12, 2481–2499, https://doi.org/10.5194/gmd-12-2481-2019\n\n- [[5](https://doi.org/10.3390/rs11121407)] Mottram, R., Simonsen, S. B., Svendsen, S. H., Barletta, V. R., Sørensen, L. S., Nagler, T., Wuite, J., Groh, A., Horwath, M., Rosier, J., Solgaard, A., Hvidberg, C. S., and Forsberg, R. (2019). An integrated view of Greenland Ice Sheet mass changes based on models and satellite observations. Remote Sensing, 11(12), 1407. https://doi.org/10.3390/rs11121407\n\n- [[6](https://doi.org/10.1016/j.rse.2017.08.038)] Paul, F., Bolch, T., Briggs, K., Kääb, A., McMillan, M., McNabb, R., Nagler, T., Nuth, C., Rastner, P., Strozzi, T., and Wuite, J. (2017). Error sources and guidelines for quality assessment of glacier area, elevation change, and velocity products derived from satellite data in the Glaciers_cci project, Remote Sensing of Environment, 203, 256-275. https://doi.org/10.1016/j.rse.2017.08.038\n\n- [[7](https://doi.org/10.3390/rs9101062)] Lüttig, C., Neckel, N., and Humbert, A. (2017). A Combined Approach for Filtering Ice Surface Velocity Fields Derived from Remote Sensing Methods, Remote Sensing, 9(10). https://doi.org/10.3390/rs9101062", "text_with_prefix": "EQC Quality Assessment: \"Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications\"\nDataset: satellite-greenland-ice-sheet-velocity [CDS]\nAspect: uncertainty_q02 | Category: Satellite_ECVs\nSection: Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications > ℹ️ If you want to know more > References\n---\n- [[1](https://doi.org/10.3390/rs70709371)] Nagler, T., Rott, H., Hetzenecker, M., Wuite, J. and Potin, P. (2015). The Sentinel-1 Mission: New Opportunities for Ice Sheet Observations. Remote Sensing. 7(7):9371-9389. https://doi.org/10.3390/rs70709371\n\n- [[2](https://library.wmo.int/records/item/58111-the-2022-gcos-ecvs-requirements-gcos-245)] GCOS (Global Climate Observing System) (2022). The 2022 GCOS ECVs Requirements (GCOS-245). World Meteorological Organization: Geneva, Switzerland. doi: https://library.wmo.int/idurl/4/58111\n\n- [[3](https://nsidc.org/data/nsidc-0777/versions/1)] Howat, I., Chudley, T., and Noh, M. (2022). MEaSUREs Greenland Ice Velocity: Selected Glacier Site Single-Pair Velocity Maps from Optical Images, https://doi.org/10.5067/b28fm2qvvywy\n\n- [[4](https://doi.org/10.5194/gmd-12-2481-2019)] Le clec'h, S., Quiquet, A., Charbit, S., Dumas, C., Kageyama, M., and Ritz, C. (2019). A rapidly converging initialisation method to simulate the present-day Greenland ice sheet using the GRISLI ice sheet model (version 1.3), Geosci. Model Dev., 12, 2481–2499, https://doi.org/10.5194/gmd-12-2481-2019\n\n- [[5](https://doi.org/10.3390/rs11121407)] Mottram, R., Simonsen, S. B., Svendsen, S. H., Barletta, V. R., Sørensen, L. S., Nagler, T., Wuite, J., Groh, A., Horwath, M., Rosier, J., Solgaard, A., Hvidberg, C. S., and Forsberg, R. (2019). An integrated view of Greenland Ice Sheet mass changes based on models and satellite observations. Remote Sensing, 11(12), 1407. https://doi.org/10.3390/rs11121407\n\n- [[6](https://doi.org/10.1016/j.rse.2017.08.038)] Paul, F., Bolch, T., Briggs, K., Kääb, A., McMillan, M., McNabb, R., Nagler, T., Nuth, C., Rastner, P., Strozzi, T., and Wuite, J. (2017). Error sources and guidelines for quality assessment of glacier area, elevation change, and velocity products derived from satellite data in the Glaciers_cci project, Remote Sensing of Environment, 203, 256-275. https://doi.org/10.1016/j.rse.2017.08.038\n\n- [[7](https://doi.org/10.3390/rs9101062)] Lüttig, C., Neckel, N., and Humbert, A. (2017). A Combined Approach for Filtering Ice Surface Velocity Fields Derived from Remote Sensing Methods, Remote Sensing, 9(10). https://doi.org/10.3390/rs9101062"} {"chunk_id": "satellite_satellite-greenland-ice-sheet-velocity_uncertainty_q02__43aea705671d", "report_id": "satellite_satellite-greenland-ice-sheet-velocity_uncertainty_q02", "dataset_id": "satellite-greenland-ice-sheet-velocity", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty_q02", "aspect_base": "uncertainty", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications > ℹ️ If you want to know more > References", "title": "Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications", "chunk_index": 21, "token_count": 909, "text_raw": "A Combined Approach for Filtering Ice Surface Velocity Fields Derived from Remote Sensing Methods, Remote Sensing, 9(10). https://doi.org/10.3390/rs9101062\n\n- [[8](https://doi.org/10.1175/BAMS-D-21-0109.1)] Yang, C. X., Cagnazzo, C., Artale, V., Nardelli, B. B., Buontempo, C., Busatto, J., Caporaso, L., Cesarini, C., Cionni, I., Coll, J., Crezee, B., Cristofanelli, P., de Toma, V., Essa, Y. H., Eyring, V., Fierli, F., Grant, L., Hassler, B., Hirschi, M., Huybrechts, P., Le Merle, E., Leonelli, F. E., Lin, X., Madonna, F., Mason, E., Massonnet, F., Marcos, M., Marullo, S., Muller, B., Obregon, A., Organelli, E., Palacz, A., Pascual, A., Pisano, A., Putero, D., Rana, A., Sanchez-Roman, A., Seneviratne, S. I., Serva, F., Storto, A., Thiery, W., Throne, P., Van Tricht, L., Verhaegen, Y., Volpe, G., and Santoleri, R. (2022). Independent Quality Assessment of Essential Climate Variables: Lessons Learned from the Copernicus Climate Change Service, B. Am. Meteorol. Soc., 103, E2032–E2049, https://doi.org/10.1175/Bams-D-21-0109.1\n\n- [[9](https://doi.org/10.1029/2019GL083826)] Mouginot, J., E. Rignot, and B. Scheuchl (2019). Continent-wide, interferometric SAR phase, mapping of Antarctic ice velocity, Geophys. Res. Lett., 46, 9710-9718, https://doi.org/10.1029/2019GL083826\n\n- [[10](https://doi.org/10.5194/tc-6-1561-2012)] Gillet-Chaulet, F., Gagliardini, O., Seddik, H., Nodet, M., Durand, G., Ritz, C., Zwinger, T., Greve, R., and Vaughan, D. G. (2012). Greenland ice sheet contribution to sea-level rise from a new-generation ice-sheet model, The Cryosphere, 6, 1561–1576, https://doi.org/10.5194/tc-6-1561-2012\n\n- [[11](https://doi.org/10.1016/j.rse.2025.115092)] Wuite, J., Nagler, T., Hetzenecker, M., and Rott, H. (2025). Ten years of polar ice velocity mapping using Copernicus Sentinel-1. Remote Sensing of Environment, 332. https://doi.org/10.1016/j.rse.2025.115092\n\n- [[12](https://doi.org/10.1109/IGARSS.2006.956)] Wegmuller, U., Werner, C., Strozzi, T., and Wiesmann, A. (2006). Ionospheric electron concentration effects on SAR and insar. In Proceedings of the 2006 IEEE International Symposium on Geoscience and Remote Sensing, Denver, CO, USA, 31 July–4 August 2006. https://doi.org/10.1109/IGARSS.2006.956", "text_with_prefix": "EQC Quality Assessment: \"Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications\"\nDataset: satellite-greenland-ice-sheet-velocity [CDS]\nAspect: uncertainty_q02 | Category: Satellite_ECVs\nSection: Ice sheet velocity data from satellite observations: temporal and spatial variability of velocity-related uncertainties for glaciological applications > ℹ️ If you want to know more > References\n---\nA Combined Approach for Filtering Ice Surface Velocity Fields Derived from Remote Sensing Methods, Remote Sensing, 9(10). https://doi.org/10.3390/rs9101062\n\n- [[8](https://doi.org/10.1175/BAMS-D-21-0109.1)] Yang, C. X., Cagnazzo, C., Artale, V., Nardelli, B. B., Buontempo, C., Busatto, J., Caporaso, L., Cesarini, C., Cionni, I., Coll, J., Crezee, B., Cristofanelli, P., de Toma, V., Essa, Y. H., Eyring, V., Fierli, F., Grant, L., Hassler, B., Hirschi, M., Huybrechts, P., Le Merle, E., Leonelli, F. E., Lin, X., Madonna, F., Mason, E., Massonnet, F., Marcos, M., Marullo, S., Muller, B., Obregon, A., Organelli, E., Palacz, A., Pascual, A., Pisano, A., Putero, D., Rana, A., Sanchez-Roman, A., Seneviratne, S. I., Serva, F., Storto, A., Thiery, W., Throne, P., Van Tricht, L., Verhaegen, Y., Volpe, G., and Santoleri, R. (2022). Independent Quality Assessment of Essential Climate Variables: Lessons Learned from the Copernicus Climate Change Service, B. Am. Meteorol. Soc., 103, E2032–E2049, https://doi.org/10.1175/Bams-D-21-0109.1\n\n- [[9](https://doi.org/10.1029/2019GL083826)] Mouginot, J., E. Rignot, and B. Scheuchl (2019). Continent-wide, interferometric SAR phase, mapping of Antarctic ice velocity, Geophys. Res. Lett., 46, 9710-9718, https://doi.org/10.1029/2019GL083826\n\n- [[10](https://doi.org/10.5194/tc-6-1561-2012)] Gillet-Chaulet, F., Gagliardini, O., Seddik, H., Nodet, M., Durand, G., Ritz, C., Zwinger, T., Greve, R., and Vaughan, D. G. (2012). Greenland ice sheet contribution to sea-level rise from a new-generation ice-sheet model, The Cryosphere, 6, 1561–1576, https://doi.org/10.5194/tc-6-1561-2012\n\n- [[11](https://doi.org/10.1016/j.rse.2025.115092)] Wuite, J., Nagler, T., Hetzenecker, M., and Rott, H. (2025). Ten years of polar ice velocity mapping using Copernicus Sentinel-1. Remote Sensing of Environment, 332. https://doi.org/10.1016/j.rse.2025.115092\n\n- [[12](https://doi.org/10.1109/IGARSS.2006.956)] Wegmuller, U., Werner, C., Strozzi, T., and Wiesmann, A. (2006). Ionospheric electron concentration effects on SAR and insar. In Proceedings of the 2006 IEEE International Symposium on Geoscience and Remote Sensing, Denver, CO, USA, 31 July–4 August 2006. https://doi.org/10.1109/IGARSS.2006.956"} {"chunk_id": "satellite_satellite-ice-sheet-elevation-change_consistency-assessment_q02__c225d5effd8a", "report_id": "satellite_satellite-ice-sheet-elevation-change_consistency-assessment_q02", "dataset_id": "satellite-ice-sheet-elevation-change", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency-assessment_q02", "aspect_base": "consistency-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Spatio-temporal data completeness and consistency of remote sensing-derived ice sheet surface elevation changes over Antarctica", "title": "Spatio-temporal data completeness and consistency of remote sensing-derived ice sheet surface elevation changes over Antarctica", "chunk_index": 0, "token_count": 124, "text_raw": "Production date: 31-05-2025\n\nDataset version: 4.0\n\nProduced by: Yoni Verhaegen and Philippe Huybrechts (Vrije Universiteit Brussel)", "text_with_prefix": "EQC Quality Assessment: \"Spatio-temporal data completeness and consistency of remote sensing-derived ice sheet surface elevation changes over Antarctica\"\nDataset: satellite-ice-sheet-elevation-change [CDS]\nAspect: consistency-assessment_q02 | Category: Satellite_ECVs\nSection: Spatio-temporal data completeness and consistency of remote sensing-derived ice sheet surface elevation changes over Antarctica\n---\nProduction date: 31-05-2025\n\nDataset version: 4.0\n\nProduced by: Yoni Verhaegen and Philippe Huybrechts (Vrije Universiteit Brussel)"} {"chunk_id": "satellite_satellite-ice-sheet-elevation-change_consistency-assessment_q02__43265957559c", "report_id": "satellite_satellite-ice-sheet-elevation-change_consistency-assessment_q02", "dataset_id": "satellite-ice-sheet-elevation-change", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency-assessment_q02", "aspect_base": "consistency-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Spatio-temporal data completeness and consistency of remote sensing-derived ice sheet surface elevation changes over Antarctica > Quality assessment question", "title": "Spatio-temporal data completeness and consistency of remote sensing-derived ice sheet surface elevation changes over Antarctica", "chunk_index": 1, "token_count": 472, "text_raw": "* **\"How does the spatial coverage of the ice sheet surface elevation change data, aggregated from distinct satellite sensors and missions over time, change and how does it affect estimates of (cumulative) ice sheet surface elevation changes?\"**\n\nThe C3S surface elevation change (SEC) product quantifies changes in the surface elevation of ice sheets, providing crucial insights into their response to climate change and changing ice sheet dynamics. Satellite remote sensing is a valueable and practical method for regularly monitoring such surface elevation changes over those large, remote ice sheet areas. SEC data on the Climate Data Store (CDS) are derived from satellite radar altimetry, where the time delay between a transmitted pulse from the radar and its surface echo is converted into distance, adjusted for the satellite's known elevation. Repeated measurements of the surface elevation from multiple satellite missions, combined with corrections and interpolation, therefore allows to create a consistent time series of gridded ice sheet surface elevation changes [[1](https://doi.org/10.1016/j.epsl.2018.05.015), [2](https://doi.org/10.5194/tc-13-427-2019)].\n\nWhile these techniques generally offer extensive spatial and temporal coverage, they also have limitations, such as inconsistencies due to data aggregation from different satellite sensors and missions, and difficulties with data processing and data acquisition over complex terrain. This notebook assesses the suitability of the CDS SEC dataset (version 4.0) as an indicator of ice sheet imbalance under climate change. It specifically evaluates the dataset's maturity and quality related to data completeness, focusing on its spatio-temporal coverage and consistency across time series derived from multiple satellite sources. In this notebook, we only focus on the Antarctic Ice Sheet (AIS).\n\n🚨 **Although this notebook deals with the Antarctic Ice Sheet only, surface elevation change data for the Greenland Ice Sheet are also available.**", "text_with_prefix": "EQC Quality Assessment: \"Spatio-temporal data completeness and consistency of remote sensing-derived ice sheet surface elevation changes over Antarctica\"\nDataset: satellite-ice-sheet-elevation-change [CDS]\nAspect: consistency-assessment_q02 | Category: Satellite_ECVs\nSection: Spatio-temporal data completeness and consistency of remote sensing-derived ice sheet surface elevation changes over Antarctica > Quality assessment question\n---\n* **\"How does the spatial coverage of the ice sheet surface elevation change data, aggregated from distinct satellite sensors and missions over time, change and how does it affect estimates of (cumulative) ice sheet surface elevation changes?\"**\n\nThe C3S surface elevation change (SEC) product quantifies changes in the surface elevation of ice sheets, providing crucial insights into their response to climate change and changing ice sheet dynamics. Satellite remote sensing is a valueable and practical method for regularly monitoring such surface elevation changes over those large, remote ice sheet areas. SEC data on the Climate Data Store (CDS) are derived from satellite radar altimetry, where the time delay between a transmitted pulse from the radar and its surface echo is converted into distance, adjusted for the satellite's known elevation. Repeated measurements of the surface elevation from multiple satellite missions, combined with corrections and interpolation, therefore allows to create a consistent time series of gridded ice sheet surface elevation changes [[1](https://doi.org/10.1016/j.epsl.2018.05.015), [2](https://doi.org/10.5194/tc-13-427-2019)].\n\nWhile these techniques generally offer extensive spatial and temporal coverage, they also have limitations, such as inconsistencies due to data aggregation from different satellite sensors and missions, and difficulties with data processing and data acquisition over complex terrain. This notebook assesses the suitability of the CDS SEC dataset (version 4.0) as an indicator of ice sheet imbalance under climate change. It specifically evaluates the dataset's maturity and quality related to data completeness, focusing on its spatio-temporal coverage and consistency across time series derived from multiple satellite sources. In this notebook, we only focus on the Antarctic Ice Sheet (AIS).\n\n🚨 **Although this notebook deals with the Antarctic Ice Sheet only, surface elevation change data for the Greenland Ice Sheet are also available.**"} {"chunk_id": "satellite_satellite-ice-sheet-elevation-change_consistency-assessment_q02__270359a5f92f", "report_id": "satellite_satellite-ice-sheet-elevation-change_consistency-assessment_q02", "dataset_id": "satellite-ice-sheet-elevation-change", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency-assessment_q02", "aspect_base": "consistency-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Spatio-temporal data completeness and consistency of remote sensing-derived ice sheet surface elevation changes over Antarctica > Quality assessment statements", "title": "Spatio-temporal data completeness and consistency of remote sensing-derived ice sheet surface elevation changes over Antarctica", "chunk_index": 2, "token_count": 423, "text_raw": "These are the key outcomes of this assessment\n\n- Surface elevation change (SEC) detection by radar altimetry is a valuable tool for assessing the impacts of climate change on ice sheets, but it has notable limitations that users should be aware of. The C3S SEC data have been merged from different satellite missions over time, each having their own (dis)advantages in terms of data acquisition limitations and spatio-temporal sampling coverage/frequency. For the C3S AIS SEC dataset in particular, the lack of post-processing to fill remaining data gaps results in a rather immature final product. Even though significant improvements with respect to data completeness have occured during the more recent years, persistent data gaps remain around the polar gap and over complex terrain around the ice sheet margins. Since some of the areas with the most prominent SEC are within the areas of least coverage, special attention is required. \n- Without additional processing, the C3S SEC data for the AIS are found to not be suitable for reliable statistical analysis, such as calculating means, variability, trends, due to the high proportion of missing values, noisier data and relatively higher error values when compared to the GrIS SEC dataset. Due to the lack of post-processing, some signals in the SEC time series may therefore reflect changing sampling conditions and/or processing limitations rather than actual SEC. In the context of long-term monitoring and climate change studies, the C3S AIS SEC dataset mostly lacks the robustness and reliability required to accurately track surface elevation changes and estimate cumulative mass changes. Users of the C3S AIS SEC data product are therefore most likely required to implement additional processing to derive glaciologically interpretable surface elevation changes.\n```", "text_with_prefix": "EQC Quality Assessment: \"Spatio-temporal data completeness and consistency of remote sensing-derived ice sheet surface elevation changes over Antarctica\"\nDataset: satellite-ice-sheet-elevation-change [CDS]\nAspect: consistency-assessment_q02 | Category: Satellite_ECVs\nSection: Spatio-temporal data completeness and consistency of remote sensing-derived ice sheet surface elevation changes over Antarctica > Quality assessment statements\n---\nThese are the key outcomes of this assessment\n\n- Surface elevation change (SEC) detection by radar altimetry is a valuable tool for assessing the impacts of climate change on ice sheets, but it has notable limitations that users should be aware of. The C3S SEC data have been merged from different satellite missions over time, each having their own (dis)advantages in terms of data acquisition limitations and spatio-temporal sampling coverage/frequency. For the C3S AIS SEC dataset in particular, the lack of post-processing to fill remaining data gaps results in a rather immature final product. Even though significant improvements with respect to data completeness have occured during the more recent years, persistent data gaps remain around the polar gap and over complex terrain around the ice sheet margins. Since some of the areas with the most prominent SEC are within the areas of least coverage, special attention is required. \n- Without additional processing, the C3S SEC data for the AIS are found to not be suitable for reliable statistical analysis, such as calculating means, variability, trends, due to the high proportion of missing values, noisier data and relatively higher error values when compared to the GrIS SEC dataset. Due to the lack of post-processing, some signals in the SEC time series may therefore reflect changing sampling conditions and/or processing limitations rather than actual SEC. In the context of long-term monitoring and climate change studies, the C3S AIS SEC dataset mostly lacks the robustness and reliability required to accurately track surface elevation changes and estimate cumulative mass changes. Users of the C3S AIS SEC data product are therefore most likely required to implement additional processing to derive glaciologically interpretable surface elevation changes.\n```"} {"chunk_id": "satellite_satellite-ice-sheet-elevation-change_consistency-assessment_q02__491322b23c58", "report_id": "satellite_satellite-ice-sheet-elevation-change_consistency-assessment_q02", "dataset_id": "satellite-ice-sheet-elevation-change", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency-assessment_q02", "aspect_base": "consistency-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Spatio-temporal data completeness and consistency of remote sensing-derived ice sheet surface elevation changes over Antarctica > Methodology > Dataset description", "title": "Spatio-temporal data completeness and consistency of remote sensing-derived ice sheet surface elevation changes over Antarctica", "chunk_index": 3, "token_count": 996, "text_raw": "Surface elevation change detection by satellite radar altimetry is a useful tool to grasp the impact of climate change on the ice sheets. In that regard, the C3S dataset on the Climate Data Store (CDS) provides monthly surface elevation change (SEC) values and their uncertainty for the Greenland (GrIS) and Antarctic Ice Sheet (AIS) on a 25 km spatial resolution grid. The core principle involves measuring surface elevations at different times and comparing them to detect changes. In this dataset, they are derived using satellite radar altimetry that contain data from multiple satellite missions, which are grouped together into a consistent time series for each pixel. Surface elevation changes are reported with units of meter per year and are available since 1992. Data are provided in NetCDF format as gridded data and are available for both the GrIS (excluding peripheral glaciers and ice caps) and AIS (including ice shelves).\n\nThese ice sheet surface elevation change rates are mathematically expressed as:\n\n$\\dfrac{dh}{dt} = \\dfrac{h_{t_2}-h_{t_1}}{t_{2}-t_{1}}$\n\nwhere:\n- $h$ is the surface elevation (m)\n- $t$ the time (yr)\n\nNote that in the dataset itself, surface elevation changes are, however, actually calculated from linear regressions to a time series. The mentioned time for such a SEC measurement is the center of a 3-year of 5-year moving window (for the GrIS) and a 5-year moving window (for the AIS) used to derive the SEC values. Users should hereby note that not only accumulation or ablation (i.e. the surface mass balance) can cause the surface elevation to increase or decrease. The surface elevation change rate at a certain pixel contains signals from various processes and are generally split up into contributions from the surface mass balance (surface ablation and accumulation), ice-flow dynamics (dynamical thinning or thickening) and other processes:\n\n$\\dfrac{dh}{dt}_{obs} = \\dfrac{SMB}{\\rho_m} + \\dfrac{D'}{\\rho_i} + $ other\n\nwhere the first term on the right-hand side ($SMB$ in units kg m$^{-2}$ yr$^{-1}$) includes elevation changes from surface mass balance processes and the second one ($D'$ in units kg m$^{-2}$ yr$^{-1}$) due to ice dynamic processes. The term $\\rho_m$ is the density of the material lost or gained (snow, firn or ice) and $\\rho_i$ that of ice. Other factors that may affect surface elevation changes are basal mass balance processes (which is generally of less importance when compared to the surface mass balance), as well as firn densification/compaction and vertical bedrock motion (which, however, do not have a direct effect on mass changes of the ice sheet) [[7](https://doi.org/10.5194/tc-5-173-2011), [8](https://doi.org/10.1029/2021JF006505)]. Ice dynamics is furthermore only relevant with respect to mass changes when ice discharge across the grounding line is considered. For the GrIS as a whole, contributions to mass changes are currently driven by both SMB and ice flow dynamics at an approximately equal magnitude (i.e. 60% SMB and 40% ice dynamics), while for the AIS they are almost entirely driven by ice flow dynamics [[1](https://doi.org/10.1016/j.epsl.2018.05.015), [2](https://doi.org/10.5194/tc-13-427-2019), [9](https://doi.org/10.1038/s41586-019-1855-2), [10](https://doi.org/10.1038/s41586-018-0179-y)].\n\nIn this notebook, we use version 4.0. For a more detailed description of the data acquisition and processing methods, we refer to the [documentation on the CDS](https://cds.climate.copernicus.eu/datasets/satellite-ice-sheet-elevation-change?tab=overview) and the [ECMWF Confluence Wiki](https://confluence.ecmwf.int/pages/viewpage.action?pageId=355345393) (Copernicus Knowledge Base).", "text_with_prefix": "EQC Quality Assessment: \"Spatio-temporal data completeness and consistency of remote sensing-derived ice sheet surface elevation changes over Antarctica\"\nDataset: satellite-ice-sheet-elevation-change [CDS]\nAspect: consistency-assessment_q02 | Category: Satellite_ECVs\nSection: Spatio-temporal data completeness and consistency of remote sensing-derived ice sheet surface elevation changes over Antarctica > Methodology > Dataset description\n---\nSurface elevation change detection by satellite radar altimetry is a useful tool to grasp the impact of climate change on the ice sheets. In that regard, the C3S dataset on the Climate Data Store (CDS) provides monthly surface elevation change (SEC) values and their uncertainty for the Greenland (GrIS) and Antarctic Ice Sheet (AIS) on a 25 km spatial resolution grid. The core principle involves measuring surface elevations at different times and comparing them to detect changes. In this dataset, they are derived using satellite radar altimetry that contain data from multiple satellite missions, which are grouped together into a consistent time series for each pixel. Surface elevation changes are reported with units of meter per year and are available since 1992. Data are provided in NetCDF format as gridded data and are available for both the GrIS (excluding peripheral glaciers and ice caps) and AIS (including ice shelves).\n\nThese ice sheet surface elevation change rates are mathematically expressed as:\n\n$\\dfrac{dh}{dt} = \\dfrac{h_{t_2}-h_{t_1}}{t_{2}-t_{1}}$\n\nwhere:\n- $h$ is the surface elevation (m)\n- $t$ the time (yr)\n\nNote that in the dataset itself, surface elevation changes are, however, actually calculated from linear regressions to a time series. The mentioned time for such a SEC measurement is the center of a 3-year of 5-year moving window (for the GrIS) and a 5-year moving window (for the AIS) used to derive the SEC values. Users should hereby note that not only accumulation or ablation (i.e. the surface mass balance) can cause the surface elevation to increase or decrease. The surface elevation change rate at a certain pixel contains signals from various processes and are generally split up into contributions from the surface mass balance (surface ablation and accumulation), ice-flow dynamics (dynamical thinning or thickening) and other processes:\n\n$\\dfrac{dh}{dt}_{obs} = \\dfrac{SMB}{\\rho_m} + \\dfrac{D'}{\\rho_i} + $ other\n\nwhere the first term on the right-hand side ($SMB$ in units kg m$^{-2}$ yr$^{-1}$) includes elevation changes from surface mass balance processes and the second one ($D'$ in units kg m$^{-2}$ yr$^{-1}$) due to ice dynamic processes. The term $\\rho_m$ is the density of the material lost or gained (snow, firn or ice) and $\\rho_i$ that of ice. Other factors that may affect surface elevation changes are basal mass balance processes (which is generally of less importance when compared to the surface mass balance), as well as firn densification/compaction and vertical bedrock motion (which, however, do not have a direct effect on mass changes of the ice sheet) [[7](https://doi.org/10.5194/tc-5-173-2011), [8](https://doi.org/10.1029/2021JF006505)]. Ice dynamics is furthermore only relevant with respect to mass changes when ice discharge across the grounding line is considered. For the GrIS as a whole, contributions to mass changes are currently driven by both SMB and ice flow dynamics at an approximately equal magnitude (i.e. 60% SMB and 40% ice dynamics), while for the AIS they are almost entirely driven by ice flow dynamics [[1](https://doi.org/10.1016/j.epsl.2018.05.015), [2](https://doi.org/10.5194/tc-13-427-2019), [9](https://doi.org/10.1038/s41586-019-1855-2), [10](https://doi.org/10.1038/s41586-018-0179-y)].\n\nIn this notebook, we use version 4.0. For a more detailed description of the data acquisition and processing methods, we refer to the [documentation on the CDS](https://cds.climate.copernicus.eu/datasets/satellite-ice-sheet-elevation-change?tab=overview) and the [ECMWF Confluence Wiki](https://confluence.ecmwf.int/pages/viewpage.action?pageId=355345393) (Copernicus Knowledge Base)."} {"chunk_id": "satellite_satellite-ice-sheet-elevation-change_consistency-assessment_q02__b6770d58a489", "report_id": "satellite_satellite-ice-sheet-elevation-change_consistency-assessment_q02", "dataset_id": "satellite-ice-sheet-elevation-change", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency-assessment_q02", "aspect_base": "consistency-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Spatio-temporal data completeness and consistency of remote sensing-derived ice sheet surface elevation changes over Antarctica > Methodology > Structure and (sub)sections", "title": "Spatio-temporal data completeness and consistency of remote sensing-derived ice sheet surface elevation changes over Antarctica", "chunk_index": 4, "token_count": 189, "text_raw": "**[](section-1)**\n\n* [](section-1-1)\n* [](section-1-2)\n* [](section-1-3)\n* [](section-1-4)\n\n**[](section-2)**\n* [](section-2-1)\n* [](section-2-2)\n\n**[](section-3)**\n* [](section-3-1)\n* [](section-3-2)\n* [](section-3-3)\n\n**[](section-4)**", "text_with_prefix": "EQC Quality Assessment: \"Spatio-temporal data completeness and consistency of remote sensing-derived ice sheet surface elevation changes over Antarctica\"\nDataset: satellite-ice-sheet-elevation-change [CDS]\nAspect: consistency-assessment_q02 | Category: Satellite_ECVs\nSection: Spatio-temporal data completeness and consistency of remote sensing-derived ice sheet surface elevation changes over Antarctica > Methodology > Structure and (sub)sections\n---\n**[](section-1)**\n\n* [](section-1-1)\n* [](section-1-2)\n* [](section-1-3)\n* [](section-1-4)\n\n**[](section-2)**\n* [](section-2-1)\n* [](section-2-2)\n\n**[](section-3)**\n* [](section-3-1)\n* [](section-3-2)\n* [](section-3-3)\n\n**[](section-4)**"} {"chunk_id": "satellite_satellite-ice-sheet-elevation-change_consistency-assessment_q02__0afa8a4804ca", "report_id": "satellite_satellite-ice-sheet-elevation-change_consistency-assessment_q02", "dataset_id": "satellite-ice-sheet-elevation-change", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency-assessment_q02", "aspect_base": "consistency-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Spatio-temporal data completeness and consistency of remote sensing-derived ice sheet surface elevation changes over Antarctica > Analysis and results > 1. Data preparation and processing > 1.2 Define request and download", "title": "Spatio-temporal data completeness and consistency of remote sensing-derived ice sheet surface elevation changes over Antarctica", "chunk_index": 5, "token_count": 198, "text_raw": "Then we define requests for download from the CDS and download the ice sheet surface elevation change data.\n\nSelect the domain: \"greenland\" or \"antarctica\"\nDefine the request\nDownload the data\n\n```text\ndomain='antarctica'\n```\n\n```text\n100%|██████████| 1/1 [00:00<00:00, 20.94it/s]\n```\n\n```text\nDownload completed.\n```\n\n(section-1-3)=", "text_with_prefix": "EQC Quality Assessment: \"Spatio-temporal data completeness and consistency of remote sensing-derived ice sheet surface elevation changes over Antarctica\"\nDataset: satellite-ice-sheet-elevation-change [CDS]\nAspect: consistency-assessment_q02 | Category: Satellite_ECVs\nSection: Spatio-temporal data completeness and consistency of remote sensing-derived ice sheet surface elevation changes over Antarctica > Analysis and results > 1. Data preparation and processing > 1.2 Define request and download\n---\nThen we define requests for download from the CDS and download the ice sheet surface elevation change data.\n\nSelect the domain: \"greenland\" or \"antarctica\"\nDefine the request\nDownload the data\n\n```text\ndomain='antarctica'\n```\n\n```text\n100%|██████████| 1/1 [00:00<00:00, 20.94it/s]\n```\n\n```text\nDownload completed.\n```\n\n(section-1-3)="} {"chunk_id": "satellite_satellite-ice-sheet-elevation-change_consistency-assessment_q02__f04ba34a11e7", "report_id": "satellite_satellite-ice-sheet-elevation-change_consistency-assessment_q02", "dataset_id": "satellite-ice-sheet-elevation-change", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency-assessment_q02", "aspect_base": "consistency-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Spatio-temporal data completeness and consistency of remote sensing-derived ice sheet surface elevation changes over Antarctica > Analysis and results > 1. Data preparation and processing > 1.3 Display and inspect data", "title": "Spatio-temporal data completeness and consistency of remote sensing-derived ice sheet surface elevation changes over Antarctica", "chunk_index": 6, "token_count": 951, "text_raw": "We can read and inspect the data. Let us print out the data to inspect its structure:\n\n```text\n{'antarctica': Size: 108MB\n Dimensions: (y: 180, x: 216, time: 307, bounds: 2)\n Coordinates:\n * time (time) datetime64[ns] 2kB 1994-11-01T09:21:19.687500 ......\n grid_x_bounds (bounds, x) float32 2kB -2.6e+06 -2.575e+06 ... 2.8e+06\n grid_y_bounds (bounds, y) float32 1kB -2.2e+06 -2.175e+06 ... 2.3e+06\n grid_lon_bounds (bounds, y, x) float32 311kB 229.8 229.5 ... 50.35 50.6\n grid_lat_bounds (bounds, y, x) float32 311kB -59.37 -59.54 ... -57.51\n time_bounds (bounds, time) datetime64[ns] 5kB 1992-05-02T06:21:26.71...\n * x (x) float32 864B -2.588e+06 -2.562e+06 ... 2.788e+06\n * y (y) float32 720B -2.188e+06 -2.162e+06 ... 2.288e+06\n Dimensions without coordinates: bounds\n Data variables:\n longitude (y, x) float32 156kB 229.8 229.5 229.2 ... 50.37 50.63\n latitude (y, x) float32 156kB -59.53 -59.69 -59.85 ... -57.83 -57.66\n sec (y, x, time) float32 48MB nan nan nan nan ... nan nan nan\n sec_uncert (y, x, time) float32 48MB nan nan nan nan ... nan nan nan\n sec_ok (y, x, time) int8 12MB 0 0 0 0 0 0 0 0 ... 0 0 0 0 0 0 0 0\n surface_type (y, x) int8 39kB 0 0 0 0 0 0 0 0 0 0 ... 0 0 0 0 0 0 0 0 0\n high_slope (y, x) int8 39kB 0 0 0 0 0 0 0 0 0 0 ... 0 0 0 0 0 0 0 0 0\n grid_projection |S1 1B b''\n Attributes: (12/13)\n Conventions: CF-1.7\n title: Surface Elevation Change Rate of the Antarctic Ice Sheet\n references: Main: Wingham, Shepherd, Muir and Marshall, Phil Trans R ...\n source: ESA Radar altimeters: ERS-1, ERS-2, Envisat, CryoSat-2, S...\n institution: Copernicus Climate Change Service\n contact: copernicus-support@ecmwf.int\n ... ...\n creation_date: 2022-12-16T:10:07:43Z\n comment: Data is geophysically corrected, instruments power correc...\n history: Product version 4.0\n summary: Surface elevation change rate derived for Antarctica in 2...\n keywords: EARTH SCIENCE CLIMATE INDICATORS CRYOSPHERIC INDICATORS G...\n license: C3S general license}\n```", "text_with_prefix": "EQC Quality Assessment: \"Spatio-temporal data completeness and consistency of remote sensing-derived ice sheet surface elevation changes over Antarctica\"\nDataset: satellite-ice-sheet-elevation-change [CDS]\nAspect: consistency-assessment_q02 | Category: Satellite_ECVs\nSection: Spatio-temporal data completeness and consistency of remote sensing-derived ice sheet surface elevation changes over Antarctica > Analysis and results > 1. Data preparation and processing > 1.3 Display and inspect data\n---\nWe can read and inspect the data. Let us print out the data to inspect its structure:\n\n```text\n{'antarctica': Size: 108MB\n Dimensions: (y: 180, x: 216, time: 307, bounds: 2)\n Coordinates:\n * time (time) datetime64[ns] 2kB 1994-11-01T09:21:19.687500 ......\n grid_x_bounds (bounds, x) float32 2kB -2.6e+06 -2.575e+06 ... 2.8e+06\n grid_y_bounds (bounds, y) float32 1kB -2.2e+06 -2.175e+06 ... 2.3e+06\n grid_lon_bounds (bounds, y, x) float32 311kB 229.8 229.5 ... 50.35 50.6\n grid_lat_bounds (bounds, y, x) float32 311kB -59.37 -59.54 ... -57.51\n time_bounds (bounds, time) datetime64[ns] 5kB 1992-05-02T06:21:26.71...\n * x (x) float32 864B -2.588e+06 -2.562e+06 ... 2.788e+06\n * y (y) float32 720B -2.188e+06 -2.162e+06 ... 2.288e+06\n Dimensions without coordinates: bounds\n Data variables:\n longitude (y, x) float32 156kB 229.8 229.5 229.2 ... 50.37 50.63\n latitude (y, x) float32 156kB -59.53 -59.69 -59.85 ... -57.83 -57.66\n sec (y, x, time) float32 48MB nan nan nan nan ... nan nan nan\n sec_uncert (y, x, time) float32 48MB nan nan nan nan ... nan nan nan\n sec_ok (y, x, time) int8 12MB 0 0 0 0 0 0 0 0 ... 0 0 0 0 0 0 0 0\n surface_type (y, x) int8 39kB 0 0 0 0 0 0 0 0 0 0 ... 0 0 0 0 0 0 0 0 0\n high_slope (y, x) int8 39kB 0 0 0 0 0 0 0 0 0 0 ... 0 0 0 0 0 0 0 0 0\n grid_projection |S1 1B b''\n Attributes: (12/13)\n Conventions: CF-1.7\n title: Surface Elevation Change Rate of the Antarctic Ice Sheet\n references: Main: Wingham, Shepherd, Muir and Marshall, Phil Trans R ...\n source: ESA Radar altimeters: ERS-1, ERS-2, Envisat, CryoSat-2, S...\n institution: Copernicus Climate Change Service\n contact: copernicus-support@ecmwf.int\n ... ...\n creation_date: 2022-12-16T:10:07:43Z\n comment: Data is geophysically corrected, instruments power correc...\n history: Product version 4.0\n summary: Surface elevation change rate derived for Antarctica in 2...\n keywords: EARTH SCIENCE CLIMATE INDICATORS CRYOSPHERIC INDICATORS G...\n license: C3S general license}\n```"} {"chunk_id": "satellite_satellite-ice-sheet-elevation-change_consistency-assessment_q02__92f85ba7212a", "report_id": "satellite_satellite-ice-sheet-elevation-change_consistency-assessment_q02", "dataset_id": "satellite-ice-sheet-elevation-change", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency-assessment_q02", "aspect_base": "consistency-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Spatio-temporal data completeness and consistency of remote sensing-derived ice sheet surface elevation changes over Antarctica > Analysis and results > 1. Data preparation and processing > 1.3 Display and inspect data", "title": "Spatio-temporal data completeness and consistency of remote sensing-derived ice sheet surface elevation changes over Antarctica", "chunk_index": 7, "token_count": 358, "text_raw": "rate derived for Antarctica in 2...\n keywords: EARTH SCIENCE CLIMATE INDICATORS CRYOSPHERIC INDICATORS G...\n license: C3S general license}\n```\n\nThe version 4.0 is a gridded dataset at a 25 km spatial resolution containing monthly values of the ice sheet surface elevation change rate $\\frac{dh}{dt}$ (`sec` in m/yr) and its uncertainty (`sec_uncert` in m/yr) since 1992. The time for a measurement mentioned in the dataset is the center of a 5-year moving window used to derive the surface elevation change values. For the AIS, uncertainties are reported as the sum of the modeling error, the cross-calibration error, and the measurement uncertainties. These can be considered 1-sigma precision errors. A land mask (`surface_type`), slope mask (`high_slope`), validity flags (`sec_ok`) are also included.\n\nLet us check the total temporal extent of the data:\n\n```text\nThe begin period of the dataset is 1992-05-02 and the end period is 2022-11-01, which is a total of time 30.50 years.\n```\n\n(section-1-4)=", "text_with_prefix": "EQC Quality Assessment: \"Spatio-temporal data completeness and consistency of remote sensing-derived ice sheet surface elevation changes over Antarctica\"\nDataset: satellite-ice-sheet-elevation-change [CDS]\nAspect: consistency-assessment_q02 | Category: Satellite_ECVs\nSection: Spatio-temporal data completeness and consistency of remote sensing-derived ice sheet surface elevation changes over Antarctica > Analysis and results > 1. Data preparation and processing > 1.3 Display and inspect data\n---\nrate derived for Antarctica in 2...\n keywords: EARTH SCIENCE CLIMATE INDICATORS CRYOSPHERIC INDICATORS G...\n license: C3S general license}\n```\n\nThe version 4.0 is a gridded dataset at a 25 km spatial resolution containing monthly values of the ice sheet surface elevation change rate $\\frac{dh}{dt}$ (`sec` in m/yr) and its uncertainty (`sec_uncert` in m/yr) since 1992. The time for a measurement mentioned in the dataset is the center of a 5-year moving window used to derive the surface elevation change values. For the AIS, uncertainties are reported as the sum of the modeling error, the cross-calibration error, and the measurement uncertainties. These can be considered 1-sigma precision errors. A land mask (`surface_type`), slope mask (`high_slope`), validity flags (`sec_ok`) are also included.\n\nLet us check the total temporal extent of the data:\n\n```text\nThe begin period of the dataset is 1992-05-02 and the end period is 2022-11-01, which is a total of time 30.50 years.\n```\n\n(section-1-4)="} {"chunk_id": "satellite_satellite-ice-sheet-elevation-change_consistency-assessment_q02__95a6748c425d", "report_id": "satellite_satellite-ice-sheet-elevation-change_consistency-assessment_q02", "dataset_id": "satellite-ice-sheet-elevation-change", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency-assessment_q02", "aspect_base": "consistency-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Spatio-temporal data completeness and consistency of remote sensing-derived ice sheet surface elevation changes over Antarctica > Analysis and results > 1. Data preparation and processing > 1.4 Data handling and creating functions", "title": "Spatio-temporal data completeness and consistency of remote sensing-derived ice sheet surface elevation changes over Antarctica", "chunk_index": 8, "token_count": 242, "text_raw": "Let us now perform some data handling and define a plotting function before getting started with the analysis:\n\nSelect the specific dataset you want to process, e.g., \"greenland\"\nDefine plotting function\nCreate subplots with Polar Stereographic projection\nPlot the data\nSet extent and plot features\nAdd colorbar\n\nNow, our dataset array only holds the most important information, such as the multiyear mean surface elevation change rates (`sec`) and the arithmetic mean precision error (`sec_err`). With the function above, we also calculated linear and quadratic trends of surface elevation changes, as well as the amount of missing values.\n\nWith everything ready, let us now begin with the analysis:\n\n(section-2)=", "text_with_prefix": "EQC Quality Assessment: \"Spatio-temporal data completeness and consistency of remote sensing-derived ice sheet surface elevation changes over Antarctica\"\nDataset: satellite-ice-sheet-elevation-change [CDS]\nAspect: consistency-assessment_q02 | Category: Satellite_ECVs\nSection: Spatio-temporal data completeness and consistency of remote sensing-derived ice sheet surface elevation changes over Antarctica > Analysis and results > 1. Data preparation and processing > 1.4 Data handling and creating functions\n---\nLet us now perform some data handling and define a plotting function before getting started with the analysis:\n\nSelect the specific dataset you want to process, e.g., \"greenland\"\nDefine plotting function\nCreate subplots with Polar Stereographic projection\nPlot the data\nSet extent and plot features\nAdd colorbar\n\nNow, our dataset array only holds the most important information, such as the multiyear mean surface elevation change rates (`sec`) and the arithmetic mean precision error (`sec_err`). With the function above, we also calculated linear and quadratic trends of surface elevation changes, as well as the amount of missing values.\n\nWith everything ready, let us now begin with the analysis:\n\n(section-2)="} {"chunk_id": "satellite_satellite-ice-sheet-elevation-change_consistency-assessment_q02__8292c62d14ad", "report_id": "satellite_satellite-ice-sheet-elevation-change_consistency-assessment_q02", "dataset_id": "satellite-ice-sheet-elevation-change", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency-assessment_q02", "aspect_base": "consistency-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Spatio-temporal data completeness and consistency of remote sensing-derived ice sheet surface elevation changes over Antarctica > Analysis and results > 2. Quantifying Antarctic Ice Sheet surface elevation changes in space and time > 2.1 Spatial distribution of surface elevation changes", "title": "Spatio-temporal data completeness and consistency of remote sensing-derived ice sheet surface elevation changes over Antarctica", "chunk_index": 9, "token_count": 503, "text_raw": "We begin by plotting the Antarctic multiyear mean surface elevation change rate $\\overline {\\frac{dh}{dt}}$ between the beginning and end period with the defined plotting function:\n\nApply the function to the surface elevation change rate data\nDefine dataset to be plotted\n\n*Figure 1. Multiyear mean surface elevation change (SEC) over Antarctica from all available data in the SEC dataset on the Climate Data Store. It must be noted that the period over which the data are averaged in this figure may differ, depending on available data (e.g. with respect to the varying polar gap or other missing data).*\n\nThe figure displays the multiyear mean surface elevation change of the Antarctic ice sheet from 1992 to 2022 in meter per year (m/yr). In many parts of the ice sheet, the pattern, however, exhibits a noisy appearance. A notable data gap is also present around the South Pole (i.e. the well-known 'polar gap', of which the extent varies depending on the satellite mission). It must be noted that the period over which the data are averaged in Figure 1 may differ, depending on the availability of the data (e.g. due to the varying polar gap or other missing data in steep marginal regions). It is therefore only a quick representation of the data and cannot be used for proper glaciological analysis.\n\nLet us quantify the ice sheet-wide average value:\n\n```text\nThe Antarctic ice sheet-wide average surface elevation change rate value between 1992-05-02 and 2022-11-01 is -0.01691 m/yr.\n```\n\nThe negative value indicates that the surface of the ice sheet, in general, has been slightly lowering during the last several decades, resulting in net negative values. Again, we note the presence of abundant noisy and missing data (see later for more details), which may impact the robustness of this value.\n\n(section-2-2)=", "text_with_prefix": "EQC Quality Assessment: \"Spatio-temporal data completeness and consistency of remote sensing-derived ice sheet surface elevation changes over Antarctica\"\nDataset: satellite-ice-sheet-elevation-change [CDS]\nAspect: consistency-assessment_q02 | Category: Satellite_ECVs\nSection: Spatio-temporal data completeness and consistency of remote sensing-derived ice sheet surface elevation changes over Antarctica > Analysis and results > 2. Quantifying Antarctic Ice Sheet surface elevation changes in space and time > 2.1 Spatial distribution of surface elevation changes\n---\nWe begin by plotting the Antarctic multiyear mean surface elevation change rate $\\overline {\\frac{dh}{dt}}$ between the beginning and end period with the defined plotting function:\n\nApply the function to the surface elevation change rate data\nDefine dataset to be plotted\n\n*Figure 1. Multiyear mean surface elevation change (SEC) over Antarctica from all available data in the SEC dataset on the Climate Data Store. It must be noted that the period over which the data are averaged in this figure may differ, depending on available data (e.g. with respect to the varying polar gap or other missing data).*\n\nThe figure displays the multiyear mean surface elevation change of the Antarctic ice sheet from 1992 to 2022 in meter per year (m/yr). In many parts of the ice sheet, the pattern, however, exhibits a noisy appearance. A notable data gap is also present around the South Pole (i.e. the well-known 'polar gap', of which the extent varies depending on the satellite mission). It must be noted that the period over which the data are averaged in Figure 1 may differ, depending on the availability of the data (e.g. due to the varying polar gap or other missing data in steep marginal regions). It is therefore only a quick representation of the data and cannot be used for proper glaciological analysis.\n\nLet us quantify the ice sheet-wide average value:\n\n```text\nThe Antarctic ice sheet-wide average surface elevation change rate value between 1992-05-02 and 2022-11-01 is -0.01691 m/yr.\n```\n\nThe negative value indicates that the surface of the ice sheet, in general, has been slightly lowering during the last several decades, resulting in net negative values. Again, we note the presence of abundant noisy and missing data (see later for more details), which may impact the robustness of this value.\n\n(section-2-2)="} {"chunk_id": "satellite_satellite-ice-sheet-elevation-change_consistency-assessment_q02__7df29dcda5ae", "report_id": "satellite_satellite-ice-sheet-elevation-change_consistency-assessment_q02", "dataset_id": "satellite-ice-sheet-elevation-change", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency-assessment_q02", "aspect_base": "consistency-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Spatio-temporal data completeness and consistency of remote sensing-derived ice sheet surface elevation changes over Antarctica > Analysis and results > 2. Quantifying Antarctic Ice Sheet surface elevation changes in space and time > 2.2 Time series of surface elevation changes", "title": "Spatio-temporal data completeness and consistency of remote sensing-derived ice sheet surface elevation changes over Antarctica", "chunk_index": 10, "token_count": 442, "text_raw": "Let us now express the average surface elevation change for the ice sheet as a whole as a time series:\n\nDefine the function\nApply mean to all variables\nCombine the results\nPlot the data\n\n*Figure 2. Time series of (left) ice sheet-wide surface elevation change (SEC) and (right) cumulative ice sheet-wide surface elevation change over Antarctica from the SEC dataset on the Climate Data Store.*\n\nOverall, the surface elevation of the Antarctic Ice Sheet has shown a long-term declining trend. The relatively consistent downward trajectory of the cumulative surface elevation change (SEC) curve in the figure above may therefore serve as a potential indicator of ongoing ice sheet imbalance, assuming that the underlying dataset is of sufficient quality, particularly with regard to spatio-temporal completeness and the consistency of time series derived from multiple satellite missions.\n\nHowever, one notable anomaly in the SEC curve is the presence of pronounced spikes around ca. 2012 (negative) and 2016 (positive), which may warrant further investigation. Hence, these peaks may not be due to actual elevation changes, but instead may have occurred due to changes in sampling and/or processing limitations, because such sharp peaks in the SEC values around these periods are not noted in other studies for the AIS (e.g. [[2](https://tc.copernicus.org/articles/13/427/2019/), [5](https://doi.org/10.5194/essd-14-3573-2022)]). In order to further investigate this, let us therefore inspect the spatio-temporal data coverage and the amount of missing data in the SEC dataset.\n\n(section-3)=", "text_with_prefix": "EQC Quality Assessment: \"Spatio-temporal data completeness and consistency of remote sensing-derived ice sheet surface elevation changes over Antarctica\"\nDataset: satellite-ice-sheet-elevation-change [CDS]\nAspect: consistency-assessment_q02 | Category: Satellite_ECVs\nSection: Spatio-temporal data completeness and consistency of remote sensing-derived ice sheet surface elevation changes over Antarctica > Analysis and results > 2. Quantifying Antarctic Ice Sheet surface elevation changes in space and time > 2.2 Time series of surface elevation changes\n---\nLet us now express the average surface elevation change for the ice sheet as a whole as a time series:\n\nDefine the function\nApply mean to all variables\nCombine the results\nPlot the data\n\n*Figure 2. Time series of (left) ice sheet-wide surface elevation change (SEC) and (right) cumulative ice sheet-wide surface elevation change over Antarctica from the SEC dataset on the Climate Data Store.*\n\nOverall, the surface elevation of the Antarctic Ice Sheet has shown a long-term declining trend. The relatively consistent downward trajectory of the cumulative surface elevation change (SEC) curve in the figure above may therefore serve as a potential indicator of ongoing ice sheet imbalance, assuming that the underlying dataset is of sufficient quality, particularly with regard to spatio-temporal completeness and the consistency of time series derived from multiple satellite missions.\n\nHowever, one notable anomaly in the SEC curve is the presence of pronounced spikes around ca. 2012 (negative) and 2016 (positive), which may warrant further investigation. Hence, these peaks may not be due to actual elevation changes, but instead may have occurred due to changes in sampling and/or processing limitations, because such sharp peaks in the SEC values around these periods are not noted in other studies for the AIS (e.g. [[2](https://tc.copernicus.org/articles/13/427/2019/), [5](https://doi.org/10.5194/essd-14-3573-2022)]). In order to further investigate this, let us therefore inspect the spatio-temporal data coverage and the amount of missing data in the SEC dataset.\n\n(section-3)="} {"chunk_id": "satellite_satellite-ice-sheet-elevation-change_consistency-assessment_q02__1d96e45c44af", "report_id": "satellite_satellite-ice-sheet-elevation-change_consistency-assessment_q02", "dataset_id": "satellite-ice-sheet-elevation-change", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency-assessment_q02", "aspect_base": "consistency-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Spatio-temporal data completeness and consistency of remote sensing-derived ice sheet surface elevation changes over Antarctica > Analysis and results > 3. Inspecting spatio-temporal data coverage and missing data > 3.1 Amount of missing data over time", "title": "Spatio-temporal data completeness and consistency of remote sensing-derived ice sheet surface elevation changes over Antarctica", "chunk_index": 11, "token_count": 386, "text_raw": "Let us now plot the amount of missing data as a time series in order to investigate the data coverage. In the dataset, this can be assessed comparing the amount of pixels with valid data to the (fixed-in-time) land mask (`surface_type`), where pixels with a value larger than 0 represent ice-covered grid points of the main ice sheet body or ice shelves. The time series looks as follows:\n\n*Figure 3. Time series of missing data of the surface elevation change data over Antarctica from the SEC dataset on the Climate Data Store.*\n\nIn the figure above, we note several jumps in the curve which are presumably due to changes in the used data acquisition methods (e.g. varying satellite missions), but can also be due to data processing limitations. When inspecting the graph, some notable increases in missing data seem to occur after 2010. Moreover, the periods around ca. 2012 and 2016 are characterized by a drastic switch in missing data, which may also be linked to the spikes in the surface elevation change time series noted earlier in Figure 2. This indicates that some signals in the SEC time series may reflect changing sampling conditions and/or processing limitations, rather than actual SEC, which is hence an important factor to consider when interprating these data due to the lack of applied corrections (such as gap filling).\n\n(section-3-2)=", "text_with_prefix": "EQC Quality Assessment: \"Spatio-temporal data completeness and consistency of remote sensing-derived ice sheet surface elevation changes over Antarctica\"\nDataset: satellite-ice-sheet-elevation-change [CDS]\nAspect: consistency-assessment_q02 | Category: Satellite_ECVs\nSection: Spatio-temporal data completeness and consistency of remote sensing-derived ice sheet surface elevation changes over Antarctica > Analysis and results > 3. Inspecting spatio-temporal data coverage and missing data > 3.1 Amount of missing data over time\n---\nLet us now plot the amount of missing data as a time series in order to investigate the data coverage. In the dataset, this can be assessed comparing the amount of pixels with valid data to the (fixed-in-time) land mask (`surface_type`), where pixels with a value larger than 0 represent ice-covered grid points of the main ice sheet body or ice shelves. The time series looks as follows:\n\n*Figure 3. Time series of missing data of the surface elevation change data over Antarctica from the SEC dataset on the Climate Data Store.*\n\nIn the figure above, we note several jumps in the curve which are presumably due to changes in the used data acquisition methods (e.g. varying satellite missions), but can also be due to data processing limitations. When inspecting the graph, some notable increases in missing data seem to occur after 2010. Moreover, the periods around ca. 2012 and 2016 are characterized by a drastic switch in missing data, which may also be linked to the spikes in the surface elevation change time series noted earlier in Figure 2. This indicates that some signals in the SEC time series may reflect changing sampling conditions and/or processing limitations, rather than actual SEC, which is hence an important factor to consider when interprating these data due to the lack of applied corrections (such as gap filling).\n\n(section-3-2)="} {"chunk_id": "satellite_satellite-ice-sheet-elevation-change_consistency-assessment_q02__ab0a5b5fea8b", "report_id": "satellite_satellite-ice-sheet-elevation-change_consistency-assessment_q02", "dataset_id": "satellite-ice-sheet-elevation-change", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency-assessment_q02", "aspect_base": "consistency-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Spatio-temporal data completeness and consistency of remote sensing-derived ice sheet surface elevation changes over Antarctica > Analysis and results > 3. Inspecting spatio-temporal data coverage and missing data > 3.2 Spastial distribution of missing data", "title": "Spatio-temporal data completeness and consistency of remote sensing-derived ice sheet surface elevation changes over Antarctica", "chunk_index": 12, "token_count": 587, "text_raw": "Let us have the spatial distribution of the missing data plotted:\n\nApply the function to the surface elevation change missing data\nPlot the data\n\n*Figure 4. Spatial distribution of valid data of the surface elevation change data over Antarctica from the SEC dataset on the Climate Data Store.*\n\nThe figure above displays the percentage of valid data in the surface elevation change (SEC) data for Antarctica, which further allows to inspect its data completeness. As discussed before, data gaps in the SEC dataset arise primarily from limited satellite coverage beyond certain mission-specific latitudes (i.e. the \"polar gap\" around the South Pole), and from processing limitations or challenging conditions, such as complex and steep terrain around the marginal regions (e.g. the Antarctic Peninsula and the transantarctic mountains). Both areas are clearly marked by a lesser amount of missing data in the figure above, providing additional insight into the pattern of missing data.\n\nThe central Antarctic region near the South Pole exhibits a low coverage due to the so-called “polar gap”, where satellite altimetry coverage is limited. Early satellite missions like ERS-1, ERS-2, and Envisat had coverage limitations south of 81.5°S, creating significant gaps. CryoSat-2 (launched in 2010) improved polar coverage up to 88°S but still left significant gaps in the central polar region and partly around the margins [[2](https://doi.org/10.5194/tc-13-427-2019), [3](https://doi.org/10.1029/2019GL082182), [4](https://climate.esa.int/media/documents/ST-UL-ESA-AISCCI-E3UB-001-v1.0.pdf)]. The Sentinel-3 mission has recently further enhanced the sampling frequency, but coverage is again rather limited in the coastal regions due to the absence of an interferometric SAR. Near the margins, challenging conditions, such as complex and steep terrain, limit the efficient acquisition of sufficient valid data [[4](https://climate.esa.int/media/documents/ST-UL-ESA-AISCCI-E3UB-001-v1.0.pdf)]. Unlike the Greenland SEC dataset, these data gaps for Antarctica have not been filled up during post-processing, resulting in more pronounced areas with missing data.\n\n(section-3-3)=", "text_with_prefix": "EQC Quality Assessment: \"Spatio-temporal data completeness and consistency of remote sensing-derived ice sheet surface elevation changes over Antarctica\"\nDataset: satellite-ice-sheet-elevation-change [CDS]\nAspect: consistency-assessment_q02 | Category: Satellite_ECVs\nSection: Spatio-temporal data completeness and consistency of remote sensing-derived ice sheet surface elevation changes over Antarctica > Analysis and results > 3. Inspecting spatio-temporal data coverage and missing data > 3.2 Spastial distribution of missing data\n---\nLet us have the spatial distribution of the missing data plotted:\n\nApply the function to the surface elevation change missing data\nPlot the data\n\n*Figure 4. Spatial distribution of valid data of the surface elevation change data over Antarctica from the SEC dataset on the Climate Data Store.*\n\nThe figure above displays the percentage of valid data in the surface elevation change (SEC) data for Antarctica, which further allows to inspect its data completeness. As discussed before, data gaps in the SEC dataset arise primarily from limited satellite coverage beyond certain mission-specific latitudes (i.e. the \"polar gap\" around the South Pole), and from processing limitations or challenging conditions, such as complex and steep terrain around the marginal regions (e.g. the Antarctic Peninsula and the transantarctic mountains). Both areas are clearly marked by a lesser amount of missing data in the figure above, providing additional insight into the pattern of missing data.\n\nThe central Antarctic region near the South Pole exhibits a low coverage due to the so-called “polar gap”, where satellite altimetry coverage is limited. Early satellite missions like ERS-1, ERS-2, and Envisat had coverage limitations south of 81.5°S, creating significant gaps. CryoSat-2 (launched in 2010) improved polar coverage up to 88°S but still left significant gaps in the central polar region and partly around the margins [[2](https://doi.org/10.5194/tc-13-427-2019), [3](https://doi.org/10.1029/2019GL082182), [4](https://climate.esa.int/media/documents/ST-UL-ESA-AISCCI-E3UB-001-v1.0.pdf)]. The Sentinel-3 mission has recently further enhanced the sampling frequency, but coverage is again rather limited in the coastal regions due to the absence of an interferometric SAR. Near the margins, challenging conditions, such as complex and steep terrain, limit the efficient acquisition of sufficient valid data [[4](https://climate.esa.int/media/documents/ST-UL-ESA-AISCCI-E3UB-001-v1.0.pdf)]. Unlike the Greenland SEC dataset, these data gaps for Antarctica have not been filled up during post-processing, resulting in more pronounced areas with missing data.\n\n(section-3-3)="} {"chunk_id": "satellite_satellite-ice-sheet-elevation-change_consistency-assessment_q02__8920e8da25c4", "report_id": "satellite_satellite-ice-sheet-elevation-change_consistency-assessment_q02", "dataset_id": "satellite-ice-sheet-elevation-change", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency-assessment_q02", "aspect_base": "consistency-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Spatio-temporal data completeness and consistency of remote sensing-derived ice sheet surface elevation changes over Antarctica > Analysis and results > 3. Inspecting spatio-temporal data coverage and missing data > 3.3 Link with satellite mission and surface characteristics", "title": "Spatio-temporal data completeness and consistency of remote sensing-derived ice sheet surface elevation changes over Antarctica", "chunk_index": 13, "token_count": 777, "text_raw": "Let us further investigate this pattern and check whether we can link the amount of missing data to the surface characteristics provided with the data. Therefore, we check how the amount of missing data outside of the polar gap (taken as 81.5°S) is linked to the surface slope (i.e. as a proxy for the terrain complexity) and the magnitude of the surface elevation changes at a certain pixel:\n\nFlatten the arrays and remove NaN values for the boxplot\nax1.set_yscale('log')\nScatter plot\nRemove NaN values for the scatter plot\nax2.set_yscale('log')\n\n*Figure 5. Relationship between the percentage missing data and (left) the slope of the terrain and (right) the magnitude of the ice sheet surface elevation change of a certain pixel over Antarctica from the SEC dataset on the Climate Data Store.*\n\nThe graphs above display the data quality of radar altimetry-derived surface elevation changes over the AIS, highlighting challenges in areas with higher surface slopes and high surface elevation changes. The first graph shows that regions with gentle slopes (<2°) have the least missing data, indicating better radar performance on flat surfaces. As slopes increase to more than 5°, practically all data are missing, reflecting the difficulty of capturing reliable measurements due to slope-induced errors in complex terrain. The second graph illustrates the relationship between the magnitude of surface elevation change at a certain pixel and the percentage of missing data in that same pixel. Areas with larger elevation changes, often found along ice sheet margins with complex, high-slope terrain, tend to have higher data gaps, indicating that dynamic regions are harder to monitor consistently. This relationship, however, is obviously not deterministic, as some significant of scatter remains.\n\nOverall, the graphs reveal that flatter, more stable regions provide more reliable radar data, while steeper and dynamic regions are more prone to significant amounts of missing data. Recognizing these limitations is crucial for accurately interpreting SEC data and assessing climate change impacts on ice sheets.\n\nLet us inspect how varying satellite missions perform with respect to the amount of missing data in the C3S product based on the time period that they are/were active:\n\nDefine Antarctica dataset\nDefine missing data per time period\nBroadcast flagged_pixels over time\nCount of pixels with surface_type >0 per timestep\nAmong them, how many are NaN in sec\nAverage over time steps\n\n*Figure 6. Average percentage of missing data for various satellite missions (time periods) over Antarctica from the SEC dataset on the Climate Data Store.*\n\nThe graphs above clearly show the different missing data characteristics of the various satellite missions, which are due to changes in the used data acquisition methods/spatial coverage and data processing limitations (e.g. the size of the polar gap, the susceptibility to slope-induced errors over complex terrain, etc.). The highest amount of missing data are found in the oldest period during the epoch of ERS-1/2. The most recent period, which combines data from CryoSat-2 and Sentinel-3, shows the lowest percentage of missing data, reflecting advancements in radar altimetry technology and data processing methods. Again, these findings display the progressive improvement in the observational capability of remote sensing-derived surface elevation changes over Antarctica, while also underlining the persistent challenges in obtaining complete records.\n\n(section-4)=", "text_with_prefix": "EQC Quality Assessment: \"Spatio-temporal data completeness and consistency of remote sensing-derived ice sheet surface elevation changes over Antarctica\"\nDataset: satellite-ice-sheet-elevation-change [CDS]\nAspect: consistency-assessment_q02 | Category: Satellite_ECVs\nSection: Spatio-temporal data completeness and consistency of remote sensing-derived ice sheet surface elevation changes over Antarctica > Analysis and results > 3. Inspecting spatio-temporal data coverage and missing data > 3.3 Link with satellite mission and surface characteristics\n---\nLet us further investigate this pattern and check whether we can link the amount of missing data to the surface characteristics provided with the data. Therefore, we check how the amount of missing data outside of the polar gap (taken as 81.5°S) is linked to the surface slope (i.e. as a proxy for the terrain complexity) and the magnitude of the surface elevation changes at a certain pixel:\n\nFlatten the arrays and remove NaN values for the boxplot\nax1.set_yscale('log')\nScatter plot\nRemove NaN values for the scatter plot\nax2.set_yscale('log')\n\n*Figure 5. Relationship between the percentage missing data and (left) the slope of the terrain and (right) the magnitude of the ice sheet surface elevation change of a certain pixel over Antarctica from the SEC dataset on the Climate Data Store.*\n\nThe graphs above display the data quality of radar altimetry-derived surface elevation changes over the AIS, highlighting challenges in areas with higher surface slopes and high surface elevation changes. The first graph shows that regions with gentle slopes (<2°) have the least missing data, indicating better radar performance on flat surfaces. As slopes increase to more than 5°, practically all data are missing, reflecting the difficulty of capturing reliable measurements due to slope-induced errors in complex terrain. The second graph illustrates the relationship between the magnitude of surface elevation change at a certain pixel and the percentage of missing data in that same pixel. Areas with larger elevation changes, often found along ice sheet margins with complex, high-slope terrain, tend to have higher data gaps, indicating that dynamic regions are harder to monitor consistently. This relationship, however, is obviously not deterministic, as some significant of scatter remains.\n\nOverall, the graphs reveal that flatter, more stable regions provide more reliable radar data, while steeper and dynamic regions are more prone to significant amounts of missing data. Recognizing these limitations is crucial for accurately interpreting SEC data and assessing climate change impacts on ice sheets.\n\nLet us inspect how varying satellite missions perform with respect to the amount of missing data in the C3S product based on the time period that they are/were active:\n\nDefine Antarctica dataset\nDefine missing data per time period\nBroadcast flagged_pixels over time\nCount of pixels with surface_type >0 per timestep\nAmong them, how many are NaN in sec\nAverage over time steps\n\n*Figure 6. Average percentage of missing data for various satellite missions (time periods) over Antarctica from the SEC dataset on the Climate Data Store.*\n\nThe graphs above clearly show the different missing data characteristics of the various satellite missions, which are due to changes in the used data acquisition methods/spatial coverage and data processing limitations (e.g. the size of the polar gap, the susceptibility to slope-induced errors over complex terrain, etc.). The highest amount of missing data are found in the oldest period during the epoch of ERS-1/2. The most recent period, which combines data from CryoSat-2 and Sentinel-3, shows the lowest percentage of missing data, reflecting advancements in radar altimetry technology and data processing methods. Again, these findings display the progressive improvement in the observational capability of remote sensing-derived surface elevation changes over Antarctica, while also underlining the persistent challenges in obtaining complete records.\n\n(section-4)="} {"chunk_id": "satellite_satellite-ice-sheet-elevation-change_consistency-assessment_q02__7f28b88f5d5f", "report_id": "satellite_satellite-ice-sheet-elevation-change_consistency-assessment_q02", "dataset_id": "satellite-ice-sheet-elevation-change", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency-assessment_q02", "aspect_base": "consistency-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Spatio-temporal data completeness and consistency of remote sensing-derived ice sheet surface elevation changes over Antarctica > Analysis and results > 4. Short summary and take-home messages", "title": "Spatio-temporal data completeness and consistency of remote sensing-derived ice sheet surface elevation changes over Antarctica", "chunk_index": 14, "token_count": 684, "text_raw": "The C3S SEC data represent surface elevation changes of the AIS acquired by satellite radar altimetry that have been merged from different satellite missions over time. Each of those satellite missions are characterized by their own (dis)advantages. The time series of the ice sheet-wide average SEC rate and the cumulative SEC changes show a general trend of surface elevation lowering, reflecting an overall net lowering of the surface. Using the C3S Antarctic Ice Sheet (AIS) surface elevation change (SEC) time series as an indicator for the Antarctic ice sheet imbalance, however, presents several implications.\n\nFor example, the varying spatio-temporal coverage of the data that are being fed into these trends should be considered for proper evaluation and interpretation. The most important causes of persistent data gaps in the C3S SEC product for the AIS are the polar gap around the South Pole and complex terrain around the margins. However, combining CryoSat-2 and Sentinel-3 during the more recent years has improved the spatio-temporal coverage of the product greatly, even though unfilled gaps remain in some areas. This is especially the case in high-slope and complex marginal regions. These persistent and unfilled data gaps significantly complicate the robustness and data completeness of SEC measurements of the C3S AIS product. Since some of the areas with the most prominent surface elevation changes are consequently within the area of least coverage, special attention is required. Users of C3S AIS SEC data product are therefore most likely required to implement additional processing to derive glaciologically interpretable surface elevation changes.\n\nIn general, radar altimeters tend to perform better in the central, flat regions of Antarctica due to simpler topography and relatively stable surfaces, while marginal zones with complex and dynamic terrain (where the most significant SEC occur) exhibit a higher invalid data acquisition and an increased uncertainty. These gaps pose challenges for assessing radar altimetry-derived volume and mass changes, making C3S SEC data for Antarctica generally less reliable. To address these limitations, integrating multiple datasets, such as gap-filling through laser altimetry over the margins and sloped terrain (e.g. ICESat/ICESat-2) or by ice sheet model output, can improve the robustness of surface elevation change estimates. Additionally, higher-resolution datasets, regional studies, and advanced data processing techniques (e.g. machine learning) can help fill data gaps and enhance the accuracy and reliability of SEC measurements, providing a more comprehensive understanding of the current changes of the surface of the AIS [[2](https://doi.org/10.5194/tc-13-427-2019), [3](https://doi.org/10.1029/2019GL082182), [5](https://doi.org/10.5194/essd-14-3573-2022), [6](https://doi.org/10.1029/2020GL090572)].", "text_with_prefix": "EQC Quality Assessment: \"Spatio-temporal data completeness and consistency of remote sensing-derived ice sheet surface elevation changes over Antarctica\"\nDataset: satellite-ice-sheet-elevation-change [CDS]\nAspect: consistency-assessment_q02 | Category: Satellite_ECVs\nSection: Spatio-temporal data completeness and consistency of remote sensing-derived ice sheet surface elevation changes over Antarctica > Analysis and results > 4. Short summary and take-home messages\n---\nThe C3S SEC data represent surface elevation changes of the AIS acquired by satellite radar altimetry that have been merged from different satellite missions over time. Each of those satellite missions are characterized by their own (dis)advantages. The time series of the ice sheet-wide average SEC rate and the cumulative SEC changes show a general trend of surface elevation lowering, reflecting an overall net lowering of the surface. Using the C3S Antarctic Ice Sheet (AIS) surface elevation change (SEC) time series as an indicator for the Antarctic ice sheet imbalance, however, presents several implications.\n\nFor example, the varying spatio-temporal coverage of the data that are being fed into these trends should be considered for proper evaluation and interpretation. The most important causes of persistent data gaps in the C3S SEC product for the AIS are the polar gap around the South Pole and complex terrain around the margins. However, combining CryoSat-2 and Sentinel-3 during the more recent years has improved the spatio-temporal coverage of the product greatly, even though unfilled gaps remain in some areas. This is especially the case in high-slope and complex marginal regions. These persistent and unfilled data gaps significantly complicate the robustness and data completeness of SEC measurements of the C3S AIS product. Since some of the areas with the most prominent surface elevation changes are consequently within the area of least coverage, special attention is required. Users of C3S AIS SEC data product are therefore most likely required to implement additional processing to derive glaciologically interpretable surface elevation changes.\n\nIn general, radar altimeters tend to perform better in the central, flat regions of Antarctica due to simpler topography and relatively stable surfaces, while marginal zones with complex and dynamic terrain (where the most significant SEC occur) exhibit a higher invalid data acquisition and an increased uncertainty. These gaps pose challenges for assessing radar altimetry-derived volume and mass changes, making C3S SEC data for Antarctica generally less reliable. To address these limitations, integrating multiple datasets, such as gap-filling through laser altimetry over the margins and sloped terrain (e.g. ICESat/ICESat-2) or by ice sheet model output, can improve the robustness of surface elevation change estimates. Additionally, higher-resolution datasets, regional studies, and advanced data processing techniques (e.g. machine learning) can help fill data gaps and enhance the accuracy and reliability of SEC measurements, providing a more comprehensive understanding of the current changes of the surface of the AIS [[2](https://doi.org/10.5194/tc-13-427-2019), [3](https://doi.org/10.1029/2019GL082182), [5](https://doi.org/10.5194/essd-14-3573-2022), [6](https://doi.org/10.1029/2020GL090572)]."} {"chunk_id": "satellite_satellite-ice-sheet-elevation-change_consistency-assessment_q02__214c1dc1f6e3", "report_id": "satellite_satellite-ice-sheet-elevation-change_consistency-assessment_q02", "dataset_id": "satellite-ice-sheet-elevation-change", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency-assessment_q02", "aspect_base": "consistency-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Spatio-temporal data completeness and consistency of remote sensing-derived ice sheet surface elevation changes over Antarctica > ℹ️ If you want to know more > Key resources", "title": "Spatio-temporal data completeness and consistency of remote sensing-derived ice sheet surface elevation changes over Antarctica", "chunk_index": 15, "token_count": 349, "text_raw": "- [\"Ice sheet surface elevation change rate for Greenland and Antarctica from 1992 to present derived from satellite observations\"](https://cds.climate.copernicus.eu/datasets/satellite-ice-sheet-elevation-change?tab=overview) on the CDS\n- [Documentation on the CDS](https://cds.climate.copernicus.eu/datasets/satellite-ice-sheet-elevation-change?tab=overview) and the [ECMWF Confluence Wiki](https://confluence.ecmwf.int/pages/viewpage.action?pageId=355345393) (Copernicus Knowledge Base).\n- [Copernicus climate change indicators: ice sheets](https://climate.copernicus.eu/climate-indicators/ice-sheets)\n- [An easy-to-read article about ice sheet altimetry](https://blogs.egu.eu/divisions/cr/2023/03/03/ice-radar-altimetry/)\n- [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control` prepared by [B-Open](https://www.bopen.eu).", "text_with_prefix": "EQC Quality Assessment: \"Spatio-temporal data completeness and consistency of remote sensing-derived ice sheet surface elevation changes over Antarctica\"\nDataset: satellite-ice-sheet-elevation-change [CDS]\nAspect: consistency-assessment_q02 | Category: Satellite_ECVs\nSection: Spatio-temporal data completeness and consistency of remote sensing-derived ice sheet surface elevation changes over Antarctica > ℹ️ If you want to know more > Key resources\n---\n- [\"Ice sheet surface elevation change rate for Greenland and Antarctica from 1992 to present derived from satellite observations\"](https://cds.climate.copernicus.eu/datasets/satellite-ice-sheet-elevation-change?tab=overview) on the CDS\n- [Documentation on the CDS](https://cds.climate.copernicus.eu/datasets/satellite-ice-sheet-elevation-change?tab=overview) and the [ECMWF Confluence Wiki](https://confluence.ecmwf.int/pages/viewpage.action?pageId=355345393) (Copernicus Knowledge Base).\n- [Copernicus climate change indicators: ice sheets](https://climate.copernicus.eu/climate-indicators/ice-sheets)\n- [An easy-to-read article about ice sheet altimetry](https://blogs.egu.eu/divisions/cr/2023/03/03/ice-radar-altimetry/)\n- [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control` prepared by [B-Open](https://www.bopen.eu)."} {"chunk_id": "satellite_satellite-ice-sheet-elevation-change_consistency-assessment_q02__dc5436e2b748", "report_id": "satellite_satellite-ice-sheet-elevation-change_consistency-assessment_q02", "dataset_id": "satellite-ice-sheet-elevation-change", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency-assessment_q02", "aspect_base": "consistency-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Spatio-temporal data completeness and consistency of remote sensing-derived ice sheet surface elevation changes over Antarctica > ℹ️ If you want to know more > References", "title": "Spatio-temporal data completeness and consistency of remote sensing-derived ice sheet surface elevation changes over Antarctica", "chunk_index": 16, "token_count": 955, "text_raw": "- [[1](https://doi.org/10.1016/j.epsl.2018.05.015)] Sørensen, L. S., Simonsen, S. B., Forsberg, R., Khvorostovsky, K., Meister, R., and Engdahl, M. E. (2018). 25 years of elevation changes of the Greenland Ice Sheet from ERS, Envisat, and CryoSat-2 radar altimetry, Earth and Planetary Science Letters. 495. https://doi.org/10.1016/j.epsl.2018.05.015\n\n- [[2](https://doi.org/10.5194/tc-13-427-2019)] Schröder, L., Horwath, M., Dietrich, R., Helm, V., van den Broeke, M.R., and Ligtenberg, S.R.M. (2019). Four decades of Antarctic surface elevation change from multi-mission satellite altimetry. The Cryosphere, 13, p. 427-449. https://doi.org/10.5194/tc-13-427-2019\n\n- [[3](https://doi.org/10.1029/2019GL082182)] Shepherd, A., Gilbert, L., Muir, A.S., Konrad, H., McMillan, M., Slater, T., Briggs, H.K., Sundal, A. V., Hogg A.E., and Engdahl, M.E. (2019). Trends in Antarctic Ice Sheet elevation and mass. Geophysical Research Letters, 46, p. 8174-8183. https://doi.org/10.1029/2019GL082182\n\n- [[4](https://climate.esa.int/media/documents/ST-UL-ESA-AISCCI-E3UB-001-v1.0.pdf)] Shepherd, A., Fantin, D., and Engdahl, M. (2020). ESA Climate Change Initiative (CCI+)\nEssential Climate Variable (ECV): Antarctic_Ice_Sheet_cci+ (AIS_cci+) End-to-end Uncertainty Budget (E3UB), ST-UL-ESA-AISCCI+-E3UB-001. https://climate.esa.int/media/documents/ST-UL-ESA-AISCCI-E3UB-001-v1.0.pdf\n\n- [[5](https://doi.org/10.5194/essd-14-3573-2022)] Nilsson, J., Gardner, A. S., and Paolo, F. S. (2022). Elevation change of the Antarctic Ice Sheet: 1985 to 2020, Earth Syst. Sci. Data, 14, 3573–3598, https://doi.org/10.5194/essd-14-3573-2022\n\n- [[6](https://doi.org/10.1029/2020GL090572)] Brunt, K. M., Smith, B. E., Sutterley, T. C., Kurtz, N. T., and Neumann, T. A. (2021). Comparisons of satellite and airborne altimetry with ground-based data from the interior of the Antarctic ice sheet. Geophysical Research Letters, 48, e2020GL090572. https://doi.org/10.1029/2020GL090572\n\n- [[7](https://doi.org/10.5194/tc-5-173-2011)] Sørensen, L. S., Simonsen, S. B., Nielsen, K., Lucas-Picher, P., Spada, G., Adalgeirsdottir, G., Forsberg, R., and Hvidberg, C. S. (2011). Mass balance of the Greenland ice sheet (2003–2008) from ICESat data – the impact of interpolation, sampling and firn density, The Cryosphere, 5, 173–186, https://doi.org/10.5194/tc-5-173-2011", "text_with_prefix": "EQC Quality Assessment: \"Spatio-temporal data completeness and consistency of remote sensing-derived ice sheet surface elevation changes over Antarctica\"\nDataset: satellite-ice-sheet-elevation-change [CDS]\nAspect: consistency-assessment_q02 | Category: Satellite_ECVs\nSection: Spatio-temporal data completeness and consistency of remote sensing-derived ice sheet surface elevation changes over Antarctica > ℹ️ If you want to know more > References\n---\n- [[1](https://doi.org/10.1016/j.epsl.2018.05.015)] Sørensen, L. S., Simonsen, S. B., Forsberg, R., Khvorostovsky, K., Meister, R., and Engdahl, M. E. (2018). 25 years of elevation changes of the Greenland Ice Sheet from ERS, Envisat, and CryoSat-2 radar altimetry, Earth and Planetary Science Letters. 495. https://doi.org/10.1016/j.epsl.2018.05.015\n\n- [[2](https://doi.org/10.5194/tc-13-427-2019)] Schröder, L., Horwath, M., Dietrich, R., Helm, V., van den Broeke, M.R., and Ligtenberg, S.R.M. (2019). Four decades of Antarctic surface elevation change from multi-mission satellite altimetry. The Cryosphere, 13, p. 427-449. https://doi.org/10.5194/tc-13-427-2019\n\n- [[3](https://doi.org/10.1029/2019GL082182)] Shepherd, A., Gilbert, L., Muir, A.S., Konrad, H., McMillan, M., Slater, T., Briggs, H.K., Sundal, A. V., Hogg A.E., and Engdahl, M.E. (2019). Trends in Antarctic Ice Sheet elevation and mass. Geophysical Research Letters, 46, p. 8174-8183. https://doi.org/10.1029/2019GL082182\n\n- [[4](https://climate.esa.int/media/documents/ST-UL-ESA-AISCCI-E3UB-001-v1.0.pdf)] Shepherd, A., Fantin, D., and Engdahl, M. (2020). ESA Climate Change Initiative (CCI+)\nEssential Climate Variable (ECV): Antarctic_Ice_Sheet_cci+ (AIS_cci+) End-to-end Uncertainty Budget (E3UB), ST-UL-ESA-AISCCI+-E3UB-001. https://climate.esa.int/media/documents/ST-UL-ESA-AISCCI-E3UB-001-v1.0.pdf\n\n- [[5](https://doi.org/10.5194/essd-14-3573-2022)] Nilsson, J., Gardner, A. S., and Paolo, F. S. (2022). Elevation change of the Antarctic Ice Sheet: 1985 to 2020, Earth Syst. Sci. Data, 14, 3573–3598, https://doi.org/10.5194/essd-14-3573-2022\n\n- [[6](https://doi.org/10.1029/2020GL090572)] Brunt, K. M., Smith, B. E., Sutterley, T. C., Kurtz, N. T., and Neumann, T. A. (2021). Comparisons of satellite and airborne altimetry with ground-based data from the interior of the Antarctic ice sheet. Geophysical Research Letters, 48, e2020GL090572. https://doi.org/10.1029/2020GL090572\n\n- [[7](https://doi.org/10.5194/tc-5-173-2011)] Sørensen, L. S., Simonsen, S. B., Nielsen, K., Lucas-Picher, P., Spada, G., Adalgeirsdottir, G., Forsberg, R., and Hvidberg, C. S. (2011). Mass balance of the Greenland ice sheet (2003–2008) from ICESat data – the impact of interpolation, sampling and firn density, The Cryosphere, 5, 173–186, https://doi.org/10.5194/tc-5-173-2011"} {"chunk_id": "satellite_satellite-ice-sheet-elevation-change_consistency-assessment_q02__db227a3a26d5", "report_id": "satellite_satellite-ice-sheet-elevation-change_consistency-assessment_q02", "dataset_id": "satellite-ice-sheet-elevation-change", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency-assessment_q02", "aspect_base": "consistency-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Spatio-temporal data completeness and consistency of remote sensing-derived ice sheet surface elevation changes over Antarctica > ℹ️ If you want to know more > References", "title": "Spatio-temporal data completeness and consistency of remote sensing-derived ice sheet surface elevation changes over Antarctica", "chunk_index": 17, "token_count": 662, "text_raw": "data – the impact of interpolation, sampling and firn density, The Cryosphere, 5, 173–186, https://doi.org/10.5194/tc-5-173-2011\n\n- [[8](https://doi.org/10.1029/2021JF006505)] Khan, S. A., Bamber, J. L., Rignot, E., Helm, V., Aschwanden, A., Holland, D. M., van den Broeke, M., King, M., Noël, B., Truffer, M., Humbert, A., Solgaard, A. M., Box, J. E., Colgan, W. T., Wuite, J., Mouginot, J., Andersen, O. B., Csatho, B., Felikson, D., Fettweis, X., Forsberg, R., Gogineni, P., Joughin, I., Kjeldsen, K. K., Kuschnerus, M., Langen, P. L., Luckman, A., Luthcke, S. B., McMillan, M., Merryman Boncori, J. P., Morlighem, M., Mottram, R., Nagler, T., Nagy, T., Paden, J., Palmer, S., Poinar, K., Shepherd, A., Smith, B., Stearns, L. A., van Angelen, J. H., van der Wal, W., van de Berg, W. J., van Wessem, M., Velicogna, I., Wahr, J., Wendt, A., Wouters, B., & Zwally, H. J. (2022). Greenland mass trends from airborne and satellite altimetry during 2011–2020. Journal of Geophysical Research: Earth Surface, 127(3). https://doi.org/10.1029/2021JF006505\n\n- [[9](https://doi.org/10.1038/s41586-019-1855-2)] The IMBIE Team (2019). Mass balance of the Greenland Ice Sheet from 1992 to 2018. Nature, 579, 233–239. https://doi.org/10.1038/s41586-019-1855-2\n\n- [[10](https://doi.org/10.1038/s41586-018-0179-y)] The IMBIE Team (2018). Mass balance of the Antarctic Ice Sheet from 1992 to 2017. Nature 558, 219–222. https://doi.org/10.1038/s41586-018-0179-y", "text_with_prefix": "EQC Quality Assessment: \"Spatio-temporal data completeness and consistency of remote sensing-derived ice sheet surface elevation changes over Antarctica\"\nDataset: satellite-ice-sheet-elevation-change [CDS]\nAspect: consistency-assessment_q02 | Category: Satellite_ECVs\nSection: Spatio-temporal data completeness and consistency of remote sensing-derived ice sheet surface elevation changes over Antarctica > ℹ️ If you want to know more > References\n---\ndata – the impact of interpolation, sampling and firn density, The Cryosphere, 5, 173–186, https://doi.org/10.5194/tc-5-173-2011\n\n- [[8](https://doi.org/10.1029/2021JF006505)] Khan, S. A., Bamber, J. L., Rignot, E., Helm, V., Aschwanden, A., Holland, D. M., van den Broeke, M., King, M., Noël, B., Truffer, M., Humbert, A., Solgaard, A. M., Box, J. E., Colgan, W. T., Wuite, J., Mouginot, J., Andersen, O. B., Csatho, B., Felikson, D., Fettweis, X., Forsberg, R., Gogineni, P., Joughin, I., Kjeldsen, K. K., Kuschnerus, M., Langen, P. L., Luckman, A., Luthcke, S. B., McMillan, M., Merryman Boncori, J. P., Morlighem, M., Mottram, R., Nagler, T., Nagy, T., Paden, J., Palmer, S., Poinar, K., Shepherd, A., Smith, B., Stearns, L. A., van Angelen, J. H., van der Wal, W., van de Berg, W. J., van Wessem, M., Velicogna, I., Wahr, J., Wendt, A., Wouters, B., & Zwally, H. J. (2022). Greenland mass trends from airborne and satellite altimetry during 2011–2020. Journal of Geophysical Research: Earth Surface, 127(3). https://doi.org/10.1029/2021JF006505\n\n- [[9](https://doi.org/10.1038/s41586-019-1855-2)] The IMBIE Team (2019). Mass balance of the Greenland Ice Sheet from 1992 to 2018. Nature, 579, 233–239. https://doi.org/10.1038/s41586-019-1855-2\n\n- [[10](https://doi.org/10.1038/s41586-018-0179-y)] The IMBIE Team (2018). Mass balance of the Antarctic Ice Sheet from 1992 to 2017. Nature 558, 219–222. https://doi.org/10.1038/s41586-018-0179-y"} {"chunk_id": "satellite_satellite-ice-sheet-elevation-change_trend-assessment_q01__d66ac5b38876", "report_id": "satellite_satellite-ice-sheet-elevation-change_trend-assessment_q01", "dataset_id": "satellite-ice-sheet-elevation-change", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q01", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments", "title": "Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments", "chunk_index": 0, "token_count": 123, "text_raw": "Production date: 31-05-2025\n\nDataset version: 5.0\n\nProduced by: Yoni Verhaegen and Philippe Huybrechts (Vrije Universiteit Brussel)", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments\"\nDataset: satellite-ice-sheet-elevation-change [CDS]\nAspect: trend-assessment_q01 | Category: Satellite_ECVs\nSection: Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments\n---\nProduction date: 31-05-2025\n\nDataset version: 5.0\n\nProduced by: Yoni Verhaegen and Philippe Huybrechts (Vrije Universiteit Brussel)"} {"chunk_id": "satellite_satellite-ice-sheet-elevation-change_trend-assessment_q01__5e81f5b5d19b", "report_id": "satellite_satellite-ice-sheet-elevation-change_trend-assessment_q01", "dataset_id": "satellite-ice-sheet-elevation-change", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q01", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments > Quality assessment question", "title": "Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments", "chunk_index": 1, "token_count": 489, "text_raw": "* **\"How adequate are the ice sheet surface elevation change data in terms of their uncertainty and how does it affect estimates of ice sheet volume/mass changes, variability and trends?\"**\n\nThe C3S surface elevation change (SEC) product quantifies changes in the surface elevation of ice sheets, providing crucial insights into their response to climate change and changing ice sheet dynamics. Satellite remote sensing is a valueable and practical method for regularly monitoring such surface elevation changes over those large, remote ice sheet areas. SEC data on the Climate Data Store (CDS) are derived from satellite radar altimetry, where the time delay between a transmitted pulse from the radar and its surface echo is converted into distance, adjusted for the satellite's known elevation. Repeated measurements of the surface elevation from multiple satellite missions, combined with corrections and interpolation, therefore allows to create a consistent time series of gridded ice sheet surface elevation changes [[1](https://doi.org/10.1016/j.epsl.2018.05.015), [2](https://doi.org/10.5194/tc-13-427-2019)].\n\nWhile these techniques offer extensive spatial and temporal coverage, they also have limitations, such as inconsistencies due to data aggregation from different satellite sensors and missions, and difficulties with data acquisition over complex terrain. In that regard, this notebook investigates how well the dataset on the CDS (here we use version 5.0) can be used to monitor the spatial distribution of Greenland Ice Sheet (GrIS) surface elevation changes over the last several decades, as well as to derive reliable values for the (inter/intra-annual) variability and temporal trends of the ice sheet surface elevation, volume and mass changes. More specifically, the notebook evaluates whether the dataset is of sufficient maturity and quality for that purpose in terms of its spatio-temporal uncertainty.\n\n🚨 **Although this notebook deals with the Greenland Ice Sheet only, surface elevation change data for the Antarctic Ice Sheet are also available.**", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments\"\nDataset: satellite-ice-sheet-elevation-change [CDS]\nAspect: trend-assessment_q01 | Category: Satellite_ECVs\nSection: Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments > Quality assessment question\n---\n* **\"How adequate are the ice sheet surface elevation change data in terms of their uncertainty and how does it affect estimates of ice sheet volume/mass changes, variability and trends?\"**\n\nThe C3S surface elevation change (SEC) product quantifies changes in the surface elevation of ice sheets, providing crucial insights into their response to climate change and changing ice sheet dynamics. Satellite remote sensing is a valueable and practical method for regularly monitoring such surface elevation changes over those large, remote ice sheet areas. SEC data on the Climate Data Store (CDS) are derived from satellite radar altimetry, where the time delay between a transmitted pulse from the radar and its surface echo is converted into distance, adjusted for the satellite's known elevation. Repeated measurements of the surface elevation from multiple satellite missions, combined with corrections and interpolation, therefore allows to create a consistent time series of gridded ice sheet surface elevation changes [[1](https://doi.org/10.1016/j.epsl.2018.05.015), [2](https://doi.org/10.5194/tc-13-427-2019)].\n\nWhile these techniques offer extensive spatial and temporal coverage, they also have limitations, such as inconsistencies due to data aggregation from different satellite sensors and missions, and difficulties with data acquisition over complex terrain. In that regard, this notebook investigates how well the dataset on the CDS (here we use version 5.0) can be used to monitor the spatial distribution of Greenland Ice Sheet (GrIS) surface elevation changes over the last several decades, as well as to derive reliable values for the (inter/intra-annual) variability and temporal trends of the ice sheet surface elevation, volume and mass changes. More specifically, the notebook evaluates whether the dataset is of sufficient maturity and quality for that purpose in terms of its spatio-temporal uncertainty.\n\n🚨 **Although this notebook deals with the Greenland Ice Sheet only, surface elevation change data for the Antarctic Ice Sheet are also available.**"} {"chunk_id": "satellite_satellite-ice-sheet-elevation-change_trend-assessment_q01__8db9f2bdbf89", "report_id": "satellite_satellite-ice-sheet-elevation-change_trend-assessment_q01", "dataset_id": "satellite-ice-sheet-elevation-change", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q01", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments > Quality assessment statements", "title": "Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments", "chunk_index": 2, "token_count": 424, "text_raw": "These are the key outcomes of this assessment\n\n- Surface elevation change (SEC) detection by radar altimetry is a valuable tool for assessing the impacts of climate change on ice sheets, but has notable (dis)advantages that users should be aware of. The C3S Greenland Ice Sheet (GrIS) surface elevation change (SEC) dataset consists of mature and quality-rich data, covering a time period of over 30 years with a consistent monthly temporal resolution, filled-up data gaps, and meeting minimum international proposed error thresholds for nearly all pixels. This makes it suitable for detecting reliable climate change signals, trends, and variability at local (pixel) to ice sheet-wide scales. Radar altimeters generally provide more robust surface elevation change measurements in the flat, stable regions of the GrIS compared to the coastal areas, which feature complex terrain. Error estimates therefore tend to be slightly higher around the margins. \n- Users should note that SEC cannot be readily converted to volume nor mass changes of the \"actual\" ice sheet, because converting SEC to volume and mass changes requires adjustments for bedrock uplift, firn densification, and material density differences. Users should also note that biases remain, where radar altimetry tends to underestimate surface lowering rates compared to laser altimetry. This underestimation is aided by the coarse spatial resolution (25 km) of the C3S product, which does not meet minimum thresholds and prevents capturing very localized high mass losses. Despite a favorable error characterization and spatial/temporal coverage, users of the C3S GrIS SEC data product are most likely required to implement additional processing to derive glaciologically interpretable volume/mass changes from the surface elevation changes.\n```", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments\"\nDataset: satellite-ice-sheet-elevation-change [CDS]\nAspect: trend-assessment_q01 | Category: Satellite_ECVs\nSection: Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments > Quality assessment statements\n---\nThese are the key outcomes of this assessment\n\n- Surface elevation change (SEC) detection by radar altimetry is a valuable tool for assessing the impacts of climate change on ice sheets, but has notable (dis)advantages that users should be aware of. The C3S Greenland Ice Sheet (GrIS) surface elevation change (SEC) dataset consists of mature and quality-rich data, covering a time period of over 30 years with a consistent monthly temporal resolution, filled-up data gaps, and meeting minimum international proposed error thresholds for nearly all pixels. This makes it suitable for detecting reliable climate change signals, trends, and variability at local (pixel) to ice sheet-wide scales. Radar altimeters generally provide more robust surface elevation change measurements in the flat, stable regions of the GrIS compared to the coastal areas, which feature complex terrain. Error estimates therefore tend to be slightly higher around the margins. \n- Users should note that SEC cannot be readily converted to volume nor mass changes of the \"actual\" ice sheet, because converting SEC to volume and mass changes requires adjustments for bedrock uplift, firn densification, and material density differences. Users should also note that biases remain, where radar altimetry tends to underestimate surface lowering rates compared to laser altimetry. This underestimation is aided by the coarse spatial resolution (25 km) of the C3S product, which does not meet minimum thresholds and prevents capturing very localized high mass losses. Despite a favorable error characterization and spatial/temporal coverage, users of the C3S GrIS SEC data product are most likely required to implement additional processing to derive glaciologically interpretable volume/mass changes from the surface elevation changes.\n```"} {"chunk_id": "satellite_satellite-ice-sheet-elevation-change_trend-assessment_q01__d1e4c685159a", "report_id": "satellite_satellite-ice-sheet-elevation-change_trend-assessment_q01", "dataset_id": "satellite-ice-sheet-elevation-change", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q01", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments > Methodology > Dataset description", "title": "Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments", "chunk_index": 3, "token_count": 1036, "text_raw": "Surface elevation change detection by satellite radar altimetry is a useful tool to grasp the impact of climate change on the ice sheets. In that regard, the C3S dataset on the Climate Data Store (CDS) provides monthly surface elevation change (SEC) values and their uncertainty for the Greenland (GrIS) and Antarctic Ice Sheet (AIS) on a 25 km spatial resolution grid. The core principle involves measuring surface elevations at different times and comparing them to detect changes. In this dataset, they are derived using satellite radar altimetry that contain data from multiple satellite missions, which are grouped together into a consistent time series for each pixel. Surface elevation changes are reported with units of meter per year and are available since 1992 at monthly-spaced intervals. Data are provided in NetCDF format as gridded data and are available for both the GrIS (excluding peripheral glaciers and ice caps) and AIS (including ice shelves).\n\nThese ice sheet surface elevation change rates are mathematically expressed as:\n\n$\\dfrac{dh}{dt} = \\dfrac{h_{t_2}-h_{t_1}}{t_{2}-t_{1}}$\n\nwhere:\n- $h$ is the surface elevation (m)\n- $t$ the time (yr)\n\nThe time for a measurement mentioned in the dataset is the center of a 3-year of 5-year moving window (for the GrIS) and a 5-year moving window (for the AIS) used to derive the SEC values. Users should hereby not that not only ablation or accumulation (i.e. the surface mass balance) can cause the surface elevation to increase or decrease. The surface elevation change rate at a certain pixel contains signals from various processes:\n\n$\\dfrac{dh}{dt}_{obs} = \\dfrac{SMB}{\\rho_m} + \\dfrac{BMB}{\\rho_i} + \\dfrac{D'}{\\rho_i} + w_C + w_B$\n\nwhere the first term on the right-hand side ($SMB$ in units kg m$^{-2}$ yr$^{-1}$) includes elevation changes from surface mass balance processes (i.e. climatically-driven accumulation and melt/runoff), the second one ($BMB$ in units kg m$^{-2}$ yr$^{-1}$) due to basal mass balance processes at the ice-bedrock interface, the third one ($D'$ in units kg m$^{-2}$ yr$^{-1}$) due to ice dynamic processes (e.g. dynamical thinning or thickening), the fourth one ($w_C$) is the vertical velocity due to snow and/or firn compaction (i.e. the transformation and densification of snow towards firn and in a later stage ice), and the last one ($w_B$) due to bedrock elevation changes (e.g. glacial isostatic adjustment) [[8](https://doi.org/10.5194/tc-5-173-2011), [9](https://doi.org/10.1029/2021JF006505)]. The term $\\rho_m$ is the density of the material lost or gained (snow, firn or ice) and $\\rho_i$ that of ice. The two latter terms do not contribute to actual mass changes of the ice sheet, while ice dynamics is only relevant with respect to mass changes when ice discharge across the grounding line is considered. For the GrIS as a whole, contributions to mass changes are currently driven by both SMB and ice flow dynamics at an approximately equal magnitude (i.e. 60% SMB and 40% ice dynamics), while for the AIS they are almost entirely driven by ice flow dynamics [[1](https://doi.org/10.1016/j.epsl.2018.05.015), [2](https://doi.org/10.5194/tc-13-427-2019), [10](https://doi.org/10.1038/s41586-019-1855-2), [11](https://doi.org/10.1038/s41586-018-0179-y)].\n\nIn this notebook, we use version 5.0. For a more detailed description of the data acquisition and processing methods, we refer to the [documentation on the CDS](https://cds.climate.copernicus.eu/datasets/satellite-ice-sheet-elevation-change?tab=overview) and the [ECMWF Confluence Wiki](https://confluence.ecmwf.int/pages/viewpage.action?pageId=355345393) (Copernicus Knowledge Base).", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments\"\nDataset: satellite-ice-sheet-elevation-change [CDS]\nAspect: trend-assessment_q01 | Category: Satellite_ECVs\nSection: Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments > Methodology > Dataset description\n---\nSurface elevation change detection by satellite radar altimetry is a useful tool to grasp the impact of climate change on the ice sheets. In that regard, the C3S dataset on the Climate Data Store (CDS) provides monthly surface elevation change (SEC) values and their uncertainty for the Greenland (GrIS) and Antarctic Ice Sheet (AIS) on a 25 km spatial resolution grid. The core principle involves measuring surface elevations at different times and comparing them to detect changes. In this dataset, they are derived using satellite radar altimetry that contain data from multiple satellite missions, which are grouped together into a consistent time series for each pixel. Surface elevation changes are reported with units of meter per year and are available since 1992 at monthly-spaced intervals. Data are provided in NetCDF format as gridded data and are available for both the GrIS (excluding peripheral glaciers and ice caps) and AIS (including ice shelves).\n\nThese ice sheet surface elevation change rates are mathematically expressed as:\n\n$\\dfrac{dh}{dt} = \\dfrac{h_{t_2}-h_{t_1}}{t_{2}-t_{1}}$\n\nwhere:\n- $h$ is the surface elevation (m)\n- $t$ the time (yr)\n\nThe time for a measurement mentioned in the dataset is the center of a 3-year of 5-year moving window (for the GrIS) and a 5-year moving window (for the AIS) used to derive the SEC values. Users should hereby not that not only ablation or accumulation (i.e. the surface mass balance) can cause the surface elevation to increase or decrease. The surface elevation change rate at a certain pixel contains signals from various processes:\n\n$\\dfrac{dh}{dt}_{obs} = \\dfrac{SMB}{\\rho_m} + \\dfrac{BMB}{\\rho_i} + \\dfrac{D'}{\\rho_i} + w_C + w_B$\n\nwhere the first term on the right-hand side ($SMB$ in units kg m$^{-2}$ yr$^{-1}$) includes elevation changes from surface mass balance processes (i.e. climatically-driven accumulation and melt/runoff), the second one ($BMB$ in units kg m$^{-2}$ yr$^{-1}$) due to basal mass balance processes at the ice-bedrock interface, the third one ($D'$ in units kg m$^{-2}$ yr$^{-1}$) due to ice dynamic processes (e.g. dynamical thinning or thickening), the fourth one ($w_C$) is the vertical velocity due to snow and/or firn compaction (i.e. the transformation and densification of snow towards firn and in a later stage ice), and the last one ($w_B$) due to bedrock elevation changes (e.g. glacial isostatic adjustment) [[8](https://doi.org/10.5194/tc-5-173-2011), [9](https://doi.org/10.1029/2021JF006505)]. The term $\\rho_m$ is the density of the material lost or gained (snow, firn or ice) and $\\rho_i$ that of ice. The two latter terms do not contribute to actual mass changes of the ice sheet, while ice dynamics is only relevant with respect to mass changes when ice discharge across the grounding line is considered. For the GrIS as a whole, contributions to mass changes are currently driven by both SMB and ice flow dynamics at an approximately equal magnitude (i.e. 60% SMB and 40% ice dynamics), while for the AIS they are almost entirely driven by ice flow dynamics [[1](https://doi.org/10.1016/j.epsl.2018.05.015), [2](https://doi.org/10.5194/tc-13-427-2019), [10](https://doi.org/10.1038/s41586-019-1855-2), [11](https://doi.org/10.1038/s41586-018-0179-y)].\n\nIn this notebook, we use version 5.0. For a more detailed description of the data acquisition and processing methods, we refer to the [documentation on the CDS](https://cds.climate.copernicus.eu/datasets/satellite-ice-sheet-elevation-change?tab=overview) and the [ECMWF Confluence Wiki](https://confluence.ecmwf.int/pages/viewpage.action?pageId=355345393) (Copernicus Knowledge Base)."} {"chunk_id": "satellite_satellite-ice-sheet-elevation-change_trend-assessment_q01__96d28a27ceeb", "report_id": "satellite_satellite-ice-sheet-elevation-change_trend-assessment_q01", "dataset_id": "satellite-ice-sheet-elevation-change", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q01", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments > Methodology > Structure and (sub)sections", "title": "Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments", "chunk_index": 4, "token_count": 211, "text_raw": "**[](section-1)**\n* [](section-1-1)\n* [](section-1-2)\n* [](section-1-3)\n* [](section-1-4)\n\n**[](section-2)**\n* [](section-2-1)\n* [](section-2-2)\n\n**[](section-3)**\n* [](section-3-1)\n* [](section-3-2)\n* [](section-3-3)\n\n**[](section-4)**\n* [](section-4-1)\n* [](section-4-2)\n\n**[](section-5)**", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments\"\nDataset: satellite-ice-sheet-elevation-change [CDS]\nAspect: trend-assessment_q01 | Category: Satellite_ECVs\nSection: Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments > Methodology > Structure and (sub)sections\n---\n**[](section-1)**\n* [](section-1-1)\n* [](section-1-2)\n* [](section-1-3)\n* [](section-1-4)\n\n**[](section-2)**\n* [](section-2-1)\n* [](section-2-2)\n\n**[](section-3)**\n* [](section-3-1)\n* [](section-3-2)\n* [](section-3-3)\n\n**[](section-4)**\n* [](section-4-1)\n* [](section-4-2)\n\n**[](section-5)**"} {"chunk_id": "satellite_satellite-ice-sheet-elevation-change_trend-assessment_q01__29b039413baa", "report_id": "satellite_satellite-ice-sheet-elevation-change_trend-assessment_q01", "dataset_id": "satellite-ice-sheet-elevation-change", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q01", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments > Analysis and results > 1. Data preparation and processing > 1.2 Define request and download", "title": "Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments", "chunk_index": 5, "token_count": 198, "text_raw": "Then we define requests for download from the CDS and download the ice sheet surface elevation change data.\n\nSelect the domain: \"greenland\" or \"antarctica\"\nDefine the request\nDownload the data\n\n```text\ndomain='greenland'\n```\n\n```text\n100%|██████████| 1/1 [00:00<00:00, 6.93it/s]\n```\n\n```text\nDownload completed.\n```\n\n(section-1-3)=", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments\"\nDataset: satellite-ice-sheet-elevation-change [CDS]\nAspect: trend-assessment_q01 | Category: Satellite_ECVs\nSection: Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments > Analysis and results > 1. Data preparation and processing > 1.2 Define request and download\n---\nThen we define requests for download from the CDS and download the ice sheet surface elevation change data.\n\nSelect the domain: \"greenland\" or \"antarctica\"\nDefine the request\nDownload the data\n\n```text\ndomain='greenland'\n```\n\n```text\n100%|██████████| 1/1 [00:00<00:00, 6.93it/s]\n```\n\n```text\nDownload completed.\n```\n\n(section-1-3)="} {"chunk_id": "satellite_satellite-ice-sheet-elevation-change_trend-assessment_q01__c55c97e55d06", "report_id": "satellite_satellite-ice-sheet-elevation-change_trend-assessment_q01", "dataset_id": "satellite-ice-sheet-elevation-change", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q01", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments > Analysis and results > 1. Data preparation and processing > 1.3 Display and inspect data", "title": "Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments", "chunk_index": 6, "token_count": 1081, "text_raw": "We can read and inspect the data. Let us print out the data to inspect its structure:\n\n```text\n{'greenland': Size: 76MB\n Dimensions: (time: 378, y: 123, x: 65)\n Coordinates:\n * time (time) datetime64[ns] 3kB 1992-01-01 ... 2023-06-01\n * x (x) float32 260B -7.393e+05 -7.143e+05 ... 8.607e+05\n * y (y) float32 492B -3.478e+06 -3.453e+06 ... -4.281e+05\n Data variables: (12/15)\n start_time (time) datetime64[ns] 3kB 1991-08-01 ... 2014-08-01\n end_time (time) datetime64[ns] 3kB 1996-10-31 ... 2023-07-31\n grid_projection |S1 1B b''\n lat (y, x) float32 32kB 58.0 58.04 58.09 ... 81.55 81.35 81.14\n lon (y, x) float32 32kB -57.0 -56.61 -56.21 ... 17.87 18.55\n dh (y, x, time) float32 12MB nan nan nan nan ... nan nan nan\n ... ...\n dhdt_stabil (y, x, time) float32 12MB nan nan nan nan ... nan nan nan\n dhdt_ok (y, x, time) int8 3MB 0 0 0 0 0 0 0 0 0 ... 0 0 0 0 0 0 0 0\n dist (y, x, time) float32 12MB nan nan nan nan ... nan nan nan\n land_mask (y, x) int8 8kB 0 0 0 0 0 0 0 0 0 0 ... 0 0 0 0 0 0 0 0 0 0\n high_slope (y, x) int8 8kB 0 0 0 0 0 0 0 0 0 0 ... 0 0 0 0 0 0 0 0 0 0\n area (y, x) float32 32kB nan nan nan nan nan ... nan nan nan nan\n Attributes: (12/32)\n Title: Surface Elevation change of the Greenland ice sheet...\n institution: Copernicus Climate Change Service, DTU Space - Div....\n reference: Simonsen and Sørensen (2017), Sørensen et al. (2018)\n contact: copernicus-support@ecmwf.int\n file_creation_date: 2023-12-15 16:56:53.584072\n project: C3S_312b_Lot4_ice_sheets_and_shelves\n ... ...\n netCDF_version: NETCDF4\n product_version: v5\n Conventions: CF-1.7\n keywords: EARTH SCIENCE CRYOSPHERE GLACIERS/ICE SHEETS/GLACIE...\n license: C3S general license\n summary: Surface elevation change rate derived for Greenland...}\n```\n\nThe version 5.0 is a gridded dataset at a 25 km spatial resolution containing monthly values of the ice sheet surface elevation change rate $\\frac{dh}{dt}$ (`dhdt` in m/yr) and its uncertainty (`dhdt_uncert` in m/yr), as well as the surface elevation change between two measurements $dh$ (`dh` in m, and hence not normalized to \"per year\") of a grid cell since 1992. The uncertainties are here reported as 1-sigma precision errors or standard deviations. A land mask (`land_mask`), slope mask (`high_slope`), validity flags (`dhdt_ok`) and the surface area of a grid cell (`area`) are also included. The time for a measurement mentioned in the dataset is the center of a 3-year or 5-year moving window used to derive the surface elevation change values.\n\nLet us check the total temporal extent of the data:", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments\"\nDataset: satellite-ice-sheet-elevation-change [CDS]\nAspect: trend-assessment_q01 | Category: Satellite_ECVs\nSection: Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments > Analysis and results > 1. Data preparation and processing > 1.3 Display and inspect data\n---\nWe can read and inspect the data. Let us print out the data to inspect its structure:\n\n```text\n{'greenland': Size: 76MB\n Dimensions: (time: 378, y: 123, x: 65)\n Coordinates:\n * time (time) datetime64[ns] 3kB 1992-01-01 ... 2023-06-01\n * x (x) float32 260B -7.393e+05 -7.143e+05 ... 8.607e+05\n * y (y) float32 492B -3.478e+06 -3.453e+06 ... -4.281e+05\n Data variables: (12/15)\n start_time (time) datetime64[ns] 3kB 1991-08-01 ... 2014-08-01\n end_time (time) datetime64[ns] 3kB 1996-10-31 ... 2023-07-31\n grid_projection |S1 1B b''\n lat (y, x) float32 32kB 58.0 58.04 58.09 ... 81.55 81.35 81.14\n lon (y, x) float32 32kB -57.0 -56.61 -56.21 ... 17.87 18.55\n dh (y, x, time) float32 12MB nan nan nan nan ... nan nan nan\n ... ...\n dhdt_stabil (y, x, time) float32 12MB nan nan nan nan ... nan nan nan\n dhdt_ok (y, x, time) int8 3MB 0 0 0 0 0 0 0 0 0 ... 0 0 0 0 0 0 0 0\n dist (y, x, time) float32 12MB nan nan nan nan ... nan nan nan\n land_mask (y, x) int8 8kB 0 0 0 0 0 0 0 0 0 0 ... 0 0 0 0 0 0 0 0 0 0\n high_slope (y, x) int8 8kB 0 0 0 0 0 0 0 0 0 0 ... 0 0 0 0 0 0 0 0 0 0\n area (y, x) float32 32kB nan nan nan nan nan ... nan nan nan nan\n Attributes: (12/32)\n Title: Surface Elevation change of the Greenland ice sheet...\n institution: Copernicus Climate Change Service, DTU Space - Div....\n reference: Simonsen and Sørensen (2017), Sørensen et al. (2018)\n contact: copernicus-support@ecmwf.int\n file_creation_date: 2023-12-15 16:56:53.584072\n project: C3S_312b_Lot4_ice_sheets_and_shelves\n ... ...\n netCDF_version: NETCDF4\n product_version: v5\n Conventions: CF-1.7\n keywords: EARTH SCIENCE CRYOSPHERE GLACIERS/ICE SHEETS/GLACIE...\n license: C3S general license\n summary: Surface elevation change rate derived for Greenland...}\n```\n\nThe version 5.0 is a gridded dataset at a 25 km spatial resolution containing monthly values of the ice sheet surface elevation change rate $\\frac{dh}{dt}$ (`dhdt` in m/yr) and its uncertainty (`dhdt_uncert` in m/yr), as well as the surface elevation change between two measurements $dh$ (`dh` in m, and hence not normalized to \"per year\") of a grid cell since 1992. The uncertainties are here reported as 1-sigma precision errors or standard deviations. A land mask (`land_mask`), slope mask (`high_slope`), validity flags (`dhdt_ok`) and the surface area of a grid cell (`area`) are also included. The time for a measurement mentioned in the dataset is the center of a 3-year or 5-year moving window used to derive the surface elevation change values.\n\nLet us check the total temporal extent of the data:"} {"chunk_id": "satellite_satellite-ice-sheet-elevation-change_trend-assessment_q01__0afe530944de", "report_id": "satellite_satellite-ice-sheet-elevation-change_trend-assessment_q01", "dataset_id": "satellite-ice-sheet-elevation-change", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q01", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments > Analysis and results > 1. Data preparation and processing > 1.3 Display and inspect data", "title": "Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments", "chunk_index": 7, "token_count": 199, "text_raw": "area`) are also included. The time for a measurement mentioned in the dataset is the center of a 3-year or 5-year moving window used to derive the surface elevation change values.\n\nLet us check the total temporal extent of the data:\n\n```text\nThe begin period of the dataset is 1992-01-01 and the end period is 2023-06-01, which is a total of time 31.41 years.\n```\n\n(section-1-4)=", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments\"\nDataset: satellite-ice-sheet-elevation-change [CDS]\nAspect: trend-assessment_q01 | Category: Satellite_ECVs\nSection: Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments > Analysis and results > 1. Data preparation and processing > 1.3 Display and inspect data\n---\narea`) are also included. The time for a measurement mentioned in the dataset is the center of a 3-year or 5-year moving window used to derive the surface elevation change values.\n\nLet us check the total temporal extent of the data:\n\n```text\nThe begin period of the dataset is 1992-01-01 and the end period is 2023-06-01, which is a total of time 31.41 years.\n```\n\n(section-1-4)="} {"chunk_id": "satellite_satellite-ice-sheet-elevation-change_trend-assessment_q01__d87196181c0b", "report_id": "satellite_satellite-ice-sheet-elevation-change_trend-assessment_q01", "dataset_id": "satellite-ice-sheet-elevation-change", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q01", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments > Analysis and results > 1. Data preparation and processing > 1.4 Data handling and creating functions", "title": "Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments", "chunk_index": 8, "token_count": 269, "text_raw": "Let us now perform some data handling and define a plotting function before getting started with the analysis:\n\nSelect the specific dataset you want to process, e.g., \"greenland\"\nDefine plotting function\nCreate subplots with Polar Stereographic projection\nPlot the data\nSet extent and plot features\nAdd colorbar\nDefine the function for plotting a time series\nApply mean to all variables except 'dvol', which will use sum\nCombine the results\n\nNow, our dataset array only holds the most important information, such as the multiyear mean surface elevation change rates (`sec`) and the arithmetic mean precision error (`sec_err`). With the function above, we also calculated linear and quadratic trends of surface elevation changes, as well as the amount of missing values.\n\nWith everything ready, let us now begin with the analysis:\n\n(section-2)=", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments\"\nDataset: satellite-ice-sheet-elevation-change [CDS]\nAspect: trend-assessment_q01 | Category: Satellite_ECVs\nSection: Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments > Analysis and results > 1. Data preparation and processing > 1.4 Data handling and creating functions\n---\nLet us now perform some data handling and define a plotting function before getting started with the analysis:\n\nSelect the specific dataset you want to process, e.g., \"greenland\"\nDefine plotting function\nCreate subplots with Polar Stereographic projection\nPlot the data\nSet extent and plot features\nAdd colorbar\nDefine the function for plotting a time series\nApply mean to all variables except 'dvol', which will use sum\nCombine the results\n\nNow, our dataset array only holds the most important information, such as the multiyear mean surface elevation change rates (`sec`) and the arithmetic mean precision error (`sec_err`). With the function above, we also calculated linear and quadratic trends of surface elevation changes, as well as the amount of missing values.\n\nWith everything ready, let us now begin with the analysis:\n\n(section-2)="} {"chunk_id": "satellite_satellite-ice-sheet-elevation-change_trend-assessment_q01__09af134a29c5", "report_id": "satellite_satellite-ice-sheet-elevation-change_trend-assessment_q01", "dataset_id": "satellite-ice-sheet-elevation-change", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q01", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments > Analysis and results > 2. Greenland Ice Sheet surface elevation and altimetric volume changes in space and time > 2.1 Multiyear mean surface elevation changes", "title": "Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments", "chunk_index": 9, "token_count": 481, "text_raw": "We begin by plotting the Greenland multiyear mean surface elevation change $\\overline {\\frac{dh}{dt}}$ between the beginning and end period with the defined plotting function:\n\nApply the function to the surface elevation change rate data\nDefine dataset to be plotted\n\n*Figure 1. Multiyear mean surface elevation change (SEC) over Greenland from the SEC dataset on the Climate Data Store.*\n\nOverall, the map provides a visual representation of the changes of the surface elevation of Greenland's ice sheet over the last several decades. The red regions, particularly around the margins, indicate areas of significant surface lowering, which is consistent with observations of accelerated glacier flow and ice thinning in these regions [[1](https://doi.org/10.1016/j.epsl.2018.05.015), [3](https://doi.org/10.1016/j.rse.2016.12.012)].\n\nLet us quantify the ice sheet-wide average value:\n\n```text\nThe Greenland ice sheet-wide average surface elevation change rate value between 1992-01-01 and 2023-06-01 is -0.04856 m/yr.\n```\n\nThe negative value indicates that the surface of the ice sheet, in general, has been lowering during the last several decades, resulting in net negative values. This is consistent with overall trends in the literature [[1](https://doi.org/10.1016/j.epsl.2018.05.015), [5](https://doi.org/10.1029/2020GL091216), [8](https://doi.org/10.5194/tc-5-173-2011), [10](https://doi.org/10.1038/s41586-019-1855-2)].\n\n(section-2-2)=", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments\"\nDataset: satellite-ice-sheet-elevation-change [CDS]\nAspect: trend-assessment_q01 | Category: Satellite_ECVs\nSection: Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments > Analysis and results > 2. Greenland Ice Sheet surface elevation and altimetric volume changes in space and time > 2.1 Multiyear mean surface elevation changes\n---\nWe begin by plotting the Greenland multiyear mean surface elevation change $\\overline {\\frac{dh}{dt}}$ between the beginning and end period with the defined plotting function:\n\nApply the function to the surface elevation change rate data\nDefine dataset to be plotted\n\n*Figure 1. Multiyear mean surface elevation change (SEC) over Greenland from the SEC dataset on the Climate Data Store.*\n\nOverall, the map provides a visual representation of the changes of the surface elevation of Greenland's ice sheet over the last several decades. The red regions, particularly around the margins, indicate areas of significant surface lowering, which is consistent with observations of accelerated glacier flow and ice thinning in these regions [[1](https://doi.org/10.1016/j.epsl.2018.05.015), [3](https://doi.org/10.1016/j.rse.2016.12.012)].\n\nLet us quantify the ice sheet-wide average value:\n\n```text\nThe Greenland ice sheet-wide average surface elevation change rate value between 1992-01-01 and 2023-06-01 is -0.04856 m/yr.\n```\n\nThe negative value indicates that the surface of the ice sheet, in general, has been lowering during the last several decades, resulting in net negative values. This is consistent with overall trends in the literature [[1](https://doi.org/10.1016/j.epsl.2018.05.015), [5](https://doi.org/10.1029/2020GL091216), [8](https://doi.org/10.5194/tc-5-173-2011), [10](https://doi.org/10.1038/s41586-019-1855-2)].\n\n(section-2-2)="} {"chunk_id": "satellite_satellite-ice-sheet-elevation-change_trend-assessment_q01__6836b89ce817", "report_id": "satellite_satellite-ice-sheet-elevation-change_trend-assessment_q01", "dataset_id": "satellite-ice-sheet-elevation-change", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q01", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments > Analysis and results > 2. Greenland Ice Sheet surface elevation and altimetric volume changes in space and time > 2.2 Conversion into altimetric and \"actual\" volume changes", "title": "Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments", "chunk_index": 10, "token_count": 1000, "text_raw": "We can also convert the surface elevation change values to cumulative altimetric volume changes $V_a$ (with subscript 'a' meaning altimetric) by combining the variables `dh` and `area`. We therefore have to make the assumption that surface elevation changes due to bedrock uplift and firn compaction rates are unsignificant or unknown:\n\n$dV_{a,i} \n$\n[m$^{3}$]\n$ = \\sum\\limits^{x,y}{dV_{x,y,i}} = \\sum\\limits^{x,y}(d{h_{x,y,i}}*A_{x,y})$\n\nwhere $d{h_{x,y}}$ is the height change between two measurement intervals $i$ (from `dh`, in m) and $A_{x,y}$ the surface area (from `area`, in m$^2$) of pixel $x,y$ .\n\nWe then take the cumulative value over time to get the cumulative altimetric volume change:\n\n$\n{V_a} \n$\n[km$^{3}$]\n$\n= \\sum\\limits_{i={1}}^{{{n}}} \\left(\\dfrac{dV_{a,i}}{1 \\cdot 10^9}\\right)\n$\n\nwith $dV_{a,i}$ the ice sheet-wide altimetric volume change (in m$^{3}$) between two measurements (as calculated above) and $n$ the number of temporal intervals in the time series.\n\nLet us plot this as a time series:\n\nGiven datasets_original is the dataset dictionary\nExtract the 'dh' and 'area' variables\nSum over the x and y dimensions and convert to km^3\nPlot the resulting time series\n\n*Figure 2. Cumulative altimetric volume change of the Greenland ice sheet from the SEC dataset on the Climate Data Store.*\n\nAs expected, the plot shows a downward pattern of the altimetric volume with a clear trend towards volume losses over time, especially during the more recent years. However, the curve should be interpreted with care. Greenland’s bedrock is namely uplifting currently, due to slow mantle-deformation processes and elastic processes associated with the ongoing ice loss today. Especially around the margins, where ice loss is currently highest, bedrock uplift rates are in the order of several mm per year [[9](https://doi.org/10.1029/2021JF006505)]. Effects from near-surface firn density changes are even larger (i.e. up to a few cm per year) but also uncertain [[4](https://doi.org/10.3189/172756505781829007), [9](https://doi.org/10.1029/2021JF006505)]. In order to obtain \"actual\" ice sheet volume changes, corrections for firn densification (which can lower the surface without actual mass loss) and bedrock uplift (which does not impact ice thickness changes) are required [[4](https://doi.org/10.3189/172756505781829007), [5](https://doi.org/10.1029/2020GL091216), [8](https://doi.org/10.5194/tc-5-173-2011), [9](https://doi.org/10.1029/2021JF006505)]. As such, to derive pixel-by-pixel ice thickness changes $dH/dt$ from surface elevation changes $dh/dt$ over grounded ice, a correction is needed:\n\n$\\dfrac{dH}{dt} \n$\n[m yr$^{-1}$]\n$= \\dfrac{dh}{dt} - w_C - w_B$\n\nwhich can then be translated into actual volume changes $dV/dt$ by using the surface area $A$ of a grid cell:\n\n$\\dfrac{dV}{dt} \n$\n[m$^3$ yr$^{-1}$]\n$\n= A\\dfrac{dH}{dt}$\n\nThen, mass changes can be obtained by:\n\n$\\dfrac{dM}{dt} \n$\n[kg yr$^{-1}$]\n$\n= \\rho_m \\dfrac{dV}{dt}$\n\nwhere the value of the density $\\rho_m$ depends on the material/physical process involved in the mass change.", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments\"\nDataset: satellite-ice-sheet-elevation-change [CDS]\nAspect: trend-assessment_q01 | Category: Satellite_ECVs\nSection: Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments > Analysis and results > 2. Greenland Ice Sheet surface elevation and altimetric volume changes in space and time > 2.2 Conversion into altimetric and \"actual\" volume changes\n---\nWe can also convert the surface elevation change values to cumulative altimetric volume changes $V_a$ (with subscript 'a' meaning altimetric) by combining the variables `dh` and `area`. We therefore have to make the assumption that surface elevation changes due to bedrock uplift and firn compaction rates are unsignificant or unknown:\n\n$dV_{a,i} \n$\n[m$^{3}$]\n$ = \\sum\\limits^{x,y}{dV_{x,y,i}} = \\sum\\limits^{x,y}(d{h_{x,y,i}}*A_{x,y})$\n\nwhere $d{h_{x,y}}$ is the height change between two measurement intervals $i$ (from `dh`, in m) and $A_{x,y}$ the surface area (from `area`, in m$^2$) of pixel $x,y$ .\n\nWe then take the cumulative value over time to get the cumulative altimetric volume change:\n\n$\n{V_a} \n$\n[km$^{3}$]\n$\n= \\sum\\limits_{i={1}}^{{{n}}} \\left(\\dfrac{dV_{a,i}}{1 \\cdot 10^9}\\right)\n$\n\nwith $dV_{a,i}$ the ice sheet-wide altimetric volume change (in m$^{3}$) between two measurements (as calculated above) and $n$ the number of temporal intervals in the time series.\n\nLet us plot this as a time series:\n\nGiven datasets_original is the dataset dictionary\nExtract the 'dh' and 'area' variables\nSum over the x and y dimensions and convert to km^3\nPlot the resulting time series\n\n*Figure 2. Cumulative altimetric volume change of the Greenland ice sheet from the SEC dataset on the Climate Data Store.*\n\nAs expected, the plot shows a downward pattern of the altimetric volume with a clear trend towards volume losses over time, especially during the more recent years. However, the curve should be interpreted with care. Greenland’s bedrock is namely uplifting currently, due to slow mantle-deformation processes and elastic processes associated with the ongoing ice loss today. Especially around the margins, where ice loss is currently highest, bedrock uplift rates are in the order of several mm per year [[9](https://doi.org/10.1029/2021JF006505)]. Effects from near-surface firn density changes are even larger (i.e. up to a few cm per year) but also uncertain [[4](https://doi.org/10.3189/172756505781829007), [9](https://doi.org/10.1029/2021JF006505)]. In order to obtain \"actual\" ice sheet volume changes, corrections for firn densification (which can lower the surface without actual mass loss) and bedrock uplift (which does not impact ice thickness changes) are required [[4](https://doi.org/10.3189/172756505781829007), [5](https://doi.org/10.1029/2020GL091216), [8](https://doi.org/10.5194/tc-5-173-2011), [9](https://doi.org/10.1029/2021JF006505)]. As such, to derive pixel-by-pixel ice thickness changes $dH/dt$ from surface elevation changes $dh/dt$ over grounded ice, a correction is needed:\n\n$\\dfrac{dH}{dt} \n$\n[m yr$^{-1}$]\n$= \\dfrac{dh}{dt} - w_C - w_B$\n\nwhich can then be translated into actual volume changes $dV/dt$ by using the surface area $A$ of a grid cell:\n\n$\\dfrac{dV}{dt} \n$\n[m$^3$ yr$^{-1}$]\n$\n= A\\dfrac{dH}{dt}$\n\nThen, mass changes can be obtained by:\n\n$\\dfrac{dM}{dt} \n$\n[kg yr$^{-1}$]\n$\n= \\rho_m \\dfrac{dV}{dt}$\n\nwhere the value of the density $\\rho_m$ depends on the material/physical process involved in the mass change."} {"chunk_id": "satellite_satellite-ice-sheet-elevation-change_trend-assessment_q01__28e5cee3c5f7", "report_id": "satellite_satellite-ice-sheet-elevation-change_trend-assessment_q01", "dataset_id": "satellite-ice-sheet-elevation-change", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q01", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments > Analysis and results > 2. Greenland Ice Sheet surface elevation and altimetric volume changes in space and time > 2.2 Conversion into altimetric and \"actual\" volume changes", "title": "Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments", "chunk_index": 11, "token_count": 507, "text_raw": "yr$^{-1}$]\n$\n= \\rho_m \\dfrac{dV}{dt}$\n\nwhere the value of the density $\\rho_m$ depends on the material/physical process involved in the mass change.\n\nMoreover, for more significant surface lowering rates (lower than ca. -0.5 to -1 m/yr), there seems to be a bias towards an underestimation of lowering rates by the radar altimetry (as used in the C3S products) when compared to the laser altimeters (see for example the [PQAR](https://confluence.ecmwf.int/pages/viewpage.action?pageId=414590531)). This is presumably related to a variety of factors, including different modes of data acquisition (e.g. varying footprint sizes between radar and laser instruments, a varying penetration depth of the radar pulse into snow), different algorithm applications and settings, or post-processing procedures [[3](https://doi.org/10.1016/j.rse.2016.12.012)]. Additionally, in regions with high surface lowering rates, the coarse resolution of the C3S product (25 km) also tends to aid the underestimation of SEC due to the very localized high mass/volume losses of certain outlet glaciers [[5](https://doi.org/10.1029/2020GL091216)].\n\nIn that sense, the dataset on the Climate Data Store (CDS) seems to exhibit a net underestimation of the Greenland surface elevation and volume change rates. All the above processes thus tend to bias the volume change estimates of the \"actual\" ice sheet, making the curve above less suitable for the determination of actual volume or mass changes of the ice sheet. Users of the C3S GrIS SEC data product are therefore most likely required to implement additional processing to derive glaciologically interpretable mass/volume changes from the provided surface elevation changes.\n\n(section-3)=", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments\"\nDataset: satellite-ice-sheet-elevation-change [CDS]\nAspect: trend-assessment_q01 | Category: Satellite_ECVs\nSection: Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments > Analysis and results > 2. Greenland Ice Sheet surface elevation and altimetric volume changes in space and time > 2.2 Conversion into altimetric and \"actual\" volume changes\n---\nyr$^{-1}$]\n$\n= \\rho_m \\dfrac{dV}{dt}$\n\nwhere the value of the density $\\rho_m$ depends on the material/physical process involved in the mass change.\n\nMoreover, for more significant surface lowering rates (lower than ca. -0.5 to -1 m/yr), there seems to be a bias towards an underestimation of lowering rates by the radar altimetry (as used in the C3S products) when compared to the laser altimeters (see for example the [PQAR](https://confluence.ecmwf.int/pages/viewpage.action?pageId=414590531)). This is presumably related to a variety of factors, including different modes of data acquisition (e.g. varying footprint sizes between radar and laser instruments, a varying penetration depth of the radar pulse into snow), different algorithm applications and settings, or post-processing procedures [[3](https://doi.org/10.1016/j.rse.2016.12.012)]. Additionally, in regions with high surface lowering rates, the coarse resolution of the C3S product (25 km) also tends to aid the underestimation of SEC due to the very localized high mass/volume losses of certain outlet glaciers [[5](https://doi.org/10.1029/2020GL091216)].\n\nIn that sense, the dataset on the Climate Data Store (CDS) seems to exhibit a net underestimation of the Greenland surface elevation and volume change rates. All the above processes thus tend to bias the volume change estimates of the \"actual\" ice sheet, making the curve above less suitable for the determination of actual volume or mass changes of the ice sheet. Users of the C3S GrIS SEC data product are therefore most likely required to implement additional processing to derive glaciologically interpretable mass/volume changes from the provided surface elevation changes.\n\n(section-3)="} {"chunk_id": "satellite_satellite-ice-sheet-elevation-change_trend-assessment_q01__967f3903fb73", "report_id": "satellite_satellite-ice-sheet-elevation-change_trend-assessment_q01", "dataset_id": "satellite-ice-sheet-elevation-change", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q01", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments > Analysis and results > 3. Greenland Ice Sheet surface elevation change error estimates > 3.1 Uncertainty estimates: accuracy and precision", "title": "Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments", "chunk_index": 12, "token_count": 470, "text_raw": "Let us now consider the errors of the data. The total error of a surface elevation change estimate is theoretically given by the sum of the precision (random) and the accuracy (systematic) error:\n\n$\n\\varepsilon = a\\sigma + \\delta\n$\n\nwhere $a$ is the critical z-score related to a certain statistical confidence interval, $\\sigma$ is the random error (i.e. standard deviation) and $\\delta$ the systematic error.\n\nIn the C3S surface elevation change dataset for Greenland, precision errors are reported as the standard deviation (i.e. the 68% confidence interval with $a$ = 1) and the accuracy error is not directly considered in the error estimates provided with the data. Therefore, in our case, $\\delta$ = 0, meaning that $\\varepsilon_{\\frac{dh}{dt}}$ = $\\sigma_{\\frac{dh}{dt}}$.\n\nLet us now explore a bit more the uncertainty of the data. Quantitative pixel-by-pixel error estimates are namely available for the dataset. Let us begin by plotting a histogram of the error values:\n\n*Figure 3. Histogram of surface elevation change error estimates for the Greenland ice sheet from the SEC dataset on the Climate Data Store.*\n\nMost errors (standard deviations) seem to be situated between 0 and 0.01 m/yr. Given that the GCOS (2022) requires a precision error threshold (expressed as 2$\\sigma$) of at least 0.1 m/yr, it can be stated that this requirement is clearly met for the Greenland Ice Sheet [[6](https://library.wmo.int/idurl/4/58111)]. Let us check the spatial and temporal distribution of these errors below.\n\n(section-3-2)=", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments\"\nDataset: satellite-ice-sheet-elevation-change [CDS]\nAspect: trend-assessment_q01 | Category: Satellite_ECVs\nSection: Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments > Analysis and results > 3. Greenland Ice Sheet surface elevation change error estimates > 3.1 Uncertainty estimates: accuracy and precision\n---\nLet us now consider the errors of the data. The total error of a surface elevation change estimate is theoretically given by the sum of the precision (random) and the accuracy (systematic) error:\n\n$\n\\varepsilon = a\\sigma + \\delta\n$\n\nwhere $a$ is the critical z-score related to a certain statistical confidence interval, $\\sigma$ is the random error (i.e. standard deviation) and $\\delta$ the systematic error.\n\nIn the C3S surface elevation change dataset for Greenland, precision errors are reported as the standard deviation (i.e. the 68% confidence interval with $a$ = 1) and the accuracy error is not directly considered in the error estimates provided with the data. Therefore, in our case, $\\delta$ = 0, meaning that $\\varepsilon_{\\frac{dh}{dt}}$ = $\\sigma_{\\frac{dh}{dt}}$.\n\nLet us now explore a bit more the uncertainty of the data. Quantitative pixel-by-pixel error estimates are namely available for the dataset. Let us begin by plotting a histogram of the error values:\n\n*Figure 3. Histogram of surface elevation change error estimates for the Greenland ice sheet from the SEC dataset on the Climate Data Store.*\n\nMost errors (standard deviations) seem to be situated between 0 and 0.01 m/yr. Given that the GCOS (2022) requires a precision error threshold (expressed as 2$\\sigma$) of at least 0.1 m/yr, it can be stated that this requirement is clearly met for the Greenland Ice Sheet [[6](https://library.wmo.int/idurl/4/58111)]. Let us check the spatial and temporal distribution of these errors below.\n\n(section-3-2)="} {"chunk_id": "satellite_satellite-ice-sheet-elevation-change_trend-assessment_q01__54d90b6dac9a", "report_id": "satellite_satellite-ice-sheet-elevation-change_trend-assessment_q01", "dataset_id": "satellite-ice-sheet-elevation-change", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q01", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments > Analysis and results > 3. Greenland Ice Sheet surface elevation change error estimates > 3.2 Surface elevation change errors in space and time", "title": "Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments", "chunk_index": 13, "token_count": 383, "text_raw": "To get a better idea, we can also plot the arithmetic mean values over time for each pixel to check their spatial distribution:\n\nApply the function to the surface elevation change rate error data\nPlot the data\n\n*Figure 4. Spatial distribution of temporarily averaged surface elevation change errors for the Greenland ice sheet from the SEC dataset on the Climate Data Store.*\n\nLet us quantify the ice sheet-wide average value:\n\n```text\nThe Greenland ice sheet-wide average surface elevation change rate error (standard deviation) between 1992-01-01 and 2023-06-01 is 0.008 m/yr.\n```\n\nWe can accordingly plot the data as a time series:\n\n*Figure 5. Time series of the spatially averaged surface elevation change error for the Greenland ice sheet from the SEC dataset on the Climate Data Store.*\n\nThe highest errors are clearly present around the marginal errors of the ice sheet. Moreover, no clear trend in the average error values is present in the time series. However, given that the GCOS (2022) requires a precision error threshold (expressed as 2$\\sigma$) of at least 0.1 m/yr, it can be stated that this requirement is clearly met for the Greenland Ice Sheet [[6](https://library.wmo.int/idurl/4/58111)].\n\n(section-3-3)=", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments\"\nDataset: satellite-ice-sheet-elevation-change [CDS]\nAspect: trend-assessment_q01 | Category: Satellite_ECVs\nSection: Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments > Analysis and results > 3. Greenland Ice Sheet surface elevation change error estimates > 3.2 Surface elevation change errors in space and time\n---\nTo get a better idea, we can also plot the arithmetic mean values over time for each pixel to check their spatial distribution:\n\nApply the function to the surface elevation change rate error data\nPlot the data\n\n*Figure 4. Spatial distribution of temporarily averaged surface elevation change errors for the Greenland ice sheet from the SEC dataset on the Climate Data Store.*\n\nLet us quantify the ice sheet-wide average value:\n\n```text\nThe Greenland ice sheet-wide average surface elevation change rate error (standard deviation) between 1992-01-01 and 2023-06-01 is 0.008 m/yr.\n```\n\nWe can accordingly plot the data as a time series:\n\n*Figure 5. Time series of the spatially averaged surface elevation change error for the Greenland ice sheet from the SEC dataset on the Climate Data Store.*\n\nThe highest errors are clearly present around the marginal errors of the ice sheet. Moreover, no clear trend in the average error values is present in the time series. However, given that the GCOS (2022) requires a precision error threshold (expressed as 2$\\sigma$) of at least 0.1 m/yr, it can be stated that this requirement is clearly met for the Greenland Ice Sheet [[6](https://library.wmo.int/idurl/4/58111)].\n\n(section-3-3)="} {"chunk_id": "satellite_satellite-ice-sheet-elevation-change_trend-assessment_q01__8e98470722bb", "report_id": "satellite_satellite-ice-sheet-elevation-change_trend-assessment_q01", "dataset_id": "satellite-ice-sheet-elevation-change", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q01", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments > Analysis and results > 3. Greenland Ice Sheet surface elevation change error estimates > 3.3 Link with surface characteristics", "title": "Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments", "chunk_index": 14, "token_count": 624, "text_raw": "Let us further investigate this pattern and check whether we can link the amount of missing data to the surface characteristics provided with the data. Therefore, we check how the amount of missing data is linked to the surface slope (i.e. as a proxy for the terrain complexity) and the magnitude of the surface elevation changes at a certain pixel:\n\nFlatten the arrays and remove NaN values for the boxplot\nScatter plot\nRemove NaN values for the scatter plot\n\n*Figure 6. Relationship between the surface elevation change error and (left) the slope of the terrain and (right) the magnitude of the ice sheet surface elevation change over Greenland from the SEC dataset on the Climate Data Store.*\n\nThe results are displayed in Figure 6. The highest errors are clearly present over complex and steep terrain around the marginal regions, which arise due to processing limitations or challenging data acquisition conditions. The first graph shows that regions with gentle slopes (<2°) have the lowest overall error, indicating better radar performance on flat surfaces. As slopes increase to more than 2 and 5°, the errors increase, reflecting the difficulty of capturing reliable measurements due to slope-induced errors. Moreover, the error also increases with the magnitude of the surface elevation changes themselves, as shown by the right plot. Areas with larger surface elevation changes, often found along ice sheet margins with complex terrain that are more prone to melting, tend to have higher error values. This occurs, amongst others, because radar reflections may occur from different snowpack layers (e.g. from liquid water within the snow or from refrozen ice lenses). This indicates that dynamic regions are harder to monitor consistently. In general, radar altimeters thus tend to perform better in the central, flat regions of Greenland due to simpler topography and relatively stable surfaces, while marginal zones with complex and dynamic terrain (where the most significant SEC occur) exhibit an increased uncertainty.\n\nLet us at last quantify the amount of pixels that do not meet the GCOS precision error requirement of 0.1 m/yr. We therefore multiply the errors accompanying the data by 2 to get a precision error in the form of 2 standard deviations (as proposed by GCOS):\n\n```text\nThe absolute number of pixels with a surface elevation change rate precision error (2σ) < 0.1 m/yr is 1025570 pixels or 99.41%.\n```\n\nAs already discussed above, almost all pixels exhibit error values that fall within international proposed thresholds, indicating a robust error characterization of the C3S SEC product for the GrIS.\n\n(section-4)=", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments\"\nDataset: satellite-ice-sheet-elevation-change [CDS]\nAspect: trend-assessment_q01 | Category: Satellite_ECVs\nSection: Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments > Analysis and results > 3. Greenland Ice Sheet surface elevation change error estimates > 3.3 Link with surface characteristics\n---\nLet us further investigate this pattern and check whether we can link the amount of missing data to the surface characteristics provided with the data. Therefore, we check how the amount of missing data is linked to the surface slope (i.e. as a proxy for the terrain complexity) and the magnitude of the surface elevation changes at a certain pixel:\n\nFlatten the arrays and remove NaN values for the boxplot\nScatter plot\nRemove NaN values for the scatter plot\n\n*Figure 6. Relationship between the surface elevation change error and (left) the slope of the terrain and (right) the magnitude of the ice sheet surface elevation change over Greenland from the SEC dataset on the Climate Data Store.*\n\nThe results are displayed in Figure 6. The highest errors are clearly present over complex and steep terrain around the marginal regions, which arise due to processing limitations or challenging data acquisition conditions. The first graph shows that regions with gentle slopes (<2°) have the lowest overall error, indicating better radar performance on flat surfaces. As slopes increase to more than 2 and 5°, the errors increase, reflecting the difficulty of capturing reliable measurements due to slope-induced errors. Moreover, the error also increases with the magnitude of the surface elevation changes themselves, as shown by the right plot. Areas with larger surface elevation changes, often found along ice sheet margins with complex terrain that are more prone to melting, tend to have higher error values. This occurs, amongst others, because radar reflections may occur from different snowpack layers (e.g. from liquid water within the snow or from refrozen ice lenses). This indicates that dynamic regions are harder to monitor consistently. In general, radar altimeters thus tend to perform better in the central, flat regions of Greenland due to simpler topography and relatively stable surfaces, while marginal zones with complex and dynamic terrain (where the most significant SEC occur) exhibit an increased uncertainty.\n\nLet us at last quantify the amount of pixels that do not meet the GCOS precision error requirement of 0.1 m/yr. We therefore multiply the errors accompanying the data by 2 to get a precision error in the form of 2 standard deviations (as proposed by GCOS):\n\n```text\nThe absolute number of pixels with a surface elevation change rate precision error (2σ) < 0.1 m/yr is 1025570 pixels or 99.41%.\n```\n\nAs already discussed above, almost all pixels exhibit error values that fall within international proposed thresholds, indicating a robust error characterization of the C3S SEC product for the GrIS.\n\n(section-4)="} {"chunk_id": "satellite_satellite-ice-sheet-elevation-change_trend-assessment_q01__78611312d58a", "report_id": "satellite_satellite-ice-sheet-elevation-change_trend-assessment_q01", "dataset_id": "satellite-ice-sheet-elevation-change", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q01", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments > Analysis and results > 4. Ice sheet surface elevation and altimetric volume change trends > 4.1 Spatial distribution of linear and quadratic trends", "title": "Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments", "chunk_index": 15, "token_count": 496, "text_raw": "With the patterns of the surface elevation changes and their uncertainty known, let us proceed with the analysis. In this subsection, we explore the spatial distribution of surface elevation change rate trends across the Greenland Ice Sheet. Let us start by exploring the linear trends:\n\nApply the function to the surface elevation change linear trends\nPlot the data\n\n*Figure 7. Linear trends of the ice sheet surface elevation change over Greenland from the SEC dataset on the Climate Data Store.*\n\nThe maps shows a similar pattern as the multiyear mean surface elevation change rates. The observed patterns highlight regional variability in surface elevation change trends, with coastal areas experiencing more substantial changes compared to interior regions. The data align with expected climate change impacts, where accelerating outlet glaciers and warming temperatures lead to increased melting and surface lowering, particularly in the low-elevation coastal areas [[1](https://doi.org/10.1016/j.epsl.2018.05.015), [3](https://doi.org/10.1016/j.rse.2016.12.012), [5](https://doi.org/10.1029/2020GL091216), [8](https://doi.org/10.5194/tc-5-173-2011), [10](https://doi.org/10.1038/s41586-019-1855-2)].\n\nLet us now consider quadratic trends (accelerations):\n\nApply the function to the surface elevation change quadratic trends\nPlot the data\n\n*Figure 8. Quadratic trends (acceleration) of the ice sheet surface elevation change over Greenland from the SEC dataset on the Climate Data Store.*\n\nAgain, the observed patterns highlight a clear regional variability in the acceleration and deceleration of surface elevation change rates. Coastal areas show more substantial changes in the acceleration of elevation change rates compared to central regions, which tend to be more stable.\n\n(section-4-2)=", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments\"\nDataset: satellite-ice-sheet-elevation-change [CDS]\nAspect: trend-assessment_q01 | Category: Satellite_ECVs\nSection: Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments > Analysis and results > 4. Ice sheet surface elevation and altimetric volume change trends > 4.1 Spatial distribution of linear and quadratic trends\n---\nWith the patterns of the surface elevation changes and their uncertainty known, let us proceed with the analysis. In this subsection, we explore the spatial distribution of surface elevation change rate trends across the Greenland Ice Sheet. Let us start by exploring the linear trends:\n\nApply the function to the surface elevation change linear trends\nPlot the data\n\n*Figure 7. Linear trends of the ice sheet surface elevation change over Greenland from the SEC dataset on the Climate Data Store.*\n\nThe maps shows a similar pattern as the multiyear mean surface elevation change rates. The observed patterns highlight regional variability in surface elevation change trends, with coastal areas experiencing more substantial changes compared to interior regions. The data align with expected climate change impacts, where accelerating outlet glaciers and warming temperatures lead to increased melting and surface lowering, particularly in the low-elevation coastal areas [[1](https://doi.org/10.1016/j.epsl.2018.05.015), [3](https://doi.org/10.1016/j.rse.2016.12.012), [5](https://doi.org/10.1029/2020GL091216), [8](https://doi.org/10.5194/tc-5-173-2011), [10](https://doi.org/10.1038/s41586-019-1855-2)].\n\nLet us now consider quadratic trends (accelerations):\n\nApply the function to the surface elevation change quadratic trends\nPlot the data\n\n*Figure 8. Quadratic trends (acceleration) of the ice sheet surface elevation change over Greenland from the SEC dataset on the Climate Data Store.*\n\nAgain, the observed patterns highlight a clear regional variability in the acceleration and deceleration of surface elevation change rates. Coastal areas show more substantial changes in the acceleration of elevation change rates compared to central regions, which tend to be more stable.\n\n(section-4-2)="} {"chunk_id": "satellite_satellite-ice-sheet-elevation-change_trend-assessment_q01__c74ce8534c0b", "report_id": "satellite_satellite-ice-sheet-elevation-change_trend-assessment_q01", "dataset_id": "satellite-ice-sheet-elevation-change", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q01", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments > Analysis and results > 4. Ice sheet surface elevation and altimetric volume change trends > 4.2 Time series of surface elevation and volume change trends", "title": "Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments", "chunk_index": 16, "token_count": 630, "text_raw": "Let us now quantify the ice sheet-wide surface elevation change and volume change trends as a time series:\n\nCompute coefficients\nPlot trends and print stats\n\n```text\nThe slope of the ice sheet cumulative average surface elevation change is -0.055 m/yr.\nThe trend is significant at an alpha level of 0.05, i.e. a monotonic linear trend is present.\nThe acceleration of the ice sheet cumulative average surface elevation change is -0.004 m/yr^2.\nThe slope of the ice sheet cumulative altimetric volume change is -94.701 km³/yr.\nThe trend is significant at an alpha level of 0.05, i.e. a monotonic linear trend is present.\nThe acceleration of the ice sheet cumulative altimetric volume change is -6.478 km³/yr^2.\n```\n\n*Figure 9. Linear and quadratic trends of the cumulative (above) average ice sheet surface elevation change and (below) altimetric volume change over Greenland from the SEC dataset on the Climate Data Store.*\n\nAs seen earlier, the significant negative and downward trend, for both the linear and quadratic trend, implies that the surface is generally lowering over the GrIS, and that this lowering has been accelerating over the past several decades. As discussed before, the favorable error characterization and spatial/temporal coverage (with filled-up data gaps) of the C3S GrIS SEC product allow for a reliable and robust statistical analysis. The above-derived trends can thus be interpreted as being credible. Although surface elevation changes are not solely impacted by surface mass balance processes, the observed trends are moreover consistent with the expected impacts of climate change, where accelerating outlet glaciers and rising global temperatures result in increased melting, runoff and surface lowering [e.g. [1](https://doi.org/10.1016/j.epsl.2018.05.015), [4](https://doi.org/10.3189/172756505781829007), [8](https://doi.org/10.5194/tc-5-173-2011), [9](https://doi.org/10.1029/2021JF006505), [10](https://doi.org/10.1038/s41586-019-1855-2)]. Nevertheless, a notable bias exists when comparing the radar altimetry-derived SEC to those from laser altimetry, and the coarse spatial resolution of the C3S product may aid the corresponding underestimation of surface lowering rates as discussed before.\n\n(section-5)=", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments\"\nDataset: satellite-ice-sheet-elevation-change [CDS]\nAspect: trend-assessment_q01 | Category: Satellite_ECVs\nSection: Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments > Analysis and results > 4. Ice sheet surface elevation and altimetric volume change trends > 4.2 Time series of surface elevation and volume change trends\n---\nLet us now quantify the ice sheet-wide surface elevation change and volume change trends as a time series:\n\nCompute coefficients\nPlot trends and print stats\n\n```text\nThe slope of the ice sheet cumulative average surface elevation change is -0.055 m/yr.\nThe trend is significant at an alpha level of 0.05, i.e. a monotonic linear trend is present.\nThe acceleration of the ice sheet cumulative average surface elevation change is -0.004 m/yr^2.\nThe slope of the ice sheet cumulative altimetric volume change is -94.701 km³/yr.\nThe trend is significant at an alpha level of 0.05, i.e. a monotonic linear trend is present.\nThe acceleration of the ice sheet cumulative altimetric volume change is -6.478 km³/yr^2.\n```\n\n*Figure 9. Linear and quadratic trends of the cumulative (above) average ice sheet surface elevation change and (below) altimetric volume change over Greenland from the SEC dataset on the Climate Data Store.*\n\nAs seen earlier, the significant negative and downward trend, for both the linear and quadratic trend, implies that the surface is generally lowering over the GrIS, and that this lowering has been accelerating over the past several decades. As discussed before, the favorable error characterization and spatial/temporal coverage (with filled-up data gaps) of the C3S GrIS SEC product allow for a reliable and robust statistical analysis. The above-derived trends can thus be interpreted as being credible. Although surface elevation changes are not solely impacted by surface mass balance processes, the observed trends are moreover consistent with the expected impacts of climate change, where accelerating outlet glaciers and rising global temperatures result in increased melting, runoff and surface lowering [e.g. [1](https://doi.org/10.1016/j.epsl.2018.05.015), [4](https://doi.org/10.3189/172756505781829007), [8](https://doi.org/10.5194/tc-5-173-2011), [9](https://doi.org/10.1029/2021JF006505), [10](https://doi.org/10.1038/s41586-019-1855-2)]. Nevertheless, a notable bias exists when comparing the radar altimetry-derived SEC to those from laser altimetry, and the coarse spatial resolution of the C3S product may aid the corresponding underestimation of surface lowering rates as discussed before.\n\n(section-5)="} {"chunk_id": "satellite_satellite-ice-sheet-elevation-change_trend-assessment_q01__8f03bf6fec21", "report_id": "satellite_satellite-ice-sheet-elevation-change_trend-assessment_q01", "dataset_id": "satellite-ice-sheet-elevation-change", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q01", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments > Analysis and results > 5. Short summary and take-home messages", "title": "Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments", "chunk_index": 17, "token_count": 818, "text_raw": "When measured over an extended period and over the complete ice sheet, trends in Greenland Ice Sheet surface elevation changes can serve as a clear indicator of global climate change. To ensure these trends are accurate, representative and reliable, the dataset should be sufficiently adequate in terms of its quality and maturity. Hence, the dataset should at least exhibit a comprehensive spatial coverage (i.e. ice sheet-wide), a long and continuous temporal coverage (> 30 years), quantified and transparent pixel-by-pixel uncertainty estimates that meet international proposed thresholds [[6](https://library.wmo.int/idurl/4/58111)], a validation effort, and an adequate spatio-temporal resolution (cfr. the \"Maturity Matrix\" [[7](https://doi.org/10.1175/BAMS-D-21-0109.1)]).\n\nAll in all, it can be said that the Greenland Ice Sheet (GrIS) surface elevation change dataset serves a robust indicator of global climate change, as several criteria for certain key quality aspects have been met. The temporal window of the data spams over 30 years at a consistent monthly resolution, which makes the surface elevation change dataset highly applicable for trend analysis. Furthermore, there is a complete spatio-temporal coverage since 1992 with filled-up data gaps. Error estimates (presented as 1-sigma precision errors) align with international proposed thresholds, confirming the dataset's high quality in terms of data uncertainty. Nevertheless, radar altimeters tend to perform better in the central, flat regions of the Greenland Ice Sheet, which have a simpler topography and a more stable surface, compared to the coastal areas with a more complex terrain (with steep slopes, valleys and ridges and large seasonal cycles of melt and accumulation) where errors are generally higher.\n\nHowever, a notable bias exists when comparing radar altimetry-derived surface elevation changes to laser altimetry, particularly for significant lowering rates (below -0.5 to -1 m/yr), leading to underestimated surface elevation change and volume loss rates [[3](https://doi.org/10.1016/j.rse.2016.12.012)]. Apart from that, the coarse spatial resolution of 25 km is insufficient to capture the very localized mass loss of outlet glaciers, further complicating the data and underestimating surface elevation changes and volume losses to some degree [[5](https://doi.org/10.1029/2020GL091216)]. Users should furthermore note that surface elevation changes do not necessarily equal \"actual\" ice sheet volume nor mass changes, as processes such as bedrock elevation changes, firn densificaiton and density corrections for volume to mass changes have to be taken into account as well. The surface elevation change data can thus only be converted to \"actual\" ice sheet volume or mass changes if certain corrections and/or assumptions are made. Users of the C3S GrIS SEC data product are therefore most likely required to implement additional processing to derive glaciologically interpretable volume/mass changes from the provided surface elevation changes [e.g. [1](https://doi.org/10.1016/j.epsl.2018.05.015), [4](https://doi.org/10.3189/172756505781829007), [8](https://doi.org/10.5194/tc-5-173-2011), [9](https://doi.org/10.1029/2021JF006505), [10](https://doi.org/10.1038/s41586-019-1855-2)].", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments\"\nDataset: satellite-ice-sheet-elevation-change [CDS]\nAspect: trend-assessment_q01 | Category: Satellite_ECVs\nSection: Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments > Analysis and results > 5. Short summary and take-home messages\n---\nWhen measured over an extended period and over the complete ice sheet, trends in Greenland Ice Sheet surface elevation changes can serve as a clear indicator of global climate change. To ensure these trends are accurate, representative and reliable, the dataset should be sufficiently adequate in terms of its quality and maturity. Hence, the dataset should at least exhibit a comprehensive spatial coverage (i.e. ice sheet-wide), a long and continuous temporal coverage (> 30 years), quantified and transparent pixel-by-pixel uncertainty estimates that meet international proposed thresholds [[6](https://library.wmo.int/idurl/4/58111)], a validation effort, and an adequate spatio-temporal resolution (cfr. the \"Maturity Matrix\" [[7](https://doi.org/10.1175/BAMS-D-21-0109.1)]).\n\nAll in all, it can be said that the Greenland Ice Sheet (GrIS) surface elevation change dataset serves a robust indicator of global climate change, as several criteria for certain key quality aspects have been met. The temporal window of the data spams over 30 years at a consistent monthly resolution, which makes the surface elevation change dataset highly applicable for trend analysis. Furthermore, there is a complete spatio-temporal coverage since 1992 with filled-up data gaps. Error estimates (presented as 1-sigma precision errors) align with international proposed thresholds, confirming the dataset's high quality in terms of data uncertainty. Nevertheless, radar altimeters tend to perform better in the central, flat regions of the Greenland Ice Sheet, which have a simpler topography and a more stable surface, compared to the coastal areas with a more complex terrain (with steep slopes, valleys and ridges and large seasonal cycles of melt and accumulation) where errors are generally higher.\n\nHowever, a notable bias exists when comparing radar altimetry-derived surface elevation changes to laser altimetry, particularly for significant lowering rates (below -0.5 to -1 m/yr), leading to underestimated surface elevation change and volume loss rates [[3](https://doi.org/10.1016/j.rse.2016.12.012)]. Apart from that, the coarse spatial resolution of 25 km is insufficient to capture the very localized mass loss of outlet glaciers, further complicating the data and underestimating surface elevation changes and volume losses to some degree [[5](https://doi.org/10.1029/2020GL091216)]. Users should furthermore note that surface elevation changes do not necessarily equal \"actual\" ice sheet volume nor mass changes, as processes such as bedrock elevation changes, firn densificaiton and density corrections for volume to mass changes have to be taken into account as well. The surface elevation change data can thus only be converted to \"actual\" ice sheet volume or mass changes if certain corrections and/or assumptions are made. Users of the C3S GrIS SEC data product are therefore most likely required to implement additional processing to derive glaciologically interpretable volume/mass changes from the provided surface elevation changes [e.g. [1](https://doi.org/10.1016/j.epsl.2018.05.015), [4](https://doi.org/10.3189/172756505781829007), [8](https://doi.org/10.5194/tc-5-173-2011), [9](https://doi.org/10.1029/2021JF006505), [10](https://doi.org/10.1038/s41586-019-1855-2)]."} {"chunk_id": "satellite_satellite-ice-sheet-elevation-change_trend-assessment_q01__214c1dc1f6e3", "report_id": "satellite_satellite-ice-sheet-elevation-change_trend-assessment_q01", "dataset_id": "satellite-ice-sheet-elevation-change", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q01", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments > ℹ️ If you want to know more > Key resources", "title": "Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments", "chunk_index": 18, "token_count": 348, "text_raw": "- [\"Ice sheet surface elevation change rate for Greenland and Antarctica from 1992 to present derived from satellite observations\"](https://cds.climate.copernicus.eu/datasets/satellite-ice-sheet-elevation-change?tab=overview) on the CDS\n- [Documentation on the CDS](https://cds.climate.copernicus.eu/datasets/satellite-ice-sheet-elevation-change?tab=overview) and the [ECMWF Confluence Wiki](https://confluence.ecmwf.int/pages/viewpage.action?pageId=355345393) (Copernicus Knowledge Base).\n- [Copernicus climate change indicators: ice sheets](https://climate.copernicus.eu/climate-indicators/ice-sheets)\n- [An easy-to-read article about ice sheet altimetry](https://blogs.egu.eu/divisions/cr/2023/03/03/ice-radar-altimetry/)\n- [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control` prepared by [B-Open](https://www.bopen.eu).", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments\"\nDataset: satellite-ice-sheet-elevation-change [CDS]\nAspect: trend-assessment_q01 | Category: Satellite_ECVs\nSection: Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments > ℹ️ If you want to know more > Key resources\n---\n- [\"Ice sheet surface elevation change rate for Greenland and Antarctica from 1992 to present derived from satellite observations\"](https://cds.climate.copernicus.eu/datasets/satellite-ice-sheet-elevation-change?tab=overview) on the CDS\n- [Documentation on the CDS](https://cds.climate.copernicus.eu/datasets/satellite-ice-sheet-elevation-change?tab=overview) and the [ECMWF Confluence Wiki](https://confluence.ecmwf.int/pages/viewpage.action?pageId=355345393) (Copernicus Knowledge Base).\n- [Copernicus climate change indicators: ice sheets](https://climate.copernicus.eu/climate-indicators/ice-sheets)\n- [An easy-to-read article about ice sheet altimetry](https://blogs.egu.eu/divisions/cr/2023/03/03/ice-radar-altimetry/)\n- [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control` prepared by [B-Open](https://www.bopen.eu)."} {"chunk_id": "satellite_satellite-ice-sheet-elevation-change_trend-assessment_q01__6146de44602d", "report_id": "satellite_satellite-ice-sheet-elevation-change_trend-assessment_q01", "dataset_id": "satellite-ice-sheet-elevation-change", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q01", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments > ℹ️ If you want to know more > References", "title": "Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments", "chunk_index": 19, "token_count": 1086, "text_raw": "- [[1](https://doi.org/10.1016/j.epsl.2018.05.015)] Sørensen, L. S., Simonsen, S. B., Forsberg, R., Khvorostovsky, K., Meister, R., and Engdahl, M. E. (2018). 25 years of elevation changes of the Greenland Ice Sheet from ERS, Envisat, and CryoSat-2 radar altimetry, Earth and Planetary Science Letters. 495. https://doi.org/10.1016/j.epsl.2018.05.015\n\n- [[2](https://doi.org/10.5194/tc-13-427-2019)] Schröder, L., Horwath, M., Dietrich, R., Helm, V., van den Broeke, M.R., and Ligtenberg, S.R.M. (2019). Four decades of Antarctic surface elevation change from multi-mission satellite altimetry. The Cryosphere, 13, p. 427-449. https://doi.org/10.5194/tc-13-427-2019\n\n- [[3](https://doi.org/10.1016/j.rse.2016.12.012)] Simonsen, S. B., and Sørensen, L. S. (2017). Implications of Changing Scattering Properties on Greenland Ice Sheet Volume Change from Cryosat-2 Altimetry. Remote Sensing of Environment. https://doi.org/10.1016/j.rse.2016.12.012\n\n- [[4](https://doi.org/10.3189/172756505781829007)] Zwally, H. J., Giovinetto, M.B., Li, J., Cornejo, H.G, Beckley, M.A., Brenner, A.C., Saba, J.L., and Yi, D. (2005). Mass changes of the Greenland and Antarctic ice sheets and shelves and contributions to sea-level rise: 1992–2002. J. Glaciol. 51, 509–527, https://doi.org/10.3189/172756505781829007\n\n- [[5](https://doi.org/10.1029/2020GL091216)] Simonsen, S. B., Barletta, V. R., Colgan, W. T., and Sørensen, L. S. (2021). Greenland Ice Sheet mass balance (1992–2020) from calibrated radar altimetry. Geophysical Research Letters, 48(3), e2020GL091216. https://doi.org/10.1029/2020GL091216\n\n- [[6](https://library.wmo.int/idurl/4/58111)] GCOS (Global Climate Observing System) (2022). The 2022 GCOS ECVs Requirements (GCOS-245). World Meteorological Organization: Geneva, Switzerland. doi: https://library.wmo.int/idurl/4/58111\n\n- [[7](https://doi.org/10.1175/BAMS-D-21-0109.1)] Yang, C. X., Cagnazzo, C., Artale, V., Nardelli, B. B., Buontempo, C., Busatto, J., Caporaso, L., Cesarini, C., Cionni, I., Coll, J., Crezee, B., Cristofanelli, P., de Toma, V., Essa, Y. H., Eyring, V., Fierli, F., Grant, L., Hassler, B., Hirschi, M., Huybrechts, P., Le Merle, E., Leonelli, F. E., Lin, X., Madonna, F., Mason, E., Massonnet, F., Marcos, M., Marullo, S., Muller, B., Obregon, A., Organelli, E., Palacz, A., Pascual, A., Pisano, A., Putero, D., Rana, A., Sanchez-Roman, A., Seneviratne, S. I., Serva, F., Storto, A., Thiery, W., Throne, P., Van Tricht, L., Verhaegen, Y., Volpe, G., and Santoleri, R. (2022). Independent Quality Assessment of Essential Climate Variables: Lessons Learned from the Copernicus Climate Change Service, B. Am. Meteorol. Soc., 103, E2032–E2049, doi: 10.1175/Bams-D-21-0109.1.", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments\"\nDataset: satellite-ice-sheet-elevation-change [CDS]\nAspect: trend-assessment_q01 | Category: Satellite_ECVs\nSection: Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments > ℹ️ If you want to know more > References\n---\n- [[1](https://doi.org/10.1016/j.epsl.2018.05.015)] Sørensen, L. S., Simonsen, S. B., Forsberg, R., Khvorostovsky, K., Meister, R., and Engdahl, M. E. (2018). 25 years of elevation changes of the Greenland Ice Sheet from ERS, Envisat, and CryoSat-2 radar altimetry, Earth and Planetary Science Letters. 495. https://doi.org/10.1016/j.epsl.2018.05.015\n\n- [[2](https://doi.org/10.5194/tc-13-427-2019)] Schröder, L., Horwath, M., Dietrich, R., Helm, V., van den Broeke, M.R., and Ligtenberg, S.R.M. (2019). Four decades of Antarctic surface elevation change from multi-mission satellite altimetry. The Cryosphere, 13, p. 427-449. https://doi.org/10.5194/tc-13-427-2019\n\n- [[3](https://doi.org/10.1016/j.rse.2016.12.012)] Simonsen, S. B., and Sørensen, L. S. (2017). Implications of Changing Scattering Properties on Greenland Ice Sheet Volume Change from Cryosat-2 Altimetry. Remote Sensing of Environment. https://doi.org/10.1016/j.rse.2016.12.012\n\n- [[4](https://doi.org/10.3189/172756505781829007)] Zwally, H. J., Giovinetto, M.B., Li, J., Cornejo, H.G, Beckley, M.A., Brenner, A.C., Saba, J.L., and Yi, D. (2005). Mass changes of the Greenland and Antarctic ice sheets and shelves and contributions to sea-level rise: 1992–2002. J. Glaciol. 51, 509–527, https://doi.org/10.3189/172756505781829007\n\n- [[5](https://doi.org/10.1029/2020GL091216)] Simonsen, S. B., Barletta, V. R., Colgan, W. T., and Sørensen, L. S. (2021). Greenland Ice Sheet mass balance (1992–2020) from calibrated radar altimetry. Geophysical Research Letters, 48(3), e2020GL091216. https://doi.org/10.1029/2020GL091216\n\n- [[6](https://library.wmo.int/idurl/4/58111)] GCOS (Global Climate Observing System) (2022). The 2022 GCOS ECVs Requirements (GCOS-245). World Meteorological Organization: Geneva, Switzerland. doi: https://library.wmo.int/idurl/4/58111\n\n- [[7](https://doi.org/10.1175/BAMS-D-21-0109.1)] Yang, C. X., Cagnazzo, C., Artale, V., Nardelli, B. B., Buontempo, C., Busatto, J., Caporaso, L., Cesarini, C., Cionni, I., Coll, J., Crezee, B., Cristofanelli, P., de Toma, V., Essa, Y. H., Eyring, V., Fierli, F., Grant, L., Hassler, B., Hirschi, M., Huybrechts, P., Le Merle, E., Leonelli, F. E., Lin, X., Madonna, F., Mason, E., Massonnet, F., Marcos, M., Marullo, S., Muller, B., Obregon, A., Organelli, E., Palacz, A., Pascual, A., Pisano, A., Putero, D., Rana, A., Sanchez-Roman, A., Seneviratne, S. I., Serva, F., Storto, A., Thiery, W., Throne, P., Van Tricht, L., Verhaegen, Y., Volpe, G., and Santoleri, R. (2022). Independent Quality Assessment of Essential Climate Variables: Lessons Learned from the Copernicus Climate Change Service, B. Am. Meteorol. Soc., 103, E2032–E2049, doi: 10.1175/Bams-D-21-0109.1."} {"chunk_id": "satellite_satellite-ice-sheet-elevation-change_trend-assessment_q01__4898889d729a", "report_id": "satellite_satellite-ice-sheet-elevation-change_trend-assessment_q01", "dataset_id": "satellite-ice-sheet-elevation-change", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q01", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments > ℹ️ If you want to know more > References", "title": "Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments", "chunk_index": 20, "token_count": 812, "text_raw": "Lessons Learned from the Copernicus Climate Change Service, B. Am. Meteorol. Soc., 103, E2032–E2049, doi: 10.1175/Bams-D-21-0109.1.\n\n- [[8](https://doi.org/10.5194/tc-5-173-2011)] Sørensen, L. S., Simonsen, S. B., Nielsen, K., Lucas-Picher, P., Spada, G., Adalgeirsdottir, G., Forsberg, R., and Hvidberg, C. S. (2011). Mass balance of the Greenland ice sheet (2003–2008) from ICESat data – the impact of interpolation, sampling and firn density, The Cryosphere, 5, 173–186, https://doi.org/10.5194/tc-5-173-2011\n\n- [[9](https://doi.org/10.1029/2021JF006505)] Khan, S. A., Bamber, J. L., Rignot, E., Helm, V., Aschwanden, A., Holland, D. M., van den Broeke, M., King, M., Noël, B., Truffer, M., Humbert, A., Solgaard, A. M., Box, J. E., Colgan, W. T., Wuite, J., Mouginot, J., Andersen, O. B., Csatho, B., Felikson, D., Fettweis, X., Forsberg, R., Gogineni, P., Joughin, I., Kjeldsen, K. K., Kuschnerus, M., Langen, P. L., Luckman, A., Luthcke, S. B., McMillan, M., Merryman Boncori, J. P., Morlighem, M., Mottram, R., Nagler, T., Nagy, T., Paden, J., Palmer, S., Poinar, K., Shepherd, A., Smith, B., Stearns, L. A., van Angelen, J. H., van der Wal, W., van de Berg, W. J., van Wessem, M., Velicogna, I., Wahr, J., Wendt, A., Wouters, B., & Zwally, H. J. (2022). Greenland mass trends from airborne and satellite altimetry during 2011–2020. Journal of Geophysical Research: Earth Surface, 127(3). https://doi.org/10.1029/2021JF006505\n\n- [[10](https://doi.org/10.1038/s41586-019-1855-2)] The IMBIE Team (2019). Mass balance of the Greenland Ice Sheet from 1992 to 2018. Nature, 579, 233–239. https://doi.org/10.1038/s41586-019-1855-2\n\n- [[11](https://doi.org/10.1038/s41586-018-0179-y)] The IMBIE Team (2018). Mass balance of the Antarctic Ice Sheet from 1992 to 2017. Nature 558, 219–222. https://doi.org/10.1038/s41586-018-0179-y", "text_with_prefix": "EQC Quality Assessment: \"Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments\"\nDataset: satellite-ice-sheet-elevation-change [CDS]\nAspect: trend-assessment_q01 | Category: Satellite_ECVs\nSection: Uncertainty of ice sheet surface elevation change data over Greenland for volume and mass change trend assessments > ℹ️ If you want to know more > References\n---\nLessons Learned from the Copernicus Climate Change Service, B. Am. Meteorol. Soc., 103, E2032–E2049, doi: 10.1175/Bams-D-21-0109.1.\n\n- [[8](https://doi.org/10.5194/tc-5-173-2011)] Sørensen, L. S., Simonsen, S. B., Nielsen, K., Lucas-Picher, P., Spada, G., Adalgeirsdottir, G., Forsberg, R., and Hvidberg, C. S. (2011). Mass balance of the Greenland ice sheet (2003–2008) from ICESat data – the impact of interpolation, sampling and firn density, The Cryosphere, 5, 173–186, https://doi.org/10.5194/tc-5-173-2011\n\n- [[9](https://doi.org/10.1029/2021JF006505)] Khan, S. A., Bamber, J. L., Rignot, E., Helm, V., Aschwanden, A., Holland, D. M., van den Broeke, M., King, M., Noël, B., Truffer, M., Humbert, A., Solgaard, A. M., Box, J. E., Colgan, W. T., Wuite, J., Mouginot, J., Andersen, O. B., Csatho, B., Felikson, D., Fettweis, X., Forsberg, R., Gogineni, P., Joughin, I., Kjeldsen, K. K., Kuschnerus, M., Langen, P. L., Luckman, A., Luthcke, S. B., McMillan, M., Merryman Boncori, J. P., Morlighem, M., Mottram, R., Nagler, T., Nagy, T., Paden, J., Palmer, S., Poinar, K., Shepherd, A., Smith, B., Stearns, L. A., van Angelen, J. H., van der Wal, W., van de Berg, W. J., van Wessem, M., Velicogna, I., Wahr, J., Wendt, A., Wouters, B., & Zwally, H. J. (2022). Greenland mass trends from airborne and satellite altimetry during 2011–2020. Journal of Geophysical Research: Earth Surface, 127(3). https://doi.org/10.1029/2021JF006505\n\n- [[10](https://doi.org/10.1038/s41586-019-1855-2)] The IMBIE Team (2019). Mass balance of the Greenland Ice Sheet from 1992 to 2018. Nature, 579, 233–239. https://doi.org/10.1038/s41586-019-1855-2\n\n- [[11](https://doi.org/10.1038/s41586-018-0179-y)] The IMBIE Team (2018). Mass balance of the Antarctic Ice Sheet from 1992 to 2017. Nature 558, 219–222. https://doi.org/10.1038/s41586-018-0179-y"} {"chunk_id": "satellite_satellite-ice-sheet-mass-balance_trend-assessment_q02__0c32943c2f12", "report_id": "satellite_satellite-ice-sheet-mass-balance_trend-assessment_q02", "dataset_id": "satellite-ice-sheet-mass-balance", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q02", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Gravimetric mass balance data from satellite observations: utility analysis for mass change monitoring of the ice sheets", "title": "Gravimetric mass balance data from satellite observations: utility analysis for mass change monitoring of the ice sheets", "chunk_index": 0, "token_count": 128, "text_raw": "Production date: 31-03-2025\n\nDataset version: 4.0\n\nProduced by: Yoni Verhaegen and Philippe Huybrechts (Vrije Universiteit Brussel)", "text_with_prefix": "EQC Quality Assessment: \"Gravimetric mass balance data from satellite observations: utility analysis for mass change monitoring of the ice sheets\"\nDataset: satellite-ice-sheet-mass-balance [CDS]\nAspect: trend-assessment_q02 | Category: Satellite_ECVs\nSection: Gravimetric mass balance data from satellite observations: utility analysis for mass change monitoring of the ice sheets\n---\nProduction date: 31-03-2025\n\nDataset version: 4.0\n\nProduced by: Yoni Verhaegen and Philippe Huybrechts (Vrije Universiteit Brussel)"} {"chunk_id": "satellite_satellite-ice-sheet-mass-balance_trend-assessment_q02__90b32877a260", "report_id": "satellite_satellite-ice-sheet-mass-balance_trend-assessment_q02", "dataset_id": "satellite-ice-sheet-mass-balance", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q02", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Gravimetric mass balance data from satellite observations: utility analysis for mass change monitoring of the ice sheets > Quality assessment question", "title": "Gravimetric mass balance data from satellite observations: utility analysis for mass change monitoring of the ice sheets", "chunk_index": 1, "token_count": 471, "text_raw": "* **\"Is the dataset sufficiently consistent in terms of spatio-temporal completeness to derive multi-year trends in ice sheet mass changes and their corresponding contribution to global sea level rise?\"**\n\nIce sheets are significant contributors to current (and future) global sea-level rise, indicators of climatic changes, and key players regarding feedback mechanisms within the atmosphere, ocean, and cryosphere. Assessing ice sheet mass changes is essential for understanding these mechanisms. At present-day, the gravimetric method from satellite-based remote sensing offers one out of the three feasible ways to regularly and accurately monitor mass changes of the entire ice sheets at a regular basis (the others being the altimetric and input-output method). The [\"Gravimetric mass balance data for the Antarctic and Greenland ice sheets from 2003 to 2022 derived from satellite observations\"](https://cds.climate.copernicus.eu/datasets/satellite-ice-sheet-mass-balance?tab=overview) dataset on the Climate Data Store (CDS) therefore provides valuable insights into the ice sheet's mass changes. The dataset uses satellite gravimetry from the GRACE and GRACE-FO missions to detect changes in the Earth's gravitational field. These are then translated into cumulative mass anomalies of ice above buoyancy for the grounded ice of both the Greenland (GrIS) and Antarctic (AIS) ice sheets [[1](https://doi.org/10.1007/s10712-016-9398-7), [2](https://doi.org/10.3390/geosciences9100415)]. This notebook investigates whether the mass change dataset is of sufficient maturity and quality in terms of its spatio-temporal completeness (i.e. coverage and resolution) to derive multi-year trends in ice sheet mass changes and their corresponding contribution to global sea level rise. For that, we use dataset version 4.0.", "text_with_prefix": "EQC Quality Assessment: \"Gravimetric mass balance data from satellite observations: utility analysis for mass change monitoring of the ice sheets\"\nDataset: satellite-ice-sheet-mass-balance [CDS]\nAspect: trend-assessment_q02 | Category: Satellite_ECVs\nSection: Gravimetric mass balance data from satellite observations: utility analysis for mass change monitoring of the ice sheets > Quality assessment question\n---\n* **\"Is the dataset sufficiently consistent in terms of spatio-temporal completeness to derive multi-year trends in ice sheet mass changes and their corresponding contribution to global sea level rise?\"**\n\nIce sheets are significant contributors to current (and future) global sea-level rise, indicators of climatic changes, and key players regarding feedback mechanisms within the atmosphere, ocean, and cryosphere. Assessing ice sheet mass changes is essential for understanding these mechanisms. At present-day, the gravimetric method from satellite-based remote sensing offers one out of the three feasible ways to regularly and accurately monitor mass changes of the entire ice sheets at a regular basis (the others being the altimetric and input-output method). The [\"Gravimetric mass balance data for the Antarctic and Greenland ice sheets from 2003 to 2022 derived from satellite observations\"](https://cds.climate.copernicus.eu/datasets/satellite-ice-sheet-mass-balance?tab=overview) dataset on the Climate Data Store (CDS) therefore provides valuable insights into the ice sheet's mass changes. The dataset uses satellite gravimetry from the GRACE and GRACE-FO missions to detect changes in the Earth's gravitational field. These are then translated into cumulative mass anomalies of ice above buoyancy for the grounded ice of both the Greenland (GrIS) and Antarctic (AIS) ice sheets [[1](https://doi.org/10.1007/s10712-016-9398-7), [2](https://doi.org/10.3390/geosciences9100415)]. This notebook investigates whether the mass change dataset is of sufficient maturity and quality in terms of its spatio-temporal completeness (i.e. coverage and resolution) to derive multi-year trends in ice sheet mass changes and their corresponding contribution to global sea level rise. For that, we use dataset version 4.0."} {"chunk_id": "satellite_satellite-ice-sheet-mass-balance_trend-assessment_q02__71f8942b1831", "report_id": "satellite_satellite-ice-sheet-mass-balance_trend-assessment_q02", "dataset_id": "satellite-ice-sheet-mass-balance", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q02", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Gravimetric mass balance data from satellite observations: utility analysis for mass change monitoring of the ice sheets > Quality assessment statements", "title": "Gravimetric mass balance data from satellite observations: utility analysis for mass change monitoring of the ice sheets", "chunk_index": 2, "token_count": 565, "text_raw": "These are the key outcomes of this assessment\n\n- The ice sheet mass change dataset, making use of the GRACE(-FO) satellite missions, is a useful tool for quantifying the total ice mass change of the Greenland (GrIS) and Antarctic (AIS) ice sheets. The C3S data are presented as a time series of the ice sheet GMB data and exhibits a long temporal extent (> 20 years). The data represent mass changes aggregated from different basins or from an ice sheet-wide coverage at a (quasi) monthly temporal resolution. Since it only captures ice mass changes above buoyancy, ice mass changes from GRACE(-FO) can be directly converted to global sea level contributions, of which the dataset values agree well with values found in the literature from other methods [[9](https://iopscience.iop.org/article/10.1088/1748-9326/aac2f0/meta), [12](https://doi.org/10.1038/s41586-019-1855-2), [13](https://doi.org/10.1038/s41586-018-0179-y)]. The C3S GMB dataset therefore effectively captures the long-term mean, trends and variability of ice sheet mass changes and the corresponding global sea level contributions, and can thus serve as a clear indicator of global climate change and water cycle changes.\n\n- Data limitations include a coarse spatial resolution of the data acquisition and dataset itself (i.e. since there is a lack of gridded data at the pixel level of the final product, the term \"spatial resolution\" is actually not appropriate for this dataset), undesired signal leakage from adjacent regions (e.g. the nearby Canadian ice caps for the GrIS), and the need for complex data retrieval methods and geophysical corrections resulting in occasionally very high uncertainty values. Furthermore, data gaps exist, notably during 2017-2018 in between the transition of GRACE and its follow-up GRACE-FO. These are not flagged or filled, requiring users to manually identify missing periods. However, for linear trend estimation, gaps (even long ones) in the time series are practically irrelevant, as mass changes occurring during these gaps are still included in the subsequent solution, and secondly, there is no systematic bias between GRACE and GRACE-FO. \n```", "text_with_prefix": "EQC Quality Assessment: \"Gravimetric mass balance data from satellite observations: utility analysis for mass change monitoring of the ice sheets\"\nDataset: satellite-ice-sheet-mass-balance [CDS]\nAspect: trend-assessment_q02 | Category: Satellite_ECVs\nSection: Gravimetric mass balance data from satellite observations: utility analysis for mass change monitoring of the ice sheets > Quality assessment statements\n---\nThese are the key outcomes of this assessment\n\n- The ice sheet mass change dataset, making use of the GRACE(-FO) satellite missions, is a useful tool for quantifying the total ice mass change of the Greenland (GrIS) and Antarctic (AIS) ice sheets. The C3S data are presented as a time series of the ice sheet GMB data and exhibits a long temporal extent (> 20 years). The data represent mass changes aggregated from different basins or from an ice sheet-wide coverage at a (quasi) monthly temporal resolution. Since it only captures ice mass changes above buoyancy, ice mass changes from GRACE(-FO) can be directly converted to global sea level contributions, of which the dataset values agree well with values found in the literature from other methods [[9](https://iopscience.iop.org/article/10.1088/1748-9326/aac2f0/meta), [12](https://doi.org/10.1038/s41586-019-1855-2), [13](https://doi.org/10.1038/s41586-018-0179-y)]. The C3S GMB dataset therefore effectively captures the long-term mean, trends and variability of ice sheet mass changes and the corresponding global sea level contributions, and can thus serve as a clear indicator of global climate change and water cycle changes.\n\n- Data limitations include a coarse spatial resolution of the data acquisition and dataset itself (i.e. since there is a lack of gridded data at the pixel level of the final product, the term \"spatial resolution\" is actually not appropriate for this dataset), undesired signal leakage from adjacent regions (e.g. the nearby Canadian ice caps for the GrIS), and the need for complex data retrieval methods and geophysical corrections resulting in occasionally very high uncertainty values. Furthermore, data gaps exist, notably during 2017-2018 in between the transition of GRACE and its follow-up GRACE-FO. These are not flagged or filled, requiring users to manually identify missing periods. However, for linear trend estimation, gaps (even long ones) in the time series are practically irrelevant, as mass changes occurring during these gaps are still included in the subsequent solution, and secondly, there is no systematic bias between GRACE and GRACE-FO. \n```"} {"chunk_id": "satellite_satellite-ice-sheet-mass-balance_trend-assessment_q02__968ed148b82c", "report_id": "satellite_satellite-ice-sheet-mass-balance_trend-assessment_q02", "dataset_id": "satellite-ice-sheet-mass-balance", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q02", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Gravimetric mass balance data from satellite observations: utility analysis for mass change monitoring of the ice sheets > Methodology > Dataset description", "title": "Gravimetric mass balance data from satellite observations: utility analysis for mass change monitoring of the ice sheets", "chunk_index": 3, "token_count": 772, "text_raw": "The total mass balance of an ice sheet is the difference between mass gained (mainly from snow accumulation) and mass lost (by ablation or solid ice discharge across the grounding line), which is the same as the net mass change of the ice sheet. Remote sensing techniques, such as the use of satellites, are an important feature to derive and study the mass changes of the ice sheets. The ice sheet mass change dataset provides monthly gravimetric mass balance (GMB) values and their uncertainty for the Greenland (GrIS) and also the Antarctic ice sheet (AIS). The data represent a time series of the cumulative mass changes (mass anomalies) of the ice above buoyancy of the ice sheets and their basins (including ice caps and glaciers) that are derived using satellite gravimetry data from the GRACE(-FO) missions. Data are available as time series for the whole ice sheet, as well as at the basin level, but no gridded data are provided. Data are available since 2002 with units in Gt (Gigatonnes). For a more detailed description of the data acquisition and processing methods, we refer to the [documentation on the CDS](https://cds.climate.copernicus.eu/datasets/satellite-ice-sheet-mass-balance?tab=overview) and the [ECMWF Confluence Wiki](https://confluence.ecmwf.int/pages/viewpage.action?pageId=355348671) (Copernicus Knowledge Base).\n\nThe mass changes and their errors derived from GRACE(-FO) are expressed in Gt (Gigatonnes). It can also be translated into a volume, for example one Gt of water (density 1000 kg/m$^3$) is exactly one km$^3$, while one Gt of ice (density 917 kg/m$^3$) in volume becomes 1.091 km$^3$ of ice. GRACE(-FO) data are in fact considered to be the sum of mass changes driven by changing rates of solid ice discharge (i.e. the flux across the grounding line $D$) and mass changes driven by changing rates of ablation and accumulation (mainly the surface mass balance or $SMB$). What GRACE(-FO) actually detects are patterns of mass redistribution, indicating that a material should be redistributed (e.g. by ice dynamics or liquid discharge) in order for GRACE(-FO) to be able to detect gravity anomalies and the corresponding mass changes over certain locations. In that sense, an amount of melted ice being replaced by its mass-equivalent amount of meltwater at the same location would result in zero mass change.\n\nThe main advantage of GRACE(-FO) is that all processes directly contributing to ice sheet mass fluctuations are observed directly. Unlike the input-output method, SMB is therefore not involved explicitely in processing and it is not separated from ice flow dynamics. Moreover, gaps in the time series do not affect linear trend estimation as mass changes occurring during these gaps are still included in the subsequent solution. The main drawbacks of this technique are the large footprint of data acquisition and the fact that corrections must be made for mass redistribution in the atmosphere, ocean, soil and solid earth (e.g. glacial isostatic adjustment) in order to isolate mass changes relevant for the ice sheets.", "text_with_prefix": "EQC Quality Assessment: \"Gravimetric mass balance data from satellite observations: utility analysis for mass change monitoring of the ice sheets\"\nDataset: satellite-ice-sheet-mass-balance [CDS]\nAspect: trend-assessment_q02 | Category: Satellite_ECVs\nSection: Gravimetric mass balance data from satellite observations: utility analysis for mass change monitoring of the ice sheets > Methodology > Dataset description\n---\nThe total mass balance of an ice sheet is the difference between mass gained (mainly from snow accumulation) and mass lost (by ablation or solid ice discharge across the grounding line), which is the same as the net mass change of the ice sheet. Remote sensing techniques, such as the use of satellites, are an important feature to derive and study the mass changes of the ice sheets. The ice sheet mass change dataset provides monthly gravimetric mass balance (GMB) values and their uncertainty for the Greenland (GrIS) and also the Antarctic ice sheet (AIS). The data represent a time series of the cumulative mass changes (mass anomalies) of the ice above buoyancy of the ice sheets and their basins (including ice caps and glaciers) that are derived using satellite gravimetry data from the GRACE(-FO) missions. Data are available as time series for the whole ice sheet, as well as at the basin level, but no gridded data are provided. Data are available since 2002 with units in Gt (Gigatonnes). For a more detailed description of the data acquisition and processing methods, we refer to the [documentation on the CDS](https://cds.climate.copernicus.eu/datasets/satellite-ice-sheet-mass-balance?tab=overview) and the [ECMWF Confluence Wiki](https://confluence.ecmwf.int/pages/viewpage.action?pageId=355348671) (Copernicus Knowledge Base).\n\nThe mass changes and their errors derived from GRACE(-FO) are expressed in Gt (Gigatonnes). It can also be translated into a volume, for example one Gt of water (density 1000 kg/m$^3$) is exactly one km$^3$, while one Gt of ice (density 917 kg/m$^3$) in volume becomes 1.091 km$^3$ of ice. GRACE(-FO) data are in fact considered to be the sum of mass changes driven by changing rates of solid ice discharge (i.e. the flux across the grounding line $D$) and mass changes driven by changing rates of ablation and accumulation (mainly the surface mass balance or $SMB$). What GRACE(-FO) actually detects are patterns of mass redistribution, indicating that a material should be redistributed (e.g. by ice dynamics or liquid discharge) in order for GRACE(-FO) to be able to detect gravity anomalies and the corresponding mass changes over certain locations. In that sense, an amount of melted ice being replaced by its mass-equivalent amount of meltwater at the same location would result in zero mass change.\n\nThe main advantage of GRACE(-FO) is that all processes directly contributing to ice sheet mass fluctuations are observed directly. Unlike the input-output method, SMB is therefore not involved explicitely in processing and it is not separated from ice flow dynamics. Moreover, gaps in the time series do not affect linear trend estimation as mass changes occurring during these gaps are still included in the subsequent solution. The main drawbacks of this technique are the large footprint of data acquisition and the fact that corrections must be made for mass redistribution in the atmosphere, ocean, soil and solid earth (e.g. glacial isostatic adjustment) in order to isolate mass changes relevant for the ice sheets."} {"chunk_id": "satellite_satellite-ice-sheet-mass-balance_trend-assessment_q02__a94bf3b53c45", "report_id": "satellite_satellite-ice-sheet-mass-balance_trend-assessment_q02", "dataset_id": "satellite-ice-sheet-mass-balance", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q02", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Gravimetric mass balance data from satellite observations: utility analysis for mass change monitoring of the ice sheets > Methodology > Structure and (sub)sections", "title": "Gravimetric mass balance data from satellite observations: utility analysis for mass change monitoring of the ice sheets", "chunk_index": 4, "token_count": 185, "text_raw": "**[](section-1)**\n* [](section-1-1)\n* [](section-1-2)\n* [](section-1-3)\n* [](section-1-4)\n\n**[](section-2)**\n* [](section-2-1)\n* [](section-2-2)\n\n**[](section-3)**\n* [](section-3-1)\n* [](section-3-2)\n\n**[](section-4)**", "text_with_prefix": "EQC Quality Assessment: \"Gravimetric mass balance data from satellite observations: utility analysis for mass change monitoring of the ice sheets\"\nDataset: satellite-ice-sheet-mass-balance [CDS]\nAspect: trend-assessment_q02 | Category: Satellite_ECVs\nSection: Gravimetric mass balance data from satellite observations: utility analysis for mass change monitoring of the ice sheets > Methodology > Structure and (sub)sections\n---\n**[](section-1)**\n* [](section-1-1)\n* [](section-1-2)\n* [](section-1-3)\n* [](section-1-4)\n\n**[](section-2)**\n* [](section-2-1)\n* [](section-2-2)\n\n**[](section-3)**\n* [](section-3-1)\n* [](section-3-2)\n\n**[](section-4)**"} {"chunk_id": "satellite_satellite-ice-sheet-mass-balance_trend-assessment_q02__f3e7bed0ee14", "report_id": "satellite_satellite-ice-sheet-mass-balance_trend-assessment_q02", "dataset_id": "satellite-ice-sheet-mass-balance", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q02", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Gravimetric mass balance data from satellite observations: utility analysis for mass change monitoring of the ice sheets > Analysis and results > 1. Data preparation and processing > 1.2 Define parameters and download", "title": "Gravimetric mass balance data from satellite observations: utility analysis for mass change monitoring of the ice sheets", "chunk_index": 5, "token_count": 204, "text_raw": "Then we define the parameters, i.e. for which ice sheet (or which basins of these ice sheets) we want the mass change data to be extracted:\n\nThen we define requests for download from the CDS and download and transform the glacier mass change data.\n\n```text\n100%|██████████| 1/1 [00:00<00:00, 28.68it/s]\n```\n\n```text\nDownloading done.\n```\n\n(section-1-3)=", "text_with_prefix": "EQC Quality Assessment: \"Gravimetric mass balance data from satellite observations: utility analysis for mass change monitoring of the ice sheets\"\nDataset: satellite-ice-sheet-mass-balance [CDS]\nAspect: trend-assessment_q02 | Category: Satellite_ECVs\nSection: Gravimetric mass balance data from satellite observations: utility analysis for mass change monitoring of the ice sheets > Analysis and results > 1. Data preparation and processing > 1.2 Define parameters and download\n---\nThen we define the parameters, i.e. for which ice sheet (or which basins of these ice sheets) we want the mass change data to be extracted:\n\nThen we define requests for download from the CDS and download and transform the glacier mass change data.\n\n```text\n100%|██████████| 1/1 [00:00<00:00, 28.68it/s]\n```\n\n```text\nDownloading done.\n```\n\n(section-1-3)="} {"chunk_id": "satellite_satellite-ice-sheet-mass-balance_trend-assessment_q02__ba409c95e92b", "report_id": "satellite_satellite-ice-sheet-mass-balance_trend-assessment_q02", "dataset_id": "satellite-ice-sheet-mass-balance", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q02", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Gravimetric mass balance data from satellite observations: utility analysis for mass change monitoring of the ice sheets > Analysis and results > 1. Data preparation and processing > 1.3 Display and inspect data", "title": "Gravimetric mass balance data from satellite observations: utility analysis for mass change monitoring of the ice sheets", "chunk_index": 6, "token_count": 534, "text_raw": "Let us inspect the data:\n\n```text\n Size: 3kB\nDimensions: (time: 216)\nCoordinates:\n * time (time) datetime64[ns] 2kB 2002-04-16T20:23:54.375000 ... 202...\nData variables:\n GrIS_total (time) float32 864B 1.057e+03 1.123e+03 ... -3.762e+03\n AntIS_total (time) float32 864B 345.6 488.9 467.3 ... -910.0 -843.2 nan\nAttributes:\n Title: GMB for Greenland and Antarctica ice sheets from th...\n institution: DTU Space - Geodesy and Earth Observations\n reference: Baratta et al. (2016), Groh and Horwart (2016)\n file_creation_date: Tue May 16 09:49:32 2023\n region: Greenland and Antarctica\n missions_used: GRACE and and GRACE-FO\n time_coverage_start: Apr-2002\n time_coverage_end: Dec-2022\n Tracking_id: ab2360a4-82d5-42f3-babb-90ce976a8a8e\n netCDF_version: NETCDF4\n product_version: 4.0\n Summary: This data set is prepared for the C3S project, and ...\n```\n\nIt is a dataset that consists of time series data, containing cumulative values of the total ice sheet mass change of, in this case, the entire Greenland Ice Sheet (`GrIS_total`) and the entire Antarctic Ice Sheet (`AntIS_total`) since 2002. Mass changes are expressed in units of Gt and the time period between two measurements is variable. Note that no gridded data are given in this dataset, and hence no spatial resolution can be derived.\n\n(section-1-4)=", "text_with_prefix": "EQC Quality Assessment: \"Gravimetric mass balance data from satellite observations: utility analysis for mass change monitoring of the ice sheets\"\nDataset: satellite-ice-sheet-mass-balance [CDS]\nAspect: trend-assessment_q02 | Category: Satellite_ECVs\nSection: Gravimetric mass balance data from satellite observations: utility analysis for mass change monitoring of the ice sheets > Analysis and results > 1. Data preparation and processing > 1.3 Display and inspect data\n---\nLet us inspect the data:\n\n```text\n Size: 3kB\nDimensions: (time: 216)\nCoordinates:\n * time (time) datetime64[ns] 2kB 2002-04-16T20:23:54.375000 ... 202...\nData variables:\n GrIS_total (time) float32 864B 1.057e+03 1.123e+03 ... -3.762e+03\n AntIS_total (time) float32 864B 345.6 488.9 467.3 ... -910.0 -843.2 nan\nAttributes:\n Title: GMB for Greenland and Antarctica ice sheets from th...\n institution: DTU Space - Geodesy and Earth Observations\n reference: Baratta et al. (2016), Groh and Horwart (2016)\n file_creation_date: Tue May 16 09:49:32 2023\n region: Greenland and Antarctica\n missions_used: GRACE and and GRACE-FO\n time_coverage_start: Apr-2002\n time_coverage_end: Dec-2022\n Tracking_id: ab2360a4-82d5-42f3-babb-90ce976a8a8e\n netCDF_version: NETCDF4\n product_version: 4.0\n Summary: This data set is prepared for the C3S project, and ...\n```\n\nIt is a dataset that consists of time series data, containing cumulative values of the total ice sheet mass change of, in this case, the entire Greenland Ice Sheet (`GrIS_total`) and the entire Antarctic Ice Sheet (`AntIS_total`) since 2002. Mass changes are expressed in units of Gt and the time period between two measurements is variable. Note that no gridded data are given in this dataset, and hence no spatial resolution can be derived.\n\n(section-1-4)="} {"chunk_id": "satellite_satellite-ice-sheet-mass-balance_trend-assessment_q02__6be451412379", "report_id": "satellite_satellite-ice-sheet-mass-balance_trend-assessment_q02", "dataset_id": "satellite-ice-sheet-mass-balance", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q02", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Gravimetric mass balance data from satellite observations: utility analysis for mass change monitoring of the ice sheets > Analysis and results > 1. Data preparation and processing > 1.4 Data handling and creating functions", "title": "Gravimetric mass balance data from satellite observations: utility analysis for mass change monitoring of the ice sheets", "chunk_index": 7, "token_count": 197, "text_raw": "Let us perform some data handling before getting started with the analysis. We also define a plotting function to visualize the time series:\n\nDefine specific names for the first and second variables\nDetermine the number of columns needed based on the number of variables\nConvert time to pandas datetime for easy handling\nIdentify large gaps and shade them\nDataset processing\nSet custom attributes for variables\n\nWith everything ready, let us now start with the analysis:\n\n(section-2)=", "text_with_prefix": "EQC Quality Assessment: \"Gravimetric mass balance data from satellite observations: utility analysis for mass change monitoring of the ice sheets\"\nDataset: satellite-ice-sheet-mass-balance [CDS]\nAspect: trend-assessment_q02 | Category: Satellite_ECVs\nSection: Gravimetric mass balance data from satellite observations: utility analysis for mass change monitoring of the ice sheets > Analysis and results > 1. Data preparation and processing > 1.4 Data handling and creating functions\n---\nLet us perform some data handling before getting started with the analysis. We also define a plotting function to visualize the time series:\n\nDefine specific names for the first and second variables\nDetermine the number of columns needed based on the number of variables\nConvert time to pandas datetime for easy handling\nIdentify large gaps and shade them\nDataset processing\nSet custom attributes for variables\n\nWith everything ready, let us now start with the analysis:\n\n(section-2)="} {"chunk_id": "satellite_satellite-ice-sheet-mass-balance_trend-assessment_q02__44e0d8955938", "report_id": "satellite_satellite-ice-sheet-mass-balance_trend-assessment_q02", "dataset_id": "satellite-ice-sheet-mass-balance", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q02", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Gravimetric mass balance data from satellite observations: utility analysis for mass change monitoring of the ice sheets > Analysis and results > 2. Quantifying Greenland and Antarctic Ice Sheet mass changes over time > 2.1 Time series of cumulative ice sheet mass changes", "title": "Gravimetric mass balance data from satellite observations: utility analysis for mass change monitoring of the ice sheets", "chunk_index": 8, "token_count": 462, "text_raw": "We begin by plotting the Greenland and Antarctic Ice Sheet cumulative mass change $M_{GRACE}$ between the begin and end period with the defined plotting function, where the shading in the upper part of the plot indicates when the time between two consecutive measurements is more than 1 month:\n\n*Figure 1. Greenland (top) and Antarctic (bottom) ice mass changes from GRACE(-FO), where shaded intervals depict data gaps longer than 1 month. Note the different range of the y-axis.*\n\nThe provided graphs show the cumulative mass change of the Greenland Ice Sheet and the Antarctic Ice Sheet since 2002, as measured by GRACE(-FO) satellites. The GrIS shows a clear and consistent decline in mass, having lost approximately 5000 gigatonnes (Gt) over the period. The AIS mass change graph displays more pronounced fluctuations. For Greenland, mass loss is dominated by meltwater runoff, combined with the solid ice discharge towards the ocean across the grounding line of outlet glaciers. For Antarctica, mass loss primarily occurs through the discharge of ice across the grounding line, with surface melting playing a minor role. Here, interactions between the ice sheet and the warming ocean, combined with dynamic responses within the ice sheet itself (e.g. an acceleration of outlet glaciers due to a reduced buttressing from the ice shelves), are central to understanding why Antarctica's mass is currently decreasing [[1](https://doi.org/10.1007/s10712-016-9398-7), [2](https://doi.org/10.3390/geosciences9100415), [3](https://doi.org/10.1126/science.1178176)].\n\n(section-2-2)=", "text_with_prefix": "EQC Quality Assessment: \"Gravimetric mass balance data from satellite observations: utility analysis for mass change monitoring of the ice sheets\"\nDataset: satellite-ice-sheet-mass-balance [CDS]\nAspect: trend-assessment_q02 | Category: Satellite_ECVs\nSection: Gravimetric mass balance data from satellite observations: utility analysis for mass change monitoring of the ice sheets > Analysis and results > 2. Quantifying Greenland and Antarctic Ice Sheet mass changes over time > 2.1 Time series of cumulative ice sheet mass changes\n---\nWe begin by plotting the Greenland and Antarctic Ice Sheet cumulative mass change $M_{GRACE}$ between the begin and end period with the defined plotting function, where the shading in the upper part of the plot indicates when the time between two consecutive measurements is more than 1 month:\n\n*Figure 1. Greenland (top) and Antarctic (bottom) ice mass changes from GRACE(-FO), where shaded intervals depict data gaps longer than 1 month. Note the different range of the y-axis.*\n\nThe provided graphs show the cumulative mass change of the Greenland Ice Sheet and the Antarctic Ice Sheet since 2002, as measured by GRACE(-FO) satellites. The GrIS shows a clear and consistent decline in mass, having lost approximately 5000 gigatonnes (Gt) over the period. The AIS mass change graph displays more pronounced fluctuations. For Greenland, mass loss is dominated by meltwater runoff, combined with the solid ice discharge towards the ocean across the grounding line of outlet glaciers. For Antarctica, mass loss primarily occurs through the discharge of ice across the grounding line, with surface melting playing a minor role. Here, interactions between the ice sheet and the warming ocean, combined with dynamic responses within the ice sheet itself (e.g. an acceleration of outlet glaciers due to a reduced buttressing from the ice shelves), are central to understanding why Antarctica's mass is currently decreasing [[1](https://doi.org/10.1007/s10712-016-9398-7), [2](https://doi.org/10.3390/geosciences9100415), [3](https://doi.org/10.1126/science.1178176)].\n\n(section-2-2)="} {"chunk_id": "satellite_satellite-ice-sheet-mass-balance_trend-assessment_q02__046c7a1319c6", "report_id": "satellite_satellite-ice-sheet-mass-balance_trend-assessment_q02", "dataset_id": "satellite-ice-sheet-mass-balance", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q02", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Gravimetric mass balance data from satellite observations: utility analysis for mass change monitoring of the ice sheets > Analysis and results > 2. Quantifying Greenland and Antarctic Ice Sheet mass changes over time > 2.2 Spatio-temporal resolution and coverage of the GRACE(-FO) mass change estimates", "title": "Gravimetric mass balance data from satellite observations: utility analysis for mass change monitoring of the ice sheets", "chunk_index": 9, "token_count": 1046, "text_raw": "Now that we have visualized the temporal patterns of the mass changes, we can investigate the spatio-temporal resolution and extent of the GrIS and AIS mass changes. Let us begin by zooming in more on the temporal data gaps (coverage) and the temporal resolution in the time series:\n\nGet time gaps\nCreate the plot\n\n*Figure 2. Time period between two consecutive GRACE(-FO) monthly solutions in the ice sheet GMB dataset.*\n\nThe plot displays the time gaps in the GRACE(-FO) data collection over time, which are the same for the GrIS and the AIS, highlighting periods of consistent monthly data acquisition and (significant) data gaps. Generally, a temporal resolution of approximately 1 month (which is mostly the case) is observed, aligning with the optimal requirement proposed by GCOS [[5](https://library.wmo.int/records/item/58111-the-2022-gcos-ecvs-requirements-gcos-245)]. More gaps in GRACE data, however, arise after 2011, which are due to sensor degradation. The most prominent feature is the large time gap around 2017-2018, where the gap extends up to 12 months. This significant gap corresponds to the end of the original GRACE mission and the transition period before the launch of the GRACE Follow-On (GRACE-FO) mission. During this period, there was an inability to acquire data, as the original GRACE satellites were decommissioned and the new GRACE-FO satellites were not yet operational. Because there is no systematic bias between the GRACE and GRACE-FO time series [[4](https://doi.org/10.1029/2020GL087291)], this does not impact the overall trend and magnitude of the time series after 2018. In fact, for linear trend estimation, the presence of gaps (even long ones) scattered over the entire time series are almost irrelevant, as mass changes occurring during these gaps are still included in the subsequent solution. Long-term trend stability is therefore a major advantage of GMB data, as compared to the altimetry and input-output methods for ice mass change estimations.\n\nLet us have the amount of gaps quantified to get an idea of the temporal data availability:\n\n```text\nThe start date of the time series is 16/04/2002\nThe end date of the time series is 17/12/2022\nThe amount of months with mass change measurements that is expected between these two dates is 248\nThe amount of months with mass change measurements that is present between these two dates is 213\nFor a consistent monthly temporal resolution, the amount of months with missing data is 14.11%.\n```\n\nConcerning spatial coverage, a significant challenge in analyzing the ice sheet's mass changes (especially for the GrIS) is deciding whether to include the peripheral glaciers and ice caps. The GRACE(-FO) GMB data include all ice masses in Greenland (and Antarctica), due to the coarse spatial resolution of several hundred kilometers during the data acquisition, which prevents differentiation between closely situated ice bodies. As a result, the mass changes of Greenland's peripheral glaciers and ice caps, which contribute approximately 30-35 Gt/year [[6](https://essd.copernicus.org/articles/15/1597/2023/)], are included in these measurements. To exclude these peripheral areas from the GRACE(-FO) time series, a scaling factor of 0.84 is often used for the GrIS [[7](https://doi.org/10.1016/j.rse.2015.06.016)]. It must be furthermore noted that only mass changes of ice above buoyancy are considered in the GMB dataset (of which the impact is, however, relatively limited for the GrIS but can be significant for the AIS, due to the presence of more floating ice and ice grounded below present-day sea level).\n\nThe lack of downloadable gridded data further complicates the spatial resolution issue (in fact, since there are no gridded data, the term \"spatial resolution\" is not applicable for this C3S dataset), although such pixel-by-pixel data are available from other sources like the TU Dresden website. The currently available data are provided as time series for the entire ice sheet's and their basins, which is too coarse to meet GCOS requirements and poses challenges, for example for ice sheet models that require gridded inputs.\n\n(section-3)=", "text_with_prefix": "EQC Quality Assessment: \"Gravimetric mass balance data from satellite observations: utility analysis for mass change monitoring of the ice sheets\"\nDataset: satellite-ice-sheet-mass-balance [CDS]\nAspect: trend-assessment_q02 | Category: Satellite_ECVs\nSection: Gravimetric mass balance data from satellite observations: utility analysis for mass change monitoring of the ice sheets > Analysis and results > 2. Quantifying Greenland and Antarctic Ice Sheet mass changes over time > 2.2 Spatio-temporal resolution and coverage of the GRACE(-FO) mass change estimates\n---\nNow that we have visualized the temporal patterns of the mass changes, we can investigate the spatio-temporal resolution and extent of the GrIS and AIS mass changes. Let us begin by zooming in more on the temporal data gaps (coverage) and the temporal resolution in the time series:\n\nGet time gaps\nCreate the plot\n\n*Figure 2. Time period between two consecutive GRACE(-FO) monthly solutions in the ice sheet GMB dataset.*\n\nThe plot displays the time gaps in the GRACE(-FO) data collection over time, which are the same for the GrIS and the AIS, highlighting periods of consistent monthly data acquisition and (significant) data gaps. Generally, a temporal resolution of approximately 1 month (which is mostly the case) is observed, aligning with the optimal requirement proposed by GCOS [[5](https://library.wmo.int/records/item/58111-the-2022-gcos-ecvs-requirements-gcos-245)]. More gaps in GRACE data, however, arise after 2011, which are due to sensor degradation. The most prominent feature is the large time gap around 2017-2018, where the gap extends up to 12 months. This significant gap corresponds to the end of the original GRACE mission and the transition period before the launch of the GRACE Follow-On (GRACE-FO) mission. During this period, there was an inability to acquire data, as the original GRACE satellites were decommissioned and the new GRACE-FO satellites were not yet operational. Because there is no systematic bias between the GRACE and GRACE-FO time series [[4](https://doi.org/10.1029/2020GL087291)], this does not impact the overall trend and magnitude of the time series after 2018. In fact, for linear trend estimation, the presence of gaps (even long ones) scattered over the entire time series are almost irrelevant, as mass changes occurring during these gaps are still included in the subsequent solution. Long-term trend stability is therefore a major advantage of GMB data, as compared to the altimetry and input-output methods for ice mass change estimations.\n\nLet us have the amount of gaps quantified to get an idea of the temporal data availability:\n\n```text\nThe start date of the time series is 16/04/2002\nThe end date of the time series is 17/12/2022\nThe amount of months with mass change measurements that is expected between these two dates is 248\nThe amount of months with mass change measurements that is present between these two dates is 213\nFor a consistent monthly temporal resolution, the amount of months with missing data is 14.11%.\n```\n\nConcerning spatial coverage, a significant challenge in analyzing the ice sheet's mass changes (especially for the GrIS) is deciding whether to include the peripheral glaciers and ice caps. The GRACE(-FO) GMB data include all ice masses in Greenland (and Antarctica), due to the coarse spatial resolution of several hundred kilometers during the data acquisition, which prevents differentiation between closely situated ice bodies. As a result, the mass changes of Greenland's peripheral glaciers and ice caps, which contribute approximately 30-35 Gt/year [[6](https://essd.copernicus.org/articles/15/1597/2023/)], are included in these measurements. To exclude these peripheral areas from the GRACE(-FO) time series, a scaling factor of 0.84 is often used for the GrIS [[7](https://doi.org/10.1016/j.rse.2015.06.016)]. It must be furthermore noted that only mass changes of ice above buoyancy are considered in the GMB dataset (of which the impact is, however, relatively limited for the GrIS but can be significant for the AIS, due to the presence of more floating ice and ice grounded below present-day sea level).\n\nThe lack of downloadable gridded data further complicates the spatial resolution issue (in fact, since there are no gridded data, the term \"spatial resolution\" is not applicable for this C3S dataset), although such pixel-by-pixel data are available from other sources like the TU Dresden website. The currently available data are provided as time series for the entire ice sheet's and their basins, which is too coarse to meet GCOS requirements and poses challenges, for example for ice sheet models that require gridded inputs.\n\n(section-3)="} {"chunk_id": "satellite_satellite-ice-sheet-mass-balance_trend-assessment_q02__ec820a57e120", "report_id": "satellite_satellite-ice-sheet-mass-balance_trend-assessment_q02", "dataset_id": "satellite-ice-sheet-mass-balance", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q02", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Gravimetric mass balance data from satellite observations: utility analysis for mass change monitoring of the ice sheets > Analysis and results > 3. Ice sheet mass change trends and sea level contribution over time > 3.1 Linear and quadratic ice sheet-wide mass change trends", "title": "Gravimetric mass balance data from satellite observations: utility analysis for mass change monitoring of the ice sheets", "chunk_index": 10, "token_count": 437, "text_raw": "With the imformation above, let us now go on and calculate linear and quadratic trends for the ice sheet mass change product:\n\nCompute coefficients\nPlot trends and print stats\n\n```text\nCumulative mass change of the Greenland Ice Sheet from GRACE(-FO):\n\tThe slope of the ice sheet mass change is -251.329 Gt/yr.\n\tThe trend is significant at an alpha level of 0.05, i.e. a monotonic trend is present.\n\tThe acceleration of the ice sheet mass change is 3.240 Gt/yr^2.\nCumulative mass change of the Antarctic Ice Sheet from GRACE(-FO):\n\tThe slope of the ice sheet mass change is -89.873 Gt/yr.\n\tThe trend is significant at an alpha level of 0.05, i.e. a monotonic trend is present.\n\tThe acceleration of the ice sheet mass change is -2.354 Gt/yr^2.\n```\n\n*Figure 3. Linear and quadratic trends of ice sheet mass changes for the (above) GrIS and (below) AIS from the ice sheet GMB dataset.*\n\nThe GRACE(-FO) data plots illustrate the cumulative mass changes of the Greenland and Antarctic Ice Sheets over time. Both the Greenland and Antarctic Ice Sheets exhibit significant ongoing mass loss, with the statistical significance of the trends underscoring the robustness of these findings. The linear and quadratic trend lines, depicted in red and green respectively, provide a clear visual representation of these trends, highlighting the critical changes occurring in the polar regions.\n\n(section-3-2)=", "text_with_prefix": "EQC Quality Assessment: \"Gravimetric mass balance data from satellite observations: utility analysis for mass change monitoring of the ice sheets\"\nDataset: satellite-ice-sheet-mass-balance [CDS]\nAspect: trend-assessment_q02 | Category: Satellite_ECVs\nSection: Gravimetric mass balance data from satellite observations: utility analysis for mass change monitoring of the ice sheets > Analysis and results > 3. Ice sheet mass change trends and sea level contribution over time > 3.1 Linear and quadratic ice sheet-wide mass change trends\n---\nWith the imformation above, let us now go on and calculate linear and quadratic trends for the ice sheet mass change product:\n\nCompute coefficients\nPlot trends and print stats\n\n```text\nCumulative mass change of the Greenland Ice Sheet from GRACE(-FO):\n\tThe slope of the ice sheet mass change is -251.329 Gt/yr.\n\tThe trend is significant at an alpha level of 0.05, i.e. a monotonic trend is present.\n\tThe acceleration of the ice sheet mass change is 3.240 Gt/yr^2.\nCumulative mass change of the Antarctic Ice Sheet from GRACE(-FO):\n\tThe slope of the ice sheet mass change is -89.873 Gt/yr.\n\tThe trend is significant at an alpha level of 0.05, i.e. a monotonic trend is present.\n\tThe acceleration of the ice sheet mass change is -2.354 Gt/yr^2.\n```\n\n*Figure 3. Linear and quadratic trends of ice sheet mass changes for the (above) GrIS and (below) AIS from the ice sheet GMB dataset.*\n\nThe GRACE(-FO) data plots illustrate the cumulative mass changes of the Greenland and Antarctic Ice Sheets over time. Both the Greenland and Antarctic Ice Sheets exhibit significant ongoing mass loss, with the statistical significance of the trends underscoring the robustness of these findings. The linear and quadratic trend lines, depicted in red and green respectively, provide a clear visual representation of these trends, highlighting the critical changes occurring in the polar regions.\n\n(section-3-2)="} {"chunk_id": "satellite_satellite-ice-sheet-mass-balance_trend-assessment_q02__7006156272a7", "report_id": "satellite_satellite-ice-sheet-mass-balance_trend-assessment_q02", "dataset_id": "satellite-ice-sheet-mass-balance", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q02", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Gravimetric mass balance data from satellite observations: utility analysis for mass change monitoring of the ice sheets > Analysis and results > 3. Ice sheet mass change trends and sea level contribution over time > 3.2 Quantifying ice sheet-related global sea level contributions", "title": "Gravimetric mass balance data from satellite observations: utility analysis for mass change monitoring of the ice sheets", "chunk_index": 11, "token_count": 772, "text_raw": "We furthermore note again that GRACE(-FO) cannot detect changes in floating ice and ice below buoyancy for ice grounded below sea level. Mass change processes occurring beyond the grounding line, such as a retreating calving front, ocean-induced melt beneath the ice shelves, as well as changes to the part of the ice that is grounded below sea level (or in other words ice below buouyancy) will therefore not be detectable because it is occupied by the same mass of water. In that regard, GMB measurements by GRACE(-FO) are immediately convertible to sea level contributions. For the contributions of ice sheets, we assume that 361.8 Gt of ice will raise global sea levels by 1 mm [e.g. [11](https://doi.org/10.1038/s41467-024-45726-w)] (where 361.8 Gt of ice is equivalent to 394.67 km$^3$ ice):\n\n$\nh_{SLE} \n$\n[mm]\n$\n= -\\dfrac{M_{GRACE}}{361.8}\n$\n\nwhere $M_{GRACE}$ is the cumulative mass change of the ice sheets measured by GRACE(-FO) in Gt. Let us plot this:\n\nConvert mass change to global sea level change\nPlot the data and trends\nCompute coefficients\nPlot trends and print stats\n\n```text\nCumulative global sea level contribution of the Greenland Ice Sheet from GRACE(-FO):\n\tThe slope of the global sea level contribution is 0.695 mm/yr.\n\tThe trend is significant at an alpha level of 0.05, i.e. a monotonic trend is present.\n\tThe acceleration of the global sea level contribution is -0.009 mm/yr^2.\nCumulative global sea level contribution of the Antarctic Ice Sheet from GRACE(-FO):\n\tThe slope of the global sea level contribution is 0.248 mm/yr.\n\tThe trend is significant at an alpha level of 0.05, i.e. a monotonic trend is present.\n\tThe acceleration of the global sea level contribution is 0.007 mm/yr^2.\n```\n\n*Figure 4. Contribution of ice sheet mass changes to global sea level for the (above) GrIS and (below) AIS from the ice sheet GMB dataset.*\n\nThe linear trends and their significance show a consistent ongoing increase in sea level contributions, with Greenland contributing more significantly compared to Antarctica. The trends and their magnitudes observed in the GRACE(-FO) data for both the Greenland and Antarctic Ice Sheets are broadly consistent with the findings reported in the literature from other mass change estimation methods (e.g. [[9](https://iopscience.iop.org/article/10.1088/1748-9326/aac2f0/meta), [12](https://doi.org/10.1038/s41586-019-1855-2), [13](https://doi.org/10.1038/s41586-018-0179-y)]), highlighting the reliability and quality-richness of GRACE(-FO) data with respect to capturing significant ice mass changes and their impact on global sea level rise.\n\n(section-4)=", "text_with_prefix": "EQC Quality Assessment: \"Gravimetric mass balance data from satellite observations: utility analysis for mass change monitoring of the ice sheets\"\nDataset: satellite-ice-sheet-mass-balance [CDS]\nAspect: trend-assessment_q02 | Category: Satellite_ECVs\nSection: Gravimetric mass balance data from satellite observations: utility analysis for mass change monitoring of the ice sheets > Analysis and results > 3. Ice sheet mass change trends and sea level contribution over time > 3.2 Quantifying ice sheet-related global sea level contributions\n---\nWe furthermore note again that GRACE(-FO) cannot detect changes in floating ice and ice below buoyancy for ice grounded below sea level. Mass change processes occurring beyond the grounding line, such as a retreating calving front, ocean-induced melt beneath the ice shelves, as well as changes to the part of the ice that is grounded below sea level (or in other words ice below buouyancy) will therefore not be detectable because it is occupied by the same mass of water. In that regard, GMB measurements by GRACE(-FO) are immediately convertible to sea level contributions. For the contributions of ice sheets, we assume that 361.8 Gt of ice will raise global sea levels by 1 mm [e.g. [11](https://doi.org/10.1038/s41467-024-45726-w)] (where 361.8 Gt of ice is equivalent to 394.67 km$^3$ ice):\n\n$\nh_{SLE} \n$\n[mm]\n$\n= -\\dfrac{M_{GRACE}}{361.8}\n$\n\nwhere $M_{GRACE}$ is the cumulative mass change of the ice sheets measured by GRACE(-FO) in Gt. Let us plot this:\n\nConvert mass change to global sea level change\nPlot the data and trends\nCompute coefficients\nPlot trends and print stats\n\n```text\nCumulative global sea level contribution of the Greenland Ice Sheet from GRACE(-FO):\n\tThe slope of the global sea level contribution is 0.695 mm/yr.\n\tThe trend is significant at an alpha level of 0.05, i.e. a monotonic trend is present.\n\tThe acceleration of the global sea level contribution is -0.009 mm/yr^2.\nCumulative global sea level contribution of the Antarctic Ice Sheet from GRACE(-FO):\n\tThe slope of the global sea level contribution is 0.248 mm/yr.\n\tThe trend is significant at an alpha level of 0.05, i.e. a monotonic trend is present.\n\tThe acceleration of the global sea level contribution is 0.007 mm/yr^2.\n```\n\n*Figure 4. Contribution of ice sheet mass changes to global sea level for the (above) GrIS and (below) AIS from the ice sheet GMB dataset.*\n\nThe linear trends and their significance show a consistent ongoing increase in sea level contributions, with Greenland contributing more significantly compared to Antarctica. The trends and their magnitudes observed in the GRACE(-FO) data for both the Greenland and Antarctic Ice Sheets are broadly consistent with the findings reported in the literature from other mass change estimation methods (e.g. [[9](https://iopscience.iop.org/article/10.1088/1748-9326/aac2f0/meta), [12](https://doi.org/10.1038/s41586-019-1855-2), [13](https://doi.org/10.1038/s41586-018-0179-y)]), highlighting the reliability and quality-richness of GRACE(-FO) data with respect to capturing significant ice mass changes and their impact on global sea level rise.\n\n(section-4)="} {"chunk_id": "satellite_satellite-ice-sheet-mass-balance_trend-assessment_q02__16bd5a279125", "report_id": "satellite_satellite-ice-sheet-mass-balance_trend-assessment_q02", "dataset_id": "satellite-ice-sheet-mass-balance", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q02", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Gravimetric mass balance data from satellite observations: utility analysis for mass change monitoring of the ice sheets > Analysis and results > 4. Short summary and take-home messages", "title": "Gravimetric mass balance data from satellite observations: utility analysis for mass change monitoring of the ice sheets", "chunk_index": 12, "token_count": 834, "text_raw": "GRACE(-FO) data provide a valuable long-term dataset for detecting trends in ice sheet mass changes, making them a reliable indicator of climate change. With over 20 years of (mostly) continuous data, the dataset enables the assessment of long-term variations in ice sheet mass balance at the basin and ice sheet-wide scales. Unlike other remote sensing methods that infer mass changes indirectly, GRACE(-FO) directly measures gravitational anomalies, capturing mass redistribution patterns across both the GrIS and AIS. Although the IPCC standard for climate normals is 30 years, the dataset's continuity from GRACE (2002–2017) to GRACE-FO (2018–present) ensures that clear temporal patterns can be identified, making it a crucial tool for climate and global water cycle studies.\n\nHowever, GRACE(-FO) has limitations, including a coarse spatial resolution (several hundred kilometers), which restricts its ability to resolve small-scale variations and separate closely situated ice bodies (e.g. Greenland and its peripheral glaciers/ice caps, including the adjacent Canadian ice caps). Additional uncertainties arise mainly from complex processes such as signal leakage and geophysical corrections [[10](https://tc.copernicus.org/articles/7/1411/2013/)]. The dataset also lacks pixel-level (gridded) mass change and error estimates, limiting its applicability to grid-based assessments. Additionally, users should also acknowledge what GRACE(-FO) ice mass changes represent. GRACE(-FO) only considers patterns of mass redistribution by ice masses above buoyancy (i.e. solid ice discharge and changing rates of runoff/accumulation), meaning floating ice and ice grounded below sea level do not contribute to the measured mass change. For the GrIS, peripheral glaciers and ice caps are included, whereas for the Antarctic Ice Sheet AIS, ice shelves beyond the grounding line are excluded.\n\nDespite occasional data gaps and relatively high uncertainty values in some measurements, GRACE(-FO) remains a powerful tool for quantifying the mean, variability, and trends in ice mass changes because (1) the amount of missing data is relatively limited and these data gaps do not impact the values of consecutive cumulative mass anomalies, (2) the temporal resolution is mostly consistent at 1-monthly spaced time intervals, (3) the number of consecutive years is generally sufficient to filter out inter and intrayearly variability, and (4) the presence of data gaps in the time series do not impact the overall trend and magntiude of the mass changes and sea level contributions (e.g. [[4](https://doi.org/10.1029/2020GL087291)]). The dataset’s ability to directly convert ice mass changes into sea level contributions makes it a convenient tool for climate change monitoring and global water cycle assessments. Trends in ice sheet mass changes and sea level contributions from the C3S product furthermore agree well with general findings in the literature from other mass change estimation methods [[8](https://www.ipcc.ch/report/ar6/wg1/downloads/report/IPCC_AR6_WGI_Chapter09.pdf), [9](https://iopscience.iop.org/article/10.1088/1748-9326/aac2f0), [12](https://doi.org/10.1038/s41586-018-0179-y), [13](https://doi.org/10.1038/s41586-019-1855-2)], further adding credibility to the data. While certain limitations are thus present, its role as a direct measurement tool for ice sheet mass loss makes it essential for evaluating ice sheet stability and contributions to sea level rise.", "text_with_prefix": "EQC Quality Assessment: \"Gravimetric mass balance data from satellite observations: utility analysis for mass change monitoring of the ice sheets\"\nDataset: satellite-ice-sheet-mass-balance [CDS]\nAspect: trend-assessment_q02 | Category: Satellite_ECVs\nSection: Gravimetric mass balance data from satellite observations: utility analysis for mass change monitoring of the ice sheets > Analysis and results > 4. Short summary and take-home messages\n---\nGRACE(-FO) data provide a valuable long-term dataset for detecting trends in ice sheet mass changes, making them a reliable indicator of climate change. With over 20 years of (mostly) continuous data, the dataset enables the assessment of long-term variations in ice sheet mass balance at the basin and ice sheet-wide scales. Unlike other remote sensing methods that infer mass changes indirectly, GRACE(-FO) directly measures gravitational anomalies, capturing mass redistribution patterns across both the GrIS and AIS. Although the IPCC standard for climate normals is 30 years, the dataset's continuity from GRACE (2002–2017) to GRACE-FO (2018–present) ensures that clear temporal patterns can be identified, making it a crucial tool for climate and global water cycle studies.\n\nHowever, GRACE(-FO) has limitations, including a coarse spatial resolution (several hundred kilometers), which restricts its ability to resolve small-scale variations and separate closely situated ice bodies (e.g. Greenland and its peripheral glaciers/ice caps, including the adjacent Canadian ice caps). Additional uncertainties arise mainly from complex processes such as signal leakage and geophysical corrections [[10](https://tc.copernicus.org/articles/7/1411/2013/)]. The dataset also lacks pixel-level (gridded) mass change and error estimates, limiting its applicability to grid-based assessments. Additionally, users should also acknowledge what GRACE(-FO) ice mass changes represent. GRACE(-FO) only considers patterns of mass redistribution by ice masses above buoyancy (i.e. solid ice discharge and changing rates of runoff/accumulation), meaning floating ice and ice grounded below sea level do not contribute to the measured mass change. For the GrIS, peripheral glaciers and ice caps are included, whereas for the Antarctic Ice Sheet AIS, ice shelves beyond the grounding line are excluded.\n\nDespite occasional data gaps and relatively high uncertainty values in some measurements, GRACE(-FO) remains a powerful tool for quantifying the mean, variability, and trends in ice mass changes because (1) the amount of missing data is relatively limited and these data gaps do not impact the values of consecutive cumulative mass anomalies, (2) the temporal resolution is mostly consistent at 1-monthly spaced time intervals, (3) the number of consecutive years is generally sufficient to filter out inter and intrayearly variability, and (4) the presence of data gaps in the time series do not impact the overall trend and magntiude of the mass changes and sea level contributions (e.g. [[4](https://doi.org/10.1029/2020GL087291)]). The dataset’s ability to directly convert ice mass changes into sea level contributions makes it a convenient tool for climate change monitoring and global water cycle assessments. Trends in ice sheet mass changes and sea level contributions from the C3S product furthermore agree well with general findings in the literature from other mass change estimation methods [[8](https://www.ipcc.ch/report/ar6/wg1/downloads/report/IPCC_AR6_WGI_Chapter09.pdf), [9](https://iopscience.iop.org/article/10.1088/1748-9326/aac2f0), [12](https://doi.org/10.1038/s41586-018-0179-y), [13](https://doi.org/10.1038/s41586-019-1855-2)], further adding credibility to the data. While certain limitations are thus present, its role as a direct measurement tool for ice sheet mass loss makes it essential for evaluating ice sheet stability and contributions to sea level rise."} {"chunk_id": "satellite_satellite-ice-sheet-mass-balance_trend-assessment_q02__4ceca643814b", "report_id": "satellite_satellite-ice-sheet-mass-balance_trend-assessment_q02", "dataset_id": "satellite-ice-sheet-mass-balance", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q02", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Gravimetric mass balance data from satellite observations: utility analysis for mass change monitoring of the ice sheets > ℹ️ If you want to know more > Key resources", "title": "Gravimetric mass balance data from satellite observations: utility analysis for mass change monitoring of the ice sheets", "chunk_index": 13, "token_count": 382, "text_raw": "- \"[Gravimetric mass balance data for the Antarctic and Greenland ice sheets from 2003 to 2022 derived from satellite observations](https://cds.climate.copernicus.eu/datasets/satellite-ice-sheet-mass-balance?tab=overview)\" on the CDS.\n- [Documentation on the CDS](https://cds.climate.copernicus.eu/datasets/satellite-ice-sheet-mass-balance?tab=overview) and the [ECMWF Confluence Wiki](https://confluence.ecmwf.int/pages/viewpage.action?pageId=355348671) (Copernicus Knowledge Base).\n- [Copernicus climate change indicators: ice sheets](https://climate.copernicus.eu/climate-indicators/ice-sheets)\n- The data portal with GMB data from the data provider (TU Dresden) [for the GrIS](https://data1.geo.tu-dresden.de/gis_gmb/index.html) and [the AIS](https://data1.geo.tu-dresden.de/gis_gmb/index.html).\n- [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control` prepared by [B-Open](https://www.bopen.eu).", "text_with_prefix": "EQC Quality Assessment: \"Gravimetric mass balance data from satellite observations: utility analysis for mass change monitoring of the ice sheets\"\nDataset: satellite-ice-sheet-mass-balance [CDS]\nAspect: trend-assessment_q02 | Category: Satellite_ECVs\nSection: Gravimetric mass balance data from satellite observations: utility analysis for mass change monitoring of the ice sheets > ℹ️ If you want to know more > Key resources\n---\n- \"[Gravimetric mass balance data for the Antarctic and Greenland ice sheets from 2003 to 2022 derived from satellite observations](https://cds.climate.copernicus.eu/datasets/satellite-ice-sheet-mass-balance?tab=overview)\" on the CDS.\n- [Documentation on the CDS](https://cds.climate.copernicus.eu/datasets/satellite-ice-sheet-mass-balance?tab=overview) and the [ECMWF Confluence Wiki](https://confluence.ecmwf.int/pages/viewpage.action?pageId=355348671) (Copernicus Knowledge Base).\n- [Copernicus climate change indicators: ice sheets](https://climate.copernicus.eu/climate-indicators/ice-sheets)\n- The data portal with GMB data from the data provider (TU Dresden) [for the GrIS](https://data1.geo.tu-dresden.de/gis_gmb/index.html) and [the AIS](https://data1.geo.tu-dresden.de/gis_gmb/index.html).\n- [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control` prepared by [B-Open](https://www.bopen.eu)."} {"chunk_id": "satellite_satellite-ice-sheet-mass-balance_trend-assessment_q02__1670332adb36", "report_id": "satellite_satellite-ice-sheet-mass-balance_trend-assessment_q02", "dataset_id": "satellite-ice-sheet-mass-balance", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q02", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Gravimetric mass balance data from satellite observations: utility analysis for mass change monitoring of the ice sheets > ℹ️ If you want to know more > References", "title": "Gravimetric mass balance data from satellite observations: utility analysis for mass change monitoring of the ice sheets", "chunk_index": 14, "token_count": 710, "text_raw": "- [[1](https://doi.org/10.1007/s10712-016-9398-7)] Forsberg, R., Sørensen, L.S. and Simonsen, S.B. (2017). Greenland and Antarctica Ice Sheet Mass Changes and Effects on Global Sea Level. Surv. Geophys., 38, 89–104. https://doi.org/10.1007/s10712-016-9398-7\n\n- [[2](https://doi.org/10.3390/geosciences9100415)] Groh, A., Horwath, M., Horvath, A., Meister, R., Sørensen, L.S., Barletta, V.R., Forsberg, R., Wouters, B., Ditmar, P., Ran, J., Klees, R., Su, X., Shang, K., Guo, J., Shum, C.K., Schrama, E., and Shepherd, A. (2019). Evaluating GRACE Mass Change Time Series for the Antarctic and Greenland Ice Sheet, Geosciences, 9(10). https://doi.org/10.3390/geosciences9100415\n\n- [[3](https://doi.org/10.5194/tc-10-1933-2016)] van den Broeke, M., Enderlin, E. M., Howat, I. M., Kuipers Munneke, P., Noël, B. P. Y., van de Berg, W. J., van Meijgaard, E., and Wouters, B. (2016). On the recent contribution of the Greenland ice sheet to sea level change, The Cryosphere, 10, 1933–1946, https://doi.org/10.5194/tc-10-1933-2016\n\n- [[4](https://doi.org/10.1029/2020GL087291)] Velicogna, I., Mohajerani, Y., A, G., Landerer, F., Mouginot, J., Noel, B., Rignot, E., Sutterley, T., van den Broeke, M., van Wessem, M., and Wiese, D. (2020). Continuity of Ice Sheet Mass Loss in Greenland and Antarctica from the GRACE and GRACE Follow-On Missions. Geophys. Res. Lett. 47. https://doi.org/10.1029/2020GL087291\n\n- [[5](https://library.wmo.int/records/item/58111-the-2022-gcos-ecvs-requirements-gcos-245)] GCOS (Global Climate Observing System) (2022). The 2022 GCOS ECVs Requirements (GCOS-245). World Meteorological Organization: Geneva, Switzerland. doi: https://library.wmo.int/idurl/4/58111", "text_with_prefix": "EQC Quality Assessment: \"Gravimetric mass balance data from satellite observations: utility analysis for mass change monitoring of the ice sheets\"\nDataset: satellite-ice-sheet-mass-balance [CDS]\nAspect: trend-assessment_q02 | Category: Satellite_ECVs\nSection: Gravimetric mass balance data from satellite observations: utility analysis for mass change monitoring of the ice sheets > ℹ️ If you want to know more > References\n---\n- [[1](https://doi.org/10.1007/s10712-016-9398-7)] Forsberg, R., Sørensen, L.S. and Simonsen, S.B. (2017). Greenland and Antarctica Ice Sheet Mass Changes and Effects on Global Sea Level. Surv. Geophys., 38, 89–104. https://doi.org/10.1007/s10712-016-9398-7\n\n- [[2](https://doi.org/10.3390/geosciences9100415)] Groh, A., Horwath, M., Horvath, A., Meister, R., Sørensen, L.S., Barletta, V.R., Forsberg, R., Wouters, B., Ditmar, P., Ran, J., Klees, R., Su, X., Shang, K., Guo, J., Shum, C.K., Schrama, E., and Shepherd, A. (2019). Evaluating GRACE Mass Change Time Series for the Antarctic and Greenland Ice Sheet, Geosciences, 9(10). https://doi.org/10.3390/geosciences9100415\n\n- [[3](https://doi.org/10.5194/tc-10-1933-2016)] van den Broeke, M., Enderlin, E. M., Howat, I. M., Kuipers Munneke, P., Noël, B. P. Y., van de Berg, W. J., van Meijgaard, E., and Wouters, B. (2016). On the recent contribution of the Greenland ice sheet to sea level change, The Cryosphere, 10, 1933–1946, https://doi.org/10.5194/tc-10-1933-2016\n\n- [[4](https://doi.org/10.1029/2020GL087291)] Velicogna, I., Mohajerani, Y., A, G., Landerer, F., Mouginot, J., Noel, B., Rignot, E., Sutterley, T., van den Broeke, M., van Wessem, M., and Wiese, D. (2020). Continuity of Ice Sheet Mass Loss in Greenland and Antarctica from the GRACE and GRACE Follow-On Missions. Geophys. Res. Lett. 47. https://doi.org/10.1029/2020GL087291\n\n- [[5](https://library.wmo.int/records/item/58111-the-2022-gcos-ecvs-requirements-gcos-245)] GCOS (Global Climate Observing System) (2022). The 2022 GCOS ECVs Requirements (GCOS-245). World Meteorological Organization: Geneva, Switzerland. doi: https://library.wmo.int/idurl/4/58111"} {"chunk_id": "satellite_satellite-ice-sheet-mass-balance_trend-assessment_q02__df5a84a66df3", "report_id": "satellite_satellite-ice-sheet-mass-balance_trend-assessment_q02", "dataset_id": "satellite-ice-sheet-mass-balance", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q02", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Gravimetric mass balance data from satellite observations: utility analysis for mass change monitoring of the ice sheets > ℹ️ If you want to know more > References", "title": "Gravimetric mass balance data from satellite observations: utility analysis for mass change monitoring of the ice sheets", "chunk_index": 15, "token_count": 1113, "text_raw": "(GCOS-245). World Meteorological Organization: Geneva, Switzerland. doi: https://library.wmo.int/idurl/4/58111\n\n- [[6](https://essd.copernicus.org/articles/15/1597/2023/)] Otosaka, I. N., Shepherd, A., Ivins, E. R., Schlegel, N.-J., Amory, C., van den Broeke, M. R., Horwath, M., Joughin, I., King, M. D., Krinner, G., Nowicki, S., Payne, A. J., Rignot, E., Scambos, T., Simon, K. M., Smith, B. E., Sørensen, L. S., Velicogna, I., Whitehouse, P. L., A, G., Agosta, C., Ahlstrøm, A. P., Blazquez, A., Colgan, W., Engdahl, M. E., Fettweis, X., Forsberg, R., Gallée, H., Gardner, A., Gilbert, L., Gourmelen, N., Groh, A., Gunter, B. C., Harig, C., Helm, V., Khan, S. A., Kittel, C., Konrad, H., Langen, P. L., Lecavalier, B. S., Liang, C.-C., Loomis, B. D., McMillan, M., Melini, D., Mernild, S. H., Mottram, R., Mouginot, J., Nilsson, J., Noël, B., Pattle, M. E., Peltier, W. R., Pie, N., Roca, M., Sasgen, I., Save, H. V., Seo, K.-W., Scheuchl, B., Schrama, E. J. O., Schröder, L., Simonsen, S. B., Slater, T., Spada, G., Sutterley, T. C., Vishwakarma, B. D., van Wessem, J. M., Wiese, D., van der Wal, W., and Wouters, B. (2023). Mass balance of the Greenland and Antarctic ice sheets from 1992 to 2020, Earth Syst. Sci. Data, 15, 1597–1616, https://doi.org/10.5194/essd-15-1597-2023\n\n- [[7](https://doi.org/10.1016/j.rse.2015.06.016)] Colgan, W., Abdalati, W., Citterio, M., Csatho, B., Fettweis, X., Luthcke, S., Moholdt, G., Simonsen, S.B., and Stober, M. (2015). Hybrid glacier Inventory, Gravimetry and Altimetry (HIGA) mass balance product for Greenland and the Canadian Arctic. Remote Sensing of Environment, 168, 24–39. https://doi.org/10.1016/j.rse.2015.06.016\n\n- [[8](https://www.ipcc.ch/report/ar6/wg1/downloads/report/IPCC_AR6_WGI_Chapter09.pdf)] Fox-Kemper, B., H.T. Hewitt, C. Xiao, G. Aðalgeirsdóttir, S.S. Drijfhout, T.L. Edwards, N.R. Golledge, M. Hemer, R.E. Kopp, G. Krinner, A. Mix, D. Notz, S. Nowicki, I.S. Nurhati, L. Ruiz, J.-B. Sallée, A.B.A. Slangen, and Y. Yu (2021). Ocean, Cryosphere and Sea Level Change. In Climate Change 2021: The Physical Science Basis. Contribution of Working Group I to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change [Masson-Delmotte, V., P. Zhai, A. Pirani, S.L. Connors, C. Péan, S. Berger, N. Caud, Y. Chen, L. Goldfarb, M.I. Gomis, M. Huang, K. Leitzell, E. Lonnoy, J.B.R. Matthews, T.K. Maycock, T. Waterfield, O. Yelekçi, R. Yu, and B. Zhou (eds.)]. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, pp. 1211–1362, https://doi.org/110.1017/9781009157896.011", "text_with_prefix": "EQC Quality Assessment: \"Gravimetric mass balance data from satellite observations: utility analysis for mass change monitoring of the ice sheets\"\nDataset: satellite-ice-sheet-mass-balance [CDS]\nAspect: trend-assessment_q02 | Category: Satellite_ECVs\nSection: Gravimetric mass balance data from satellite observations: utility analysis for mass change monitoring of the ice sheets > ℹ️ If you want to know more > References\n---\n(GCOS-245). World Meteorological Organization: Geneva, Switzerland. doi: https://library.wmo.int/idurl/4/58111\n\n- [[6](https://essd.copernicus.org/articles/15/1597/2023/)] Otosaka, I. N., Shepherd, A., Ivins, E. R., Schlegel, N.-J., Amory, C., van den Broeke, M. R., Horwath, M., Joughin, I., King, M. D., Krinner, G., Nowicki, S., Payne, A. J., Rignot, E., Scambos, T., Simon, K. M., Smith, B. E., Sørensen, L. S., Velicogna, I., Whitehouse, P. L., A, G., Agosta, C., Ahlstrøm, A. P., Blazquez, A., Colgan, W., Engdahl, M. E., Fettweis, X., Forsberg, R., Gallée, H., Gardner, A., Gilbert, L., Gourmelen, N., Groh, A., Gunter, B. C., Harig, C., Helm, V., Khan, S. A., Kittel, C., Konrad, H., Langen, P. L., Lecavalier, B. S., Liang, C.-C., Loomis, B. D., McMillan, M., Melini, D., Mernild, S. H., Mottram, R., Mouginot, J., Nilsson, J., Noël, B., Pattle, M. E., Peltier, W. R., Pie, N., Roca, M., Sasgen, I., Save, H. V., Seo, K.-W., Scheuchl, B., Schrama, E. J. O., Schröder, L., Simonsen, S. B., Slater, T., Spada, G., Sutterley, T. C., Vishwakarma, B. D., van Wessem, J. M., Wiese, D., van der Wal, W., and Wouters, B. (2023). Mass balance of the Greenland and Antarctic ice sheets from 1992 to 2020, Earth Syst. Sci. Data, 15, 1597–1616, https://doi.org/10.5194/essd-15-1597-2023\n\n- [[7](https://doi.org/10.1016/j.rse.2015.06.016)] Colgan, W., Abdalati, W., Citterio, M., Csatho, B., Fettweis, X., Luthcke, S., Moholdt, G., Simonsen, S.B., and Stober, M. (2015). Hybrid glacier Inventory, Gravimetry and Altimetry (HIGA) mass balance product for Greenland and the Canadian Arctic. Remote Sensing of Environment, 168, 24–39. https://doi.org/10.1016/j.rse.2015.06.016\n\n- [[8](https://www.ipcc.ch/report/ar6/wg1/downloads/report/IPCC_AR6_WGI_Chapter09.pdf)] Fox-Kemper, B., H.T. Hewitt, C. Xiao, G. Aðalgeirsdóttir, S.S. Drijfhout, T.L. Edwards, N.R. Golledge, M. Hemer, R.E. Kopp, G. Krinner, A. Mix, D. Notz, S. Nowicki, I.S. Nurhati, L. Ruiz, J.-B. Sallée, A.B.A. Slangen, and Y. Yu (2021). Ocean, Cryosphere and Sea Level Change. In Climate Change 2021: The Physical Science Basis. Contribution of Working Group I to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change [Masson-Delmotte, V., P. Zhai, A. Pirani, S.L. Connors, C. Péan, S. Berger, N. Caud, Y. Chen, L. Goldfarb, M.I. Gomis, M. Huang, K. Leitzell, E. Lonnoy, J.B.R. Matthews, T.K. Maycock, T. Waterfield, O. Yelekçi, R. Yu, and B. Zhou (eds.)]. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, pp. 1211–1362, https://doi.org/110.1017/9781009157896.011"} {"chunk_id": "satellite_satellite-ice-sheet-mass-balance_trend-assessment_q02__1a9774cd2f2b", "report_id": "satellite_satellite-ice-sheet-mass-balance_trend-assessment_q02", "dataset_id": "satellite-ice-sheet-mass-balance", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q02", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Gravimetric mass balance data from satellite observations: utility analysis for mass change monitoring of the ice sheets > ℹ️ If you want to know more > References", "title": "Gravimetric mass balance data from satellite observations: utility analysis for mass change monitoring of the ice sheets", "chunk_index": 16, "token_count": 619, "text_raw": "B. Zhou (eds.)]. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, pp. 1211–1362, https://doi.org/110.1017/9781009157896.011\n\n- [[9](https://iopscience.iop.org/article/10.1088/1748-9326/aac2f0)] Bamber, J. L., Westaway, R. M., Marzeion, B., and Wouters, B. (2018). The land ice contribution to sea level during the satellite era. Environ. Res. Lett. 13, https://doi.org/10.1088/1748-9326/aac2f0\n\n- [[10](https://tc.copernicus.org/articles/7/1411/2013/)] Barletta, V. R., Sørensen, L. S., and Forsberg, R. (2013). Scatter of mass changes estimates at basin scale for Greenland and Antarctica, The Cryosphere, 7, 1411–1432, https://doi.org/10.5194/tc-7-1411-2013\n\n- [[11](https://doi.org/10.1038/s41467-024-45726-w)] Ludwigsen, C.B., Andersen, O.B., Marzeion, B., Malles, J.H., Müller Schmied, H., Döll, P., Watson, C., and King, M.A. (2024). Global and regional ocean mass budget closure since 2003. Nat. Commun. 15, 1416 (2024). https://doi.org/10.1038/s41467-024-45726-w\n\n- [[12](https://doi.org/10.1038/s41586-019-1855-2)] The IMBIE Team (2019). Mass balance of the Greenland Ice Sheet from 1992 to 2018. Nature, 579, 233–239. https://doi.org/10.1038/s41586-019-1855-2\n\n- [[13](https://doi.org/10.1038/s41586-018-0179-y)] The IMBIE Team (2018). Mass balance of the Antarctic Ice Sheet from 1992 to 2017. Nature 558, 219–222. https://doi.org/10.1038/s41586-018-0179-y", "text_with_prefix": "EQC Quality Assessment: \"Gravimetric mass balance data from satellite observations: utility analysis for mass change monitoring of the ice sheets\"\nDataset: satellite-ice-sheet-mass-balance [CDS]\nAspect: trend-assessment_q02 | Category: Satellite_ECVs\nSection: Gravimetric mass balance data from satellite observations: utility analysis for mass change monitoring of the ice sheets > ℹ️ If you want to know more > References\n---\nB. Zhou (eds.)]. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, pp. 1211–1362, https://doi.org/110.1017/9781009157896.011\n\n- [[9](https://iopscience.iop.org/article/10.1088/1748-9326/aac2f0)] Bamber, J. L., Westaway, R. M., Marzeion, B., and Wouters, B. (2018). The land ice contribution to sea level during the satellite era. Environ. Res. Lett. 13, https://doi.org/10.1088/1748-9326/aac2f0\n\n- [[10](https://tc.copernicus.org/articles/7/1411/2013/)] Barletta, V. R., Sørensen, L. S., and Forsberg, R. (2013). Scatter of mass changes estimates at basin scale for Greenland and Antarctica, The Cryosphere, 7, 1411–1432, https://doi.org/10.5194/tc-7-1411-2013\n\n- [[11](https://doi.org/10.1038/s41467-024-45726-w)] Ludwigsen, C.B., Andersen, O.B., Marzeion, B., Malles, J.H., Müller Schmied, H., Döll, P., Watson, C., and King, M.A. (2024). Global and regional ocean mass budget closure since 2003. Nat. Commun. 15, 1416 (2024). https://doi.org/10.1038/s41467-024-45726-w\n\n- [[12](https://doi.org/10.1038/s41586-019-1855-2)] The IMBIE Team (2019). Mass balance of the Greenland Ice Sheet from 1992 to 2018. Nature, 579, 233–239. https://doi.org/10.1038/s41586-019-1855-2\n\n- [[13](https://doi.org/10.1038/s41586-018-0179-y)] The IMBIE Team (2018). Mass balance of the Antarctic Ice Sheet from 1992 to 2017. Nature 558, 219–222. https://doi.org/10.1038/s41586-018-0179-y"} {"chunk_id": "satellite_satellite-ice-sheet-mass-balance_uncertainty_q01__c225d5effd8a", "report_id": "satellite_satellite-ice-sheet-mass-balance_uncertainty_q01", "dataset_id": "satellite-ice-sheet-mass-balance", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty_q01", "aspect_base": "uncertainty", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Gravimetric mass changes from satellite data: assessment of resolution and uncertainty of Greenland ice sheet mass changes for glaciological applications", "title": "Gravimetric mass changes from satellite data: assessment of resolution and uncertainty of Greenland ice sheet mass changes for glaciological applications", "chunk_index": 0, "token_count": 136, "text_raw": "Production date: 31-05-2025\n\nDataset version: 4.0\n\nProduced by: Yoni Verhaegen and Philippe Huybrechts (Vrije Universiteit Brussel)", "text_with_prefix": "EQC Quality Assessment: \"Gravimetric mass changes from satellite data: assessment of resolution and uncertainty of Greenland ice sheet mass changes for glaciological applications\"\nDataset: satellite-ice-sheet-mass-balance [CDS]\nAspect: uncertainty_q01 | Category: Satellite_ECVs\nSection: Gravimetric mass changes from satellite data: assessment of resolution and uncertainty of Greenland ice sheet mass changes for glaciological applications\n---\nProduction date: 31-05-2025\n\nDataset version: 4.0\n\nProduced by: Yoni Verhaegen and Philippe Huybrechts (Vrije Universiteit Brussel)"} {"chunk_id": "satellite_satellite-ice-sheet-mass-balance_uncertainty_q01__2d747ed6c94c", "report_id": "satellite_satellite-ice-sheet-mass-balance_uncertainty_q01", "dataset_id": "satellite-ice-sheet-mass-balance", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty_q01", "aspect_base": "uncertainty", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Gravimetric mass changes from satellite data: assessment of resolution and uncertainty of Greenland ice sheet mass changes for glaciological applications > Quality assessment question", "title": "Gravimetric mass changes from satellite data: assessment of resolution and uncertainty of Greenland ice sheet mass changes for glaciological applications", "chunk_index": 1, "token_count": 483, "text_raw": "* **\"Is the C3S Greenland ice sheet mass balance dataset sufficiently adequate in terms of its spatio-temporal resolution and uncertainty to be used in glaciological modeling efforts?\"**\n\nIce sheets are significant contributors to current (and future) global sea-level rise, indicators of climatic changes, and key players regarding feedback mechanisms within the atmosphere, ocean, and cryosphere. Assessing ice sheet mass changes is essential for understanding these mechanisms. At present-day, the gravimetric method from satellite-based remote sensing offers one out of the three feasible ways to regularly and accurately monitor mass changes of the entire ice sheets at a regular basis (the others being the altimetric and input-output method). The [\"Gravimetric mass balance data for the Antarctic and Greenland ice sheets from 2003 to 2022 derived from satellite observations\"](https://cds.climate.copernicus.eu/datasets/satellite-ice-sheet-mass-balance?tab=overview) dataset on the Climate Data Store (CDS) therefore provides valuable insights into the ice sheet's mass changes. The dataset uses satellite gravimetry from the GRACE and GRACE-FO missions to detect changes in the Earth's gravitational field. These are then translated into cumulative mass anomalies of ice above buoyancy for the grounded ice of both the Greenland (GrIS) and Antarctic (AIS) ice sheets [[1](https://doi.org/10.1007/s10712-016-9398-7), [2](https://doi.org/10.3390/geosciences9100415)]. This notebook evaluates the dataset’s maturity for monitoring Greenland ice mass changes in the 21st century. More specifically, we will check whether the data are sufficiently adequate in terms of its spatio-temporal resolution and uncertainty to be applicable in glaciological modeling applications. For that, we use the dataset version 4.0.", "text_with_prefix": "EQC Quality Assessment: \"Gravimetric mass changes from satellite data: assessment of resolution and uncertainty of Greenland ice sheet mass changes for glaciological applications\"\nDataset: satellite-ice-sheet-mass-balance [CDS]\nAspect: uncertainty_q01 | Category: Satellite_ECVs\nSection: Gravimetric mass changes from satellite data: assessment of resolution and uncertainty of Greenland ice sheet mass changes for glaciological applications > Quality assessment question\n---\n* **\"Is the C3S Greenland ice sheet mass balance dataset sufficiently adequate in terms of its spatio-temporal resolution and uncertainty to be used in glaciological modeling efforts?\"**\n\nIce sheets are significant contributors to current (and future) global sea-level rise, indicators of climatic changes, and key players regarding feedback mechanisms within the atmosphere, ocean, and cryosphere. Assessing ice sheet mass changes is essential for understanding these mechanisms. At present-day, the gravimetric method from satellite-based remote sensing offers one out of the three feasible ways to regularly and accurately monitor mass changes of the entire ice sheets at a regular basis (the others being the altimetric and input-output method). The [\"Gravimetric mass balance data for the Antarctic and Greenland ice sheets from 2003 to 2022 derived from satellite observations\"](https://cds.climate.copernicus.eu/datasets/satellite-ice-sheet-mass-balance?tab=overview) dataset on the Climate Data Store (CDS) therefore provides valuable insights into the ice sheet's mass changes. The dataset uses satellite gravimetry from the GRACE and GRACE-FO missions to detect changes in the Earth's gravitational field. These are then translated into cumulative mass anomalies of ice above buoyancy for the grounded ice of both the Greenland (GrIS) and Antarctic (AIS) ice sheets [[1](https://doi.org/10.1007/s10712-016-9398-7), [2](https://doi.org/10.3390/geosciences9100415)]. This notebook evaluates the dataset’s maturity for monitoring Greenland ice mass changes in the 21st century. More specifically, we will check whether the data are sufficiently adequate in terms of its spatio-temporal resolution and uncertainty to be applicable in glaciological modeling applications. For that, we use the dataset version 4.0."} {"chunk_id": "satellite_satellite-ice-sheet-mass-balance_uncertainty_q01__57ddc5870e9d", "report_id": "satellite_satellite-ice-sheet-mass-balance_uncertainty_q01", "dataset_id": "satellite-ice-sheet-mass-balance", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty_q01", "aspect_base": "uncertainty", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Gravimetric mass changes from satellite data: assessment of resolution and uncertainty of Greenland ice sheet mass changes for glaciological applications > Quality assessment statements", "title": "Gravimetric mass changes from satellite data: assessment of resolution and uncertainty of Greenland ice sheet mass changes for glaciological applications", "chunk_index": 2, "token_count": 515, "text_raw": "These are the key outcomes of this assessment\n\n- Mass change detection by GRACE(-FO) is a useful tool to quantify the total ice mass change of the GrIS (and AIS), and is one of the three main data acquisition methods that can extract ice mass change data from large spatial scales, such as the complete ice sheets, at a regular basis. GRACE(-FO) directly captures all relevant processes for ice mass change (from both mass balance and ice dynamics), but it also has its limitations of which the user should take note before using the product. A large problem with the C3S dataset is that no gridded mass change and error products are included for the ice sheets, as well as the presence of (occasionally very) high error/uncertainty values, primarily due to the large data acquisition footprint and corrections needed for mass redistribution in the atmosphere, ocean, soil and solid earth. Data gaps are also present (e.g. during the transition period between GRACE and GRACE-FO in 2017-2018 CE). These are not filled up or flagged, users need to identify them themselves. \n- Concerning the specific use case and user question, the C3S GRACE(-FO) data are found to be inadequate to be used in most \"traditional\" glaciological modeling efforts because it is (1) only a spatially aggregated time series and has no gridded data, (2) does not include a surface mass balance component separately (which is, for example, needed in the prognostic continuity equation for ice thickness changes), and (3) has (very) high uncertainty in certain months. The GRACE(-FO) data can, on the other hand, be used for other purposes. The data are particularly well-suited for the direct calculation and monitoring of ice sheet mass changes and sea level contribution, and/or the validation of mass change estimates from other independent methods (e.g. the input-output method or the validation of the temporal variability of a surface mass balance model from a regional climate model). \n```", "text_with_prefix": "EQC Quality Assessment: \"Gravimetric mass changes from satellite data: assessment of resolution and uncertainty of Greenland ice sheet mass changes for glaciological applications\"\nDataset: satellite-ice-sheet-mass-balance [CDS]\nAspect: uncertainty_q01 | Category: Satellite_ECVs\nSection: Gravimetric mass changes from satellite data: assessment of resolution and uncertainty of Greenland ice sheet mass changes for glaciological applications > Quality assessment statements\n---\nThese are the key outcomes of this assessment\n\n- Mass change detection by GRACE(-FO) is a useful tool to quantify the total ice mass change of the GrIS (and AIS), and is one of the three main data acquisition methods that can extract ice mass change data from large spatial scales, such as the complete ice sheets, at a regular basis. GRACE(-FO) directly captures all relevant processes for ice mass change (from both mass balance and ice dynamics), but it also has its limitations of which the user should take note before using the product. A large problem with the C3S dataset is that no gridded mass change and error products are included for the ice sheets, as well as the presence of (occasionally very) high error/uncertainty values, primarily due to the large data acquisition footprint and corrections needed for mass redistribution in the atmosphere, ocean, soil and solid earth. Data gaps are also present (e.g. during the transition period between GRACE and GRACE-FO in 2017-2018 CE). These are not filled up or flagged, users need to identify them themselves. \n- Concerning the specific use case and user question, the C3S GRACE(-FO) data are found to be inadequate to be used in most \"traditional\" glaciological modeling efforts because it is (1) only a spatially aggregated time series and has no gridded data, (2) does not include a surface mass balance component separately (which is, for example, needed in the prognostic continuity equation for ice thickness changes), and (3) has (very) high uncertainty in certain months. The GRACE(-FO) data can, on the other hand, be used for other purposes. The data are particularly well-suited for the direct calculation and monitoring of ice sheet mass changes and sea level contribution, and/or the validation of mass change estimates from other independent methods (e.g. the input-output method or the validation of the temporal variability of a surface mass balance model from a regional climate model). \n```"} {"chunk_id": "satellite_satellite-ice-sheet-mass-balance_uncertainty_q01__968ed148b82c", "report_id": "satellite_satellite-ice-sheet-mass-balance_uncertainty_q01", "dataset_id": "satellite-ice-sheet-mass-balance", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty_q01", "aspect_base": "uncertainty", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Gravimetric mass changes from satellite data: assessment of resolution and uncertainty of Greenland ice sheet mass changes for glaciological applications > Methodology > Dataset description", "title": "Gravimetric mass changes from satellite data: assessment of resolution and uncertainty of Greenland ice sheet mass changes for glaciological applications", "chunk_index": 3, "token_count": 780, "text_raw": "The total mass balance of an ice sheet is the difference between mass gained (mainly from snow accumulation) and mass lost (by ablation or solid ice discharge across the grounding line), which is the same as the net mass change of the ice sheet. Remote sensing techniques, such as the use of satellites, are an important feature to derive and study the mass changes of the ice sheets. The ice sheet mass change dataset provides monthly gravimetric mass balance (GMB) values and their uncertainty for the Greenland (GrIS) and also the Antarctic ice sheet (AIS). The data represent a time series of the cumulative mass changes (mass anomalies) of the ice above buoyancy of the ice sheets and their basins (including ice caps and glaciers) that are derived using satellite gravimetry data from the GRACE(-FO) missions. Data are available as time series for the whole ice sheet, as well as at the basin level, but no gridded data are provided. Data are available since 2002 with units in Gt (Gigatonnes). For a more detailed description of the data acquisition and processing methods, we refer to the [documentation on the CDS](https://cds.climate.copernicus.eu/datasets/satellite-ice-sheet-mass-balance?tab=overview) and the [ECMWF Confluence Wiki](https://confluence.ecmwf.int/pages/viewpage.action?pageId=355348671) (Copernicus Knowledge Base).\n\nThe mass changes and their errors derived from GRACE(-FO) are expressed in Gt (Gigatonnes). It can also be translated into a volume, for example one Gt of water (density 1000 kg/m$^3$) is exactly one km$^3$, while one Gt of ice (density 917 kg/m$^3$) in volume becomes 1.091 km$^3$ of ice. GRACE(-FO) data are in fact considered to be the sum of mass changes driven by changing rates of solid ice discharge (i.e. the flux across the grounding line $D$) and mass changes driven by changing rates of ablation and accumulation (mainly the surface mass balance or $SMB$). What GRACE(-FO) actually detects are patterns of mass redistribution, indicating that a material should be redistributed (e.g. by ice dynamics or liquid discharge) in order for GRACE(-FO) to be able to detect gravity anomalies and the corresponding mass changes over certain locations. In that sense, an amount of melted ice being replaced by its mass-equivalent amount of meltwater at the same location would result in zero mass change.\n\nThe main advantage of GRACE(-FO) is that all processes directly contributing to ice sheet mass fluctuations are observed directly. Unlike the input-output method, SMB is therefore not involved explicitely in processing and it is not separated from ice flow dynamics. Moreover, gaps in the time series do not affect linear trend estimation as mass changes occurring during these gaps are still included in the subsequent solution. The main drawbacks of this technique are the large footprint of data acquisition and the fact that corrections must be made for mass redistribution in the atmosphere, ocean, soil and solid earth (e.g. glacial isostatic adjustment) in order to isolate mass changes relevant for the ice sheets.", "text_with_prefix": "EQC Quality Assessment: \"Gravimetric mass changes from satellite data: assessment of resolution and uncertainty of Greenland ice sheet mass changes for glaciological applications\"\nDataset: satellite-ice-sheet-mass-balance [CDS]\nAspect: uncertainty_q01 | Category: Satellite_ECVs\nSection: Gravimetric mass changes from satellite data: assessment of resolution and uncertainty of Greenland ice sheet mass changes for glaciological applications > Methodology > Dataset description\n---\nThe total mass balance of an ice sheet is the difference between mass gained (mainly from snow accumulation) and mass lost (by ablation or solid ice discharge across the grounding line), which is the same as the net mass change of the ice sheet. Remote sensing techniques, such as the use of satellites, are an important feature to derive and study the mass changes of the ice sheets. The ice sheet mass change dataset provides monthly gravimetric mass balance (GMB) values and their uncertainty for the Greenland (GrIS) and also the Antarctic ice sheet (AIS). The data represent a time series of the cumulative mass changes (mass anomalies) of the ice above buoyancy of the ice sheets and their basins (including ice caps and glaciers) that are derived using satellite gravimetry data from the GRACE(-FO) missions. Data are available as time series for the whole ice sheet, as well as at the basin level, but no gridded data are provided. Data are available since 2002 with units in Gt (Gigatonnes). For a more detailed description of the data acquisition and processing methods, we refer to the [documentation on the CDS](https://cds.climate.copernicus.eu/datasets/satellite-ice-sheet-mass-balance?tab=overview) and the [ECMWF Confluence Wiki](https://confluence.ecmwf.int/pages/viewpage.action?pageId=355348671) (Copernicus Knowledge Base).\n\nThe mass changes and their errors derived from GRACE(-FO) are expressed in Gt (Gigatonnes). It can also be translated into a volume, for example one Gt of water (density 1000 kg/m$^3$) is exactly one km$^3$, while one Gt of ice (density 917 kg/m$^3$) in volume becomes 1.091 km$^3$ of ice. GRACE(-FO) data are in fact considered to be the sum of mass changes driven by changing rates of solid ice discharge (i.e. the flux across the grounding line $D$) and mass changes driven by changing rates of ablation and accumulation (mainly the surface mass balance or $SMB$). What GRACE(-FO) actually detects are patterns of mass redistribution, indicating that a material should be redistributed (e.g. by ice dynamics or liquid discharge) in order for GRACE(-FO) to be able to detect gravity anomalies and the corresponding mass changes over certain locations. In that sense, an amount of melted ice being replaced by its mass-equivalent amount of meltwater at the same location would result in zero mass change.\n\nThe main advantage of GRACE(-FO) is that all processes directly contributing to ice sheet mass fluctuations are observed directly. Unlike the input-output method, SMB is therefore not involved explicitely in processing and it is not separated from ice flow dynamics. Moreover, gaps in the time series do not affect linear trend estimation as mass changes occurring during these gaps are still included in the subsequent solution. The main drawbacks of this technique are the large footprint of data acquisition and the fact that corrections must be made for mass redistribution in the atmosphere, ocean, soil and solid earth (e.g. glacial isostatic adjustment) in order to isolate mass changes relevant for the ice sheets."} {"chunk_id": "satellite_satellite-ice-sheet-mass-balance_uncertainty_q01__f921d3e08aaf", "report_id": "satellite_satellite-ice-sheet-mass-balance_uncertainty_q01", "dataset_id": "satellite-ice-sheet-mass-balance", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty_q01", "aspect_base": "uncertainty", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Gravimetric mass changes from satellite data: assessment of resolution and uncertainty of Greenland ice sheet mass changes for glaciological applications > Methodology > Structure and (sub)sections", "title": "Gravimetric mass changes from satellite data: assessment of resolution and uncertainty of Greenland ice sheet mass changes for glaciological applications", "chunk_index": 4, "token_count": 178, "text_raw": "**[](section-1)**\n* [](section-1-1)\n* [](section-1-2)\n* [](section-1-3)\n* [](section-1-4)\n\n**[](section-2)**\n* [](section-2-1)\n* [](section-2-2)\n* [](section-2-3)\n\n**[](section-3)**", "text_with_prefix": "EQC Quality Assessment: \"Gravimetric mass changes from satellite data: assessment of resolution and uncertainty of Greenland ice sheet mass changes for glaciological applications\"\nDataset: satellite-ice-sheet-mass-balance [CDS]\nAspect: uncertainty_q01 | Category: Satellite_ECVs\nSection: Gravimetric mass changes from satellite data: assessment of resolution and uncertainty of Greenland ice sheet mass changes for glaciological applications > Methodology > Structure and (sub)sections\n---\n**[](section-1)**\n* [](section-1-1)\n* [](section-1-2)\n* [](section-1-3)\n* [](section-1-4)\n\n**[](section-2)**\n* [](section-2-1)\n* [](section-2-2)\n* [](section-2-3)\n\n**[](section-3)**"} {"chunk_id": "satellite_satellite-ice-sheet-mass-balance_uncertainty_q01__3ace85926497", "report_id": "satellite_satellite-ice-sheet-mass-balance_uncertainty_q01", "dataset_id": "satellite-ice-sheet-mass-balance", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty_q01", "aspect_base": "uncertainty", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Gravimetric mass changes from satellite data: assessment of resolution and uncertainty of Greenland ice sheet mass changes for glaciological applications > Analysis and results > 1. Data preparation and processing > 1.2 Define request and download", "title": "Gravimetric mass changes from satellite data: assessment of resolution and uncertainty of Greenland ice sheet mass changes for glaciological applications", "chunk_index": 5, "token_count": 214, "text_raw": "Then we define the parameters, i.e. for which ice sheet (or which basins of these ice sheets) we want the mass change data to be extracted:\n\nThen we define requests for download from the CDS and download and transform the glacier mass change data.\n\n```text\n100%|██████████| 1/1 [00:04<00:00, 4.46s/it]\n```\n\n```text\nDownloading done.\n```\n\n(section-1-3)=", "text_with_prefix": "EQC Quality Assessment: \"Gravimetric mass changes from satellite data: assessment of resolution and uncertainty of Greenland ice sheet mass changes for glaciological applications\"\nDataset: satellite-ice-sheet-mass-balance [CDS]\nAspect: uncertainty_q01 | Category: Satellite_ECVs\nSection: Gravimetric mass changes from satellite data: assessment of resolution and uncertainty of Greenland ice sheet mass changes for glaciological applications > Analysis and results > 1. Data preparation and processing > 1.2 Define request and download\n---\nThen we define the parameters, i.e. for which ice sheet (or which basins of these ice sheets) we want the mass change data to be extracted:\n\nThen we define requests for download from the CDS and download and transform the glacier mass change data.\n\n```text\n100%|██████████| 1/1 [00:04<00:00, 4.46s/it]\n```\n\n```text\nDownloading done.\n```\n\n(section-1-3)="} {"chunk_id": "satellite_satellite-ice-sheet-mass-balance_uncertainty_q01__1bfa0776053a", "report_id": "satellite_satellite-ice-sheet-mass-balance_uncertainty_q01", "dataset_id": "satellite-ice-sheet-mass-balance", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty_q01", "aspect_base": "uncertainty", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Gravimetric mass changes from satellite data: assessment of resolution and uncertainty of Greenland ice sheet mass changes for glaciological applications > Analysis and results > 1. Data preparation and processing > 1.3 Display and inspect data", "title": "Gravimetric mass changes from satellite data: assessment of resolution and uncertainty of Greenland ice sheet mass changes for glaciological applications", "chunk_index": 6, "token_count": 1041, "text_raw": "Let us inspect the data:\n\n```text\n Size: 17kB\nDimensions: (time: 216)\nCoordinates:\n * time (time) datetime64[ns] 2kB 2002-04-16T20:23:54.375000 ... 2...\nData variables: (12/18)\n GrIS_total (time) float32 864B 1.057e+03 1.123e+03 ... -3.762e+03\n GrIS_1 (time) float32 864B 54.15 42.37 52.4 ... -440.0 -471.2 -441.5\n GrIS_2 (time) float32 864B 97.2 125.0 43.44 ... 28.75 80.42 17.64\n GrIS_3 (time) float32 864B 219.8 216.2 132.5 ... -467.5 -533.6\n GrIS_4 (time) float32 864B 140.1 152.3 129.0 ... -548.1 -498.9\n GrIS_5 (time) float32 864B 76.06 68.17 53.5 ... -288.9 -316.6 -274.7\n ... ...\n GrIS_3_er (time) float32 864B 45.75 25.68 20.46 ... 16.83 18.87 19.34\n GrIS_4_er (time) float32 864B 42.49 30.83 25.91 ... 18.51 18.26 22.87\n GrIS_5_er (time) float32 864B 26.84 32.99 18.78 ... 12.84 11.66 11.76\n GrIS_6_er (time) float32 864B 73.16 41.97 28.93 ... 22.61 24.26 25.49\n GrIS_7_er (time) float32 864B 57.64 41.52 19.34 ... 23.11 25.23 26.11\n GrIS_8_er (time) float32 864B 59.26 63.21 45.86 ... 22.68 22.85 21.76\nAttributes:\n Title: GMB for Greenland and Antarctica ice sheets from th...\n institution: DTU Space - Geodesy and Earth Observations\n reference: Baratta et al. (2016), Groh and Horwart (2016)\n file_creation_date: Tue May 16 09:49:32 2023\n region: Greenland and Antarctica\n missions_used: GRACE and and GRACE-FO\n time_coverage_start: Apr-2002\n time_coverage_end: Dec-2022\n Tracking_id: ab2360a4-82d5-42f3-babb-90ce976a8a8e\n netCDF_version: NETCDF4\n product_version: 4.0\n Summary: This data set is prepared for the C3S project, and ...\n```\n\nIt is a dataset that consists of several time series data, containing cumulative values of the total ice sheet mass change (cumulative mass anomalies) of, in this notebook, the entire Greenland Ice Sheet (`GrIS_total`) or its basins (`GrIS_{basin_number}`), as well as their uncertainty (`GrIS_total_er` and `GrIS_{basin_number}_er`) since 2002. For basin definitions and delineation, see Zwally et al. (2012) [[3](https://earth.gsfc.nasa.gov/cryo/data/polar-altimetry/antarctic-and-greenland-drainage-systems)]. Mass changes and their uncertainty are expressed in units of Gt and the time period between two measurements is variable but mostly at monthly-spaced intervals. Note that no gridded data are given in this dataset, and hence no spatial resolution can be derived.\n\n(section-1-4)=", "text_with_prefix": "EQC Quality Assessment: \"Gravimetric mass changes from satellite data: assessment of resolution and uncertainty of Greenland ice sheet mass changes for glaciological applications\"\nDataset: satellite-ice-sheet-mass-balance [CDS]\nAspect: uncertainty_q01 | Category: Satellite_ECVs\nSection: Gravimetric mass changes from satellite data: assessment of resolution and uncertainty of Greenland ice sheet mass changes for glaciological applications > Analysis and results > 1. Data preparation and processing > 1.3 Display and inspect data\n---\nLet us inspect the data:\n\n```text\n Size: 17kB\nDimensions: (time: 216)\nCoordinates:\n * time (time) datetime64[ns] 2kB 2002-04-16T20:23:54.375000 ... 2...\nData variables: (12/18)\n GrIS_total (time) float32 864B 1.057e+03 1.123e+03 ... -3.762e+03\n GrIS_1 (time) float32 864B 54.15 42.37 52.4 ... -440.0 -471.2 -441.5\n GrIS_2 (time) float32 864B 97.2 125.0 43.44 ... 28.75 80.42 17.64\n GrIS_3 (time) float32 864B 219.8 216.2 132.5 ... -467.5 -533.6\n GrIS_4 (time) float32 864B 140.1 152.3 129.0 ... -548.1 -498.9\n GrIS_5 (time) float32 864B 76.06 68.17 53.5 ... -288.9 -316.6 -274.7\n ... ...\n GrIS_3_er (time) float32 864B 45.75 25.68 20.46 ... 16.83 18.87 19.34\n GrIS_4_er (time) float32 864B 42.49 30.83 25.91 ... 18.51 18.26 22.87\n GrIS_5_er (time) float32 864B 26.84 32.99 18.78 ... 12.84 11.66 11.76\n GrIS_6_er (time) float32 864B 73.16 41.97 28.93 ... 22.61 24.26 25.49\n GrIS_7_er (time) float32 864B 57.64 41.52 19.34 ... 23.11 25.23 26.11\n GrIS_8_er (time) float32 864B 59.26 63.21 45.86 ... 22.68 22.85 21.76\nAttributes:\n Title: GMB for Greenland and Antarctica ice sheets from th...\n institution: DTU Space - Geodesy and Earth Observations\n reference: Baratta et al. (2016), Groh and Horwart (2016)\n file_creation_date: Tue May 16 09:49:32 2023\n region: Greenland and Antarctica\n missions_used: GRACE and and GRACE-FO\n time_coverage_start: Apr-2002\n time_coverage_end: Dec-2022\n Tracking_id: ab2360a4-82d5-42f3-babb-90ce976a8a8e\n netCDF_version: NETCDF4\n product_version: 4.0\n Summary: This data set is prepared for the C3S project, and ...\n```\n\nIt is a dataset that consists of several time series data, containing cumulative values of the total ice sheet mass change (cumulative mass anomalies) of, in this notebook, the entire Greenland Ice Sheet (`GrIS_total`) or its basins (`GrIS_{basin_number}`), as well as their uncertainty (`GrIS_total_er` and `GrIS_{basin_number}_er`) since 2002. For basin definitions and delineation, see Zwally et al. (2012) [[3](https://earth.gsfc.nasa.gov/cryo/data/polar-altimetry/antarctic-and-greenland-drainage-systems)]. Mass changes and their uncertainty are expressed in units of Gt and the time period between two measurements is variable but mostly at monthly-spaced intervals. Note that no gridded data are given in this dataset, and hence no spatial resolution can be derived.\n\n(section-1-4)="} {"chunk_id": "satellite_satellite-ice-sheet-mass-balance_uncertainty_q01__bcd883247471", "report_id": "satellite_satellite-ice-sheet-mass-balance_uncertainty_q01", "dataset_id": "satellite-ice-sheet-mass-balance", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty_q01", "aspect_base": "uncertainty", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Gravimetric mass changes from satellite data: assessment of resolution and uncertainty of Greenland ice sheet mass changes for glaciological applications > Analysis and results > 1. Data preparation and processing > 1.4 Data handling and creating functions", "title": "Gravimetric mass changes from satellite data: assessment of resolution and uncertainty of Greenland ice sheet mass changes for glaciological applications", "chunk_index": 7, "token_count": 234, "text_raw": "Let us perform some data handling before getting started with the analysis:\n\nCheck if the variable name contains a number\nCheck if the variable name contains a number\nCheck if the variable name contains a number\n\nWe also define a plotting function to visualize the time series:\n\nCreate a GridSpec layout with enough rows\nConvert time to pandas datetime for easy handling\nPlot the first variable, spanning the first row\nDefine dataset-specific y-axis bounds\nIdentify large gaps and shade them\nPlot the remaining variables, two per row\n\nWith everything ready, let us now start with the analysis:\n\n(section-2)=", "text_with_prefix": "EQC Quality Assessment: \"Gravimetric mass changes from satellite data: assessment of resolution and uncertainty of Greenland ice sheet mass changes for glaciological applications\"\nDataset: satellite-ice-sheet-mass-balance [CDS]\nAspect: uncertainty_q01 | Category: Satellite_ECVs\nSection: Gravimetric mass changes from satellite data: assessment of resolution and uncertainty of Greenland ice sheet mass changes for glaciological applications > Analysis and results > 1. Data preparation and processing > 1.4 Data handling and creating functions\n---\nLet us perform some data handling before getting started with the analysis:\n\nCheck if the variable name contains a number\nCheck if the variable name contains a number\nCheck if the variable name contains a number\n\nWe also define a plotting function to visualize the time series:\n\nCreate a GridSpec layout with enough rows\nConvert time to pandas datetime for easy handling\nPlot the first variable, spanning the first row\nDefine dataset-specific y-axis bounds\nIdentify large gaps and shade them\nPlot the remaining variables, two per row\n\nWith everything ready, let us now start with the analysis:\n\n(section-2)="} {"chunk_id": "satellite_satellite-ice-sheet-mass-balance_uncertainty_q01__285749d55a31", "report_id": "satellite_satellite-ice-sheet-mass-balance_uncertainty_q01", "dataset_id": "satellite-ice-sheet-mass-balance", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty_q01", "aspect_base": "uncertainty", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Gravimetric mass changes from satellite data: assessment of resolution and uncertainty of Greenland ice sheet mass changes for glaciological applications > Analysis and results > 2. Quantifying Greenland Ice Sheet mass changes and their errors in space and time > 2.1 Time series of cumulative ice sheet mass changes", "title": "Gravimetric mass changes from satellite data: assessment of resolution and uncertainty of Greenland ice sheet mass changes for glaciological applications", "chunk_index": 8, "token_count": 443, "text_raw": "We begin by plotting the GrIS cumulative mass change $M_{GRACE}$ between the begin and end period with the defined plotting function, where the shading in the upper part of the plot indicates when the time between two consecutive measurements is more than 1 month:\n\n*Figure 1. Greenland ice sheet mass changes from GRACE(-FO). The top plot shows the mass changes over the entire ice sheet, where shaded intervals depict data gaps longer than 1 month. The smaller graphs below show the mass changes of the individual basins.*\n\nIn the figure above, the first graph on top represents the cumulative mass change of the entire GrIS, while other graphs show the mass changes of its basins. Overall, these graphs depict a significant and widespread loss of ice mass across the practically the entire ice sheet, with different basins showing varying rates and patterns of ice loss. The findings underscore the critical state of the GrIS, serving as a clear indicator of the impacts of climate change [[2](https://doi.org/10.3390/geosciences9100415), [4](https://essd.copernicus.org/articles/15/1597/2023/)].\n\nWe recall the data are expressed as time series of cumulative values (mass anomalies) of the sum of mass changes driven by changing rates of solid ice discharge (i.e. the ice flux across the grounding line) and mass changes driven by changing rates of ablation and accumulation (mainly the SMB) at the basin-scale or ice sheet-wide scale.\n\n(section-2-2)=", "text_with_prefix": "EQC Quality Assessment: \"Gravimetric mass changes from satellite data: assessment of resolution and uncertainty of Greenland ice sheet mass changes for glaciological applications\"\nDataset: satellite-ice-sheet-mass-balance [CDS]\nAspect: uncertainty_q01 | Category: Satellite_ECVs\nSection: Gravimetric mass changes from satellite data: assessment of resolution and uncertainty of Greenland ice sheet mass changes for glaciological applications > Analysis and results > 2. Quantifying Greenland Ice Sheet mass changes and their errors in space and time > 2.1 Time series of cumulative ice sheet mass changes\n---\nWe begin by plotting the GrIS cumulative mass change $M_{GRACE}$ between the begin and end period with the defined plotting function, where the shading in the upper part of the plot indicates when the time between two consecutive measurements is more than 1 month:\n\n*Figure 1. Greenland ice sheet mass changes from GRACE(-FO). The top plot shows the mass changes over the entire ice sheet, where shaded intervals depict data gaps longer than 1 month. The smaller graphs below show the mass changes of the individual basins.*\n\nIn the figure above, the first graph on top represents the cumulative mass change of the entire GrIS, while other graphs show the mass changes of its basins. Overall, these graphs depict a significant and widespread loss of ice mass across the practically the entire ice sheet, with different basins showing varying rates and patterns of ice loss. The findings underscore the critical state of the GrIS, serving as a clear indicator of the impacts of climate change [[2](https://doi.org/10.3390/geosciences9100415), [4](https://essd.copernicus.org/articles/15/1597/2023/)].\n\nWe recall the data are expressed as time series of cumulative values (mass anomalies) of the sum of mass changes driven by changing rates of solid ice discharge (i.e. the ice flux across the grounding line) and mass changes driven by changing rates of ablation and accumulation (mainly the SMB) at the basin-scale or ice sheet-wide scale.\n\n(section-2-2)="} {"chunk_id": "satellite_satellite-ice-sheet-mass-balance_uncertainty_q01__b3455df045ec", "report_id": "satellite_satellite-ice-sheet-mass-balance_uncertainty_q01", "dataset_id": "satellite-ice-sheet-mass-balance", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty_q01", "aspect_base": "uncertainty", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Gravimetric mass changes from satellite data: assessment of resolution and uncertainty of Greenland ice sheet mass changes for glaciological applications > Analysis and results > 2. Quantifying Greenland Ice Sheet mass changes and their errors in space and time > 2.2 Ice sheet mass changes uncertainty estimates", "title": "Gravimetric mass changes from satellite data: assessment of resolution and uncertainty of Greenland ice sheet mass changes for glaciological applications", "chunk_index": 9, "token_count": 978, "text_raw": "The total error of a monthly ice sheet mass change estimate for Greenland is given by the sum of the precision (random) and the accuracy (systematic) error:\n\n$\n\\varepsilon = a\\sigma + \\delta\n$\nwhere $a$ is the critical z-value related to a certain statistical confidence interval, $\\sigma$ is the standard deviation (random error) and $\\delta$ the accuracy (systematic) error.\n\nFor the GrIS, the product deals with both precision and accuracy errors in its final error estimates [[5](https://tc.copernicus.org/articles/7/1411/2013/)]. The precision error accounts for the statistically distributed random error around one average value. The accuracy error accounts for how much the expected value deviates from the “true” value, dervied from averaging mass change time series obtained with different methods, models and corrections. For the precision error, the 95% confidence interval ($a$ = 1.96) propagated from the data, is provided. Therefore, in the case for the GrIS, the total uncertainty of the data is given by $\\varepsilon$. Hence, in the C3S products the most complete error estimate that was evaluated by Barletta et al. (2013) was given [[5](https://tc.copernicus.org/articles/7/1411/2013/)]. In the following section below, we will thus consider the uncertainty of the dataset as being $\\varepsilon_{M_{GRACE}}$.\n\nLet us plot a time series of the errors:\n\n*Figure 2. Greenland ice sheet mass change errors from GRACE(-FO). The top plot shows the mass change errors over the entire ice sheet, where shaded intervals depict data gaps longer than 1 month. The smaller graphs below show the mass change errors of the individual basins.*\n\nNote that these values are not normalized to a certain time period, as the time gap between two measurements is (slightly) variable. Uncertainties in GRACE ice mass change estimates are occasionally very high and arise from various sources. These include, amongst others, errors related to measurement noise (e.g. due to leakage of the signal outside the region of interest such as the Canadian ice caps, due to the coarse resolution of GRACE(-FO) data acquisition), uncertainties related to the Earth's gravitational field (e.g. due to the shift of the Earth's center of mass, or due to the influence of the atmosphere and oceans on the Earth's gravity field), and uncertainties related to other mass change signals that may be intertwined in the GRACE(-FO) data (e.g. those related to glacial isostatic adjustment).\n\nWe can furthermore calculate the average values (arithmetic means over time):\n\n```text\nThe arithmetic mean mass change error of the GrIS in total is 76.72 Gt with a maximum of 740.25 Gt.\nThe arithmetic mean mass change error of the GrIS basin 1 is 58.71 Gt with a maximum of 671.28 Gt.\nThe arithmetic mean mass change error of the GrIS basin 2 is 42.94 Gt with a maximum of 438.62 Gt.\nThe arithmetic mean mass change error of the GrIS basin 3 is 23.66 Gt with a maximum of 231.57 Gt.\nThe arithmetic mean mass change error of the GrIS basin 4 is 22.24 Gt with a maximum of 249.10 Gt.\nThe arithmetic mean mass change error of the GrIS basin 5 is 13.86 Gt with a maximum of 108.55 Gt.\nThe arithmetic mean mass change error of the GrIS basin 6 is 33.96 Gt with a maximum of 412.35 Gt.\nThe arithmetic mean mass change error of the GrIS basin 7 is 24.06 Gt with a maximum of 234.92 Gt.\nThe arithmetic mean mass change error of the GrIS basin 8 is 33.13 Gt with a maximum of 450.22 Gt.\n```", "text_with_prefix": "EQC Quality Assessment: \"Gravimetric mass changes from satellite data: assessment of resolution and uncertainty of Greenland ice sheet mass changes for glaciological applications\"\nDataset: satellite-ice-sheet-mass-balance [CDS]\nAspect: uncertainty_q01 | Category: Satellite_ECVs\nSection: Gravimetric mass changes from satellite data: assessment of resolution and uncertainty of Greenland ice sheet mass changes for glaciological applications > Analysis and results > 2. Quantifying Greenland Ice Sheet mass changes and their errors in space and time > 2.2 Ice sheet mass changes uncertainty estimates\n---\nThe total error of a monthly ice sheet mass change estimate for Greenland is given by the sum of the precision (random) and the accuracy (systematic) error:\n\n$\n\\varepsilon = a\\sigma + \\delta\n$\nwhere $a$ is the critical z-value related to a certain statistical confidence interval, $\\sigma$ is the standard deviation (random error) and $\\delta$ the accuracy (systematic) error.\n\nFor the GrIS, the product deals with both precision and accuracy errors in its final error estimates [[5](https://tc.copernicus.org/articles/7/1411/2013/)]. The precision error accounts for the statistically distributed random error around one average value. The accuracy error accounts for how much the expected value deviates from the “true” value, dervied from averaging mass change time series obtained with different methods, models and corrections. For the precision error, the 95% confidence interval ($a$ = 1.96) propagated from the data, is provided. Therefore, in the case for the GrIS, the total uncertainty of the data is given by $\\varepsilon$. Hence, in the C3S products the most complete error estimate that was evaluated by Barletta et al. (2013) was given [[5](https://tc.copernicus.org/articles/7/1411/2013/)]. In the following section below, we will thus consider the uncertainty of the dataset as being $\\varepsilon_{M_{GRACE}}$.\n\nLet us plot a time series of the errors:\n\n*Figure 2. Greenland ice sheet mass change errors from GRACE(-FO). The top plot shows the mass change errors over the entire ice sheet, where shaded intervals depict data gaps longer than 1 month. The smaller graphs below show the mass change errors of the individual basins.*\n\nNote that these values are not normalized to a certain time period, as the time gap between two measurements is (slightly) variable. Uncertainties in GRACE ice mass change estimates are occasionally very high and arise from various sources. These include, amongst others, errors related to measurement noise (e.g. due to leakage of the signal outside the region of interest such as the Canadian ice caps, due to the coarse resolution of GRACE(-FO) data acquisition), uncertainties related to the Earth's gravitational field (e.g. due to the shift of the Earth's center of mass, or due to the influence of the atmosphere and oceans on the Earth's gravity field), and uncertainties related to other mass change signals that may be intertwined in the GRACE(-FO) data (e.g. those related to glacial isostatic adjustment).\n\nWe can furthermore calculate the average values (arithmetic means over time):\n\n```text\nThe arithmetic mean mass change error of the GrIS in total is 76.72 Gt with a maximum of 740.25 Gt.\nThe arithmetic mean mass change error of the GrIS basin 1 is 58.71 Gt with a maximum of 671.28 Gt.\nThe arithmetic mean mass change error of the GrIS basin 2 is 42.94 Gt with a maximum of 438.62 Gt.\nThe arithmetic mean mass change error of the GrIS basin 3 is 23.66 Gt with a maximum of 231.57 Gt.\nThe arithmetic mean mass change error of the GrIS basin 4 is 22.24 Gt with a maximum of 249.10 Gt.\nThe arithmetic mean mass change error of the GrIS basin 5 is 13.86 Gt with a maximum of 108.55 Gt.\nThe arithmetic mean mass change error of the GrIS basin 6 is 33.96 Gt with a maximum of 412.35 Gt.\nThe arithmetic mean mass change error of the GrIS basin 7 is 24.06 Gt with a maximum of 234.92 Gt.\nThe arithmetic mean mass change error of the GrIS basin 8 is 33.13 Gt with a maximum of 450.22 Gt.\n```"} {"chunk_id": "satellite_satellite-ice-sheet-mass-balance_uncertainty_q01__4cdab61e2c20", "report_id": "satellite_satellite-ice-sheet-mass-balance_uncertainty_q01", "dataset_id": "satellite-ice-sheet-mass-balance", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty_q01", "aspect_base": "uncertainty", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Gravimetric mass changes from satellite data: assessment of resolution and uncertainty of Greenland ice sheet mass changes for glaciological applications > Analysis and results > 2. Quantifying Greenland Ice Sheet mass changes and their errors in space and time > 2.2 Ice sheet mass changes uncertainty estimates", "title": "Gravimetric mass changes from satellite data: assessment of resolution and uncertainty of Greenland ice sheet mass changes for glaciological applications", "chunk_index": 10, "token_count": 558, "text_raw": "of 234.92 Gt.\nThe arithmetic mean mass change error of the GrIS basin 8 is 33.13 Gt with a maximum of 450.22 Gt.\n```\n\nFor the entire ice sheet, error values mostly range between 50 and 100 gigatonnes (Gt), with occasional spikes reaching around 400 Gt or more. When comparing the errors to the respective mass change magnitudes, it can be noted that for the majority of the time series, the errors are relative large when compared to the actual changes in mass. This suggests that the data should be handled with care for most periods. The user can decide whether to leave the months with the highest errors out if desired. It is furthermore difficult to compare these errors to GCOS requirements (GCOS, 2022) because (a) the GCOS does not propose thresholds for mass change explicitely, but solely for ice sheet volume changes, (b) the GCOS reformulates error requirements in the form of precision errors (2$\\sigma$), while the error in the GrIS mass change product contains an error product that combines precision and accuracy errors [[5](https://tc.copernicus.org/articles/7/1411/2013/)], and (c) the time difference between two GRACE measurements varies over time, and the time series also exhibits time gaps which complicates normalizing the errors to a common timeframe.\n\nConcerning the errors, it must be said that in the C3S GrIS GMB products, the most complete error estimate that could be evaluated is included [[5](https://tc.copernicus.org/articles/7/1411/2013/)]. However, in contrast to other datasets, these errors are not easily formalized as the standard deviation or the 95% confidence interval. Most datasets namely do not usually provide the total error (i.e. they neglect the accuracy error). Apart all complex processes that can introduce uncertainty (as described above), this additionally provides a partial explanation for the relatively high uncertainty values for the ice sheet GMB dataset.\n\n(section-2-3)=", "text_with_prefix": "EQC Quality Assessment: \"Gravimetric mass changes from satellite data: assessment of resolution and uncertainty of Greenland ice sheet mass changes for glaciological applications\"\nDataset: satellite-ice-sheet-mass-balance [CDS]\nAspect: uncertainty_q01 | Category: Satellite_ECVs\nSection: Gravimetric mass changes from satellite data: assessment of resolution and uncertainty of Greenland ice sheet mass changes for glaciological applications > Analysis and results > 2. Quantifying Greenland Ice Sheet mass changes and their errors in space and time > 2.2 Ice sheet mass changes uncertainty estimates\n---\nof 234.92 Gt.\nThe arithmetic mean mass change error of the GrIS basin 8 is 33.13 Gt with a maximum of 450.22 Gt.\n```\n\nFor the entire ice sheet, error values mostly range between 50 and 100 gigatonnes (Gt), with occasional spikes reaching around 400 Gt or more. When comparing the errors to the respective mass change magnitudes, it can be noted that for the majority of the time series, the errors are relative large when compared to the actual changes in mass. This suggests that the data should be handled with care for most periods. The user can decide whether to leave the months with the highest errors out if desired. It is furthermore difficult to compare these errors to GCOS requirements (GCOS, 2022) because (a) the GCOS does not propose thresholds for mass change explicitely, but solely for ice sheet volume changes, (b) the GCOS reformulates error requirements in the form of precision errors (2$\\sigma$), while the error in the GrIS mass change product contains an error product that combines precision and accuracy errors [[5](https://tc.copernicus.org/articles/7/1411/2013/)], and (c) the time difference between two GRACE measurements varies over time, and the time series also exhibits time gaps which complicates normalizing the errors to a common timeframe.\n\nConcerning the errors, it must be said that in the C3S GrIS GMB products, the most complete error estimate that could be evaluated is included [[5](https://tc.copernicus.org/articles/7/1411/2013/)]. However, in contrast to other datasets, these errors are not easily formalized as the standard deviation or the 95% confidence interval. Most datasets namely do not usually provide the total error (i.e. they neglect the accuracy error). Apart all complex processes that can introduce uncertainty (as described above), this additionally provides a partial explanation for the relatively high uncertainty values for the ice sheet GMB dataset.\n\n(section-2-3)="} {"chunk_id": "satellite_satellite-ice-sheet-mass-balance_uncertainty_q01__914e00b7b51d", "report_id": "satellite_satellite-ice-sheet-mass-balance_uncertainty_q01", "dataset_id": "satellite-ice-sheet-mass-balance", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty_q01", "aspect_base": "uncertainty", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Gravimetric mass changes from satellite data: assessment of resolution and uncertainty of Greenland ice sheet mass changes for glaciological applications > Analysis and results > 2. Quantifying Greenland Ice Sheet mass changes and their errors in space and time > 2.3 Application to an ice sheet modelling framework", "title": "Gravimetric mass changes from satellite data: assessment of resolution and uncertainty of Greenland ice sheet mass changes for glaciological applications", "chunk_index": 11, "token_count": 796, "text_raw": "Let us now go back to our user question and assess whether the data are sufficiently adequate in terms of its spatio-temporal resolution and uncertainty to be applicable in glaciological modeling applications. Given the nature of the dataset and the above analysis, it is difficult to directly use GrIS (or AIS) GRACE(-FO) ice sheet mass change data into a \"traditional\" ice sheet model for several reasons:\n- Most importantly, ice sheet models that solve the continuity equation for prognostic ice thickness changes require surface mass balance ($SMB$) (and, if desired, basal mass balance ($BMB$)) data (measured in m yr⁻¹ ice equivalents). However, GRACE(-FO) data include mass changes that combine both $SMB$ and solid ice discharge ($D$) across the grounding line, complicating direct integration into models.\n- The coarse spatial resolution (actually, since no gridded data are provided, the term \"spatial resolution\" is not applicable for this dataset) and relatively high uncertainties of GRACE(-FO) data limit their ability to be useful in ice sheet models. GRACE(-FO) data are more suited for basin-scale or ice sheet-wide analyses rather than the fine-scale resolution (pixels or grid-based data) required for pixel-by-pixel modeling in ice sheet models.\n\nTo be directly usable in glaciological models, the C3S GRACE(-FO) GMB data would need a downscaling to higher spatial resolutions (i.e. reformatting into gridded products), an integration with SMB models or ice discharge data for component separation, and improved corrections/uncertainty characterization for certain months. The data can, however, be used for other purposes. GRACE(-FO) data are, for example, valuable for the direciton calculation and monitoring of ice sheet mass changes and their sea level contribution, or for the independent validation of ice sheet mass balance estimates from other methods. For example, they can help assess results from the input-output method, which quantifies the surface mass balance ($SMB$) and solid ice discharge ($D$) components seperately [[6](https://essd.copernicus.org/articles/12/1367/2020/), [7](https://tc.copernicus.org/articles/7/469/2013/)], as well as for validating surface mass balance output from regional climate model outputs for Greenland (e.g. [[7](https://tc.copernicus.org/articles/7/469/2013/), [8](https://doi.org/10.1126/science.1178176), [9](https://doi.org/10.5194/tc-10-1933-2016)]). Studies, including Van den Broeke (2016) [[9](https://doi.org/10.5194/tc-10-1933-2016)], have shown that cumulative surface mass balance ($SMB$) minus solid ice discharge ($D$) closely aligns with the trends and variability of GRACE(-FO) mass changes for Greenland, confirming the credibility and reliability of the GRACE(-FO) mass change data, despite the relatively high error/uncertainty values provided with the data.\n\n(section-3)=", "text_with_prefix": "EQC Quality Assessment: \"Gravimetric mass changes from satellite data: assessment of resolution and uncertainty of Greenland ice sheet mass changes for glaciological applications\"\nDataset: satellite-ice-sheet-mass-balance [CDS]\nAspect: uncertainty_q01 | Category: Satellite_ECVs\nSection: Gravimetric mass changes from satellite data: assessment of resolution and uncertainty of Greenland ice sheet mass changes for glaciological applications > Analysis and results > 2. Quantifying Greenland Ice Sheet mass changes and their errors in space and time > 2.3 Application to an ice sheet modelling framework\n---\nLet us now go back to our user question and assess whether the data are sufficiently adequate in terms of its spatio-temporal resolution and uncertainty to be applicable in glaciological modeling applications. Given the nature of the dataset and the above analysis, it is difficult to directly use GrIS (or AIS) GRACE(-FO) ice sheet mass change data into a \"traditional\" ice sheet model for several reasons:\n- Most importantly, ice sheet models that solve the continuity equation for prognostic ice thickness changes require surface mass balance ($SMB$) (and, if desired, basal mass balance ($BMB$)) data (measured in m yr⁻¹ ice equivalents). However, GRACE(-FO) data include mass changes that combine both $SMB$ and solid ice discharge ($D$) across the grounding line, complicating direct integration into models.\n- The coarse spatial resolution (actually, since no gridded data are provided, the term \"spatial resolution\" is not applicable for this dataset) and relatively high uncertainties of GRACE(-FO) data limit their ability to be useful in ice sheet models. GRACE(-FO) data are more suited for basin-scale or ice sheet-wide analyses rather than the fine-scale resolution (pixels or grid-based data) required for pixel-by-pixel modeling in ice sheet models.\n\nTo be directly usable in glaciological models, the C3S GRACE(-FO) GMB data would need a downscaling to higher spatial resolutions (i.e. reformatting into gridded products), an integration with SMB models or ice discharge data for component separation, and improved corrections/uncertainty characterization for certain months. The data can, however, be used for other purposes. GRACE(-FO) data are, for example, valuable for the direciton calculation and monitoring of ice sheet mass changes and their sea level contribution, or for the independent validation of ice sheet mass balance estimates from other methods. For example, they can help assess results from the input-output method, which quantifies the surface mass balance ($SMB$) and solid ice discharge ($D$) components seperately [[6](https://essd.copernicus.org/articles/12/1367/2020/), [7](https://tc.copernicus.org/articles/7/469/2013/)], as well as for validating surface mass balance output from regional climate model outputs for Greenland (e.g. [[7](https://tc.copernicus.org/articles/7/469/2013/), [8](https://doi.org/10.1126/science.1178176), [9](https://doi.org/10.5194/tc-10-1933-2016)]). Studies, including Van den Broeke (2016) [[9](https://doi.org/10.5194/tc-10-1933-2016)], have shown that cumulative surface mass balance ($SMB$) minus solid ice discharge ($D$) closely aligns with the trends and variability of GRACE(-FO) mass changes for Greenland, confirming the credibility and reliability of the GRACE(-FO) mass change data, despite the relatively high error/uncertainty values provided with the data.\n\n(section-3)="} {"chunk_id": "satellite_satellite-ice-sheet-mass-balance_uncertainty_q01__2bef3670be44", "report_id": "satellite_satellite-ice-sheet-mass-balance_uncertainty_q01", "dataset_id": "satellite-ice-sheet-mass-balance", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty_q01", "aspect_base": "uncertainty", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Gravimetric mass changes from satellite data: assessment of resolution and uncertainty of Greenland ice sheet mass changes for glaciological applications > Analysis and results > 3. Short summary and take-home messages", "title": "Gravimetric mass changes from satellite data: assessment of resolution and uncertainty of Greenland ice sheet mass changes for glaciological applications", "chunk_index": 12, "token_count": 829, "text_raw": "The ice sheet mass change dataset, making use of the GRACE(-FO) satellite missions, is a useful tool for quantifying the total ice mass change of the Greenland (GrIS) Ice Sheet. It is one of the three main data acquisition methods that can extract ice mass change data from large spatial scales, such as the complete ice sheets, at a regular basis (the others being the altimetric and the input-output methods). The C3S ice sheet GMB data, with its long temporal extent (> 20 years), ice sheet-wide coverage and (quasi) monthly temporal resolution, effectively captures the long-term mean, trends and variability of ice sheet mass changes and the corresponding global sea level contributions. The data thus serve as a clear indicator of global climate change and water cycle changes.\n\nDespite its strengths, GRACE(-FO) has certain limitations of which the user should take note, including a low spatial resolution (only ungridded ice-sheet-wide or basin-scale changes) and (occasionally very) high uncertainty values, including for example potential signal leakage from adjacent regions (e.g. the Canadian ice caps or peripheral glaciers and ice caps). Additional uncertainties furthermore arise from necessary geophysical corrections (e.g. glacial isostatic adjustment and intertwined mass changes from outer sources). Some GRACE(-FO) error values exhibit very high values, requiring careful interpretation. Additionally, data gaps (e.g. between the GRACE and GRACE-FO missions) affect the time series continuity, but these do not generally affect the overal trend and magnitude of the cumulative mass changes in the product (e.g. [[10](https://doi.org/10.1029/2020GL087291)]). Nevertheless, unlike the input-output method (where SMB is directly needed) and the altimetric method (where it is necessary to obtain the firn density profiles for the volume-to-mass conversion), the SMB is not explicitly involved in the processing of GRACE(-FO) data, nor is it separated from ice flow dynamics. This independence negates uncertainties due to potential discrepancies between SMB model outputs from different centers and is a significant advantage of this method.\n\nMoreover, due to the nature of the C3S GRACE(-FO) data presented on the CDS, they cannot be used directly as data assimilation into any glaciological model that requires the usage of pixel-based data to model ice thickness changes (e.g. an ice sheet model that solves the prognostic continuity equation on a grid). In essence, the results of this notebook therefore indicate that the product is inadequate to be used in most \"traditional\" glaciological modelling efforts because it is (1) only a spatially aggregated time series, (2) does not include SMB separately (which is to be used in the prognostic continuity equation for ice thickness changes), and (3) has high uncertainty in certain months. Nevertheless, it certainly has other applications and is highly relevant for the direct calculation and monitoring of ice sheet mass changes and their sea level contribution, and/or the validation of mass change estimates from other independent methods (e.g. the input-output method or the validation of the temporal variability of a surface mass balance model from a regional climate model) (e.g. [[7](https://tc.copernicus.org/articles/7/469/2013/), [8](https://doi.org/10.1126/science.1178176), [9](https://doi.org/10.5194/tc-10-1933-2016)]).", "text_with_prefix": "EQC Quality Assessment: \"Gravimetric mass changes from satellite data: assessment of resolution and uncertainty of Greenland ice sheet mass changes for glaciological applications\"\nDataset: satellite-ice-sheet-mass-balance [CDS]\nAspect: uncertainty_q01 | Category: Satellite_ECVs\nSection: Gravimetric mass changes from satellite data: assessment of resolution and uncertainty of Greenland ice sheet mass changes for glaciological applications > Analysis and results > 3. Short summary and take-home messages\n---\nThe ice sheet mass change dataset, making use of the GRACE(-FO) satellite missions, is a useful tool for quantifying the total ice mass change of the Greenland (GrIS) Ice Sheet. It is one of the three main data acquisition methods that can extract ice mass change data from large spatial scales, such as the complete ice sheets, at a regular basis (the others being the altimetric and the input-output methods). The C3S ice sheet GMB data, with its long temporal extent (> 20 years), ice sheet-wide coverage and (quasi) monthly temporal resolution, effectively captures the long-term mean, trends and variability of ice sheet mass changes and the corresponding global sea level contributions. The data thus serve as a clear indicator of global climate change and water cycle changes.\n\nDespite its strengths, GRACE(-FO) has certain limitations of which the user should take note, including a low spatial resolution (only ungridded ice-sheet-wide or basin-scale changes) and (occasionally very) high uncertainty values, including for example potential signal leakage from adjacent regions (e.g. the Canadian ice caps or peripheral glaciers and ice caps). Additional uncertainties furthermore arise from necessary geophysical corrections (e.g. glacial isostatic adjustment and intertwined mass changes from outer sources). Some GRACE(-FO) error values exhibit very high values, requiring careful interpretation. Additionally, data gaps (e.g. between the GRACE and GRACE-FO missions) affect the time series continuity, but these do not generally affect the overal trend and magnitude of the cumulative mass changes in the product (e.g. [[10](https://doi.org/10.1029/2020GL087291)]). Nevertheless, unlike the input-output method (where SMB is directly needed) and the altimetric method (where it is necessary to obtain the firn density profiles for the volume-to-mass conversion), the SMB is not explicitly involved in the processing of GRACE(-FO) data, nor is it separated from ice flow dynamics. This independence negates uncertainties due to potential discrepancies between SMB model outputs from different centers and is a significant advantage of this method.\n\nMoreover, due to the nature of the C3S GRACE(-FO) data presented on the CDS, they cannot be used directly as data assimilation into any glaciological model that requires the usage of pixel-based data to model ice thickness changes (e.g. an ice sheet model that solves the prognostic continuity equation on a grid). In essence, the results of this notebook therefore indicate that the product is inadequate to be used in most \"traditional\" glaciological modelling efforts because it is (1) only a spatially aggregated time series, (2) does not include SMB separately (which is to be used in the prognostic continuity equation for ice thickness changes), and (3) has high uncertainty in certain months. Nevertheless, it certainly has other applications and is highly relevant for the direct calculation and monitoring of ice sheet mass changes and their sea level contribution, and/or the validation of mass change estimates from other independent methods (e.g. the input-output method or the validation of the temporal variability of a surface mass balance model from a regional climate model) (e.g. [[7](https://tc.copernicus.org/articles/7/469/2013/), [8](https://doi.org/10.1126/science.1178176), [9](https://doi.org/10.5194/tc-10-1933-2016)])."} {"chunk_id": "satellite_satellite-ice-sheet-mass-balance_uncertainty_q01__38be60158ac7", "report_id": "satellite_satellite-ice-sheet-mass-balance_uncertainty_q01", "dataset_id": "satellite-ice-sheet-mass-balance", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty_q01", "aspect_base": "uncertainty", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Gravimetric mass changes from satellite data: assessment of resolution and uncertainty of Greenland ice sheet mass changes for glaciological applications > ℹ️ If you want to know more > Key resources", "title": "Gravimetric mass changes from satellite data: assessment of resolution and uncertainty of Greenland ice sheet mass changes for glaciological applications", "chunk_index": 13, "token_count": 366, "text_raw": "- \"[Gravimetric mass balance data for the Antarctic and Greenland ice sheets from 2003 to 2022 derived from satellite observations](https://cds.climate.copernicus.eu/datasets/satellite-ice-sheet-mass-balance?tab=overview)\" on the CDS.\n- [Documentation on the CDS](https://cds.climate.copernicus.eu/datasets/satellite-ice-sheet-mass-balance?tab=overview) and the [ECMWF Confluence Wiki](https://confluence.ecmwf.int/pages/viewpage.action?pageId=355348671) (Copernicus Knowledge Base).\n- [Copernicus climate change indicators: ice sheets](https://climate.copernicus.eu/climate-indicators/ice-sheets)\n- [The data portal with GrIS GMB data from the data provider (TU Dresden)](https://data1.geo.tu-dresden.de/gis_gmb/index.html)\n- [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control` prepared by [B-Open](https://www.bopen.eu).", "text_with_prefix": "EQC Quality Assessment: \"Gravimetric mass changes from satellite data: assessment of resolution and uncertainty of Greenland ice sheet mass changes for glaciological applications\"\nDataset: satellite-ice-sheet-mass-balance [CDS]\nAspect: uncertainty_q01 | Category: Satellite_ECVs\nSection: Gravimetric mass changes from satellite data: assessment of resolution and uncertainty of Greenland ice sheet mass changes for glaciological applications > ℹ️ If you want to know more > Key resources\n---\n- \"[Gravimetric mass balance data for the Antarctic and Greenland ice sheets from 2003 to 2022 derived from satellite observations](https://cds.climate.copernicus.eu/datasets/satellite-ice-sheet-mass-balance?tab=overview)\" on the CDS.\n- [Documentation on the CDS](https://cds.climate.copernicus.eu/datasets/satellite-ice-sheet-mass-balance?tab=overview) and the [ECMWF Confluence Wiki](https://confluence.ecmwf.int/pages/viewpage.action?pageId=355348671) (Copernicus Knowledge Base).\n- [Copernicus climate change indicators: ice sheets](https://climate.copernicus.eu/climate-indicators/ice-sheets)\n- [The data portal with GrIS GMB data from the data provider (TU Dresden)](https://data1.geo.tu-dresden.de/gis_gmb/index.html)\n- [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control` prepared by [B-Open](https://www.bopen.eu)."} {"chunk_id": "satellite_satellite-ice-sheet-mass-balance_uncertainty_q01__cb7e74852a9f", "report_id": "satellite_satellite-ice-sheet-mass-balance_uncertainty_q01", "dataset_id": "satellite-ice-sheet-mass-balance", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty_q01", "aspect_base": "uncertainty", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Gravimetric mass changes from satellite data: assessment of resolution and uncertainty of Greenland ice sheet mass changes for glaciological applications > ℹ️ If you want to know more > References", "title": "Gravimetric mass changes from satellite data: assessment of resolution and uncertainty of Greenland ice sheet mass changes for glaciological applications", "chunk_index": 14, "token_count": 1061, "text_raw": "- [[1](https://doi.org/10.1007/s10712-016-9398-7)] Forsberg, R., Sørensen, L.S. and Simonsen, S.B. (2017). Greenland and Antarctica Ice Sheet Mass Changes and Effects on Global Sea Level. Surv. Geophys., 38, 89–104. https://doi.org/10.1007/s10712-016-9398-7\n\n- [[2](https://doi.org/10.3390/geosciences9100415)] Groh, A., Horwath, M., Horvath, A., Meister, R., Sørensen, L.S., Barletta, V.R., Forsberg, R., Wouters, B., Ditmar, P., Ran, J., Klees, R., Su, X., Shang, K., Guo, J., Shum, C.K., Schrama, E., and Shepherd, A. (2019). Evaluating GRACE Mass Change Time Series for the Antarctic and Greenland Ice Sheet, Geosciences, 9(10). https://doi.org/10.3390/geosciences9100415\n\n- [[3](https://earth.gsfc.nasa.gov/cryo/data/polar-altimetry/antarctic-and-greenland-drainage-systems)] Zwally, H., Giovinetto, M., Beckley, M., and Saba, J. (2012). Antarctic and Greenland drainage systems, GSFC cryospheric sciences laboratory. URL: https://earth.gsfc.nasa.gov/cryo/data/polar-altimetry/antarctic-and-greenland-drainage-systems\n\n- [[4](https://essd.copernicus.org/articles/15/1597/2023/)] Otosaka, I. N., Shepherd, A., Ivins, E. R., Schlegel, N.-J., Amory, C., van den Broeke, M. R., Horwath, M., Joughin, I., King, M. D., Krinner, G., Nowicki, S., Payne, A. J., Rignot, E., Scambos, T., Simon, K. M., Smith, B. E., Sørensen, L. S., Velicogna, I., Whitehouse, P. L., A, G., Agosta, C., Ahlstrøm, A. P., Blazquez, A., Colgan, W., Engdahl, M. E., Fettweis, X., Forsberg, R., Gallée, H., Gardner, A., Gilbert, L., Gourmelen, N., Groh, A., Gunter, B. C., Harig, C., Helm, V., Khan, S. A., Kittel, C., Konrad, H., Langen, P. L., Lecavalier, B. S., Liang, C.-C., Loomis, B. D., McMillan, M., Melini, D., Mernild, S. H., Mottram, R., Mouginot, J., Nilsson, J., Noël, B., Pattle, M. E., Peltier, W. R., Pie, N., Roca, M., Sasgen, I., Save, H. V., Seo, K.-W., Scheuchl, B., Schrama, E. J. O., Schröder, L., Simonsen, S. B., Slater, T., Spada, G., Sutterley, T. C., Vishwakarma, B. D., van Wessem, J. M., Wiese, D., van der Wal, W., and Wouters, B. (2023). Mass balance of the Greenland and Antarctic ice sheets from 1992 to 2020, Earth Syst. Sci. Data, 15, 1597–1616, https://doi.org/10.5194/essd-15-1597-2023\n\n- [[5](https://tc.copernicus.org/articles/7/1411/2013/)] Barletta, V. R., Sørensen, L. S., and Forsberg, R. (2013). Scatter of mass changes estimates at basin scale for Greenland and Antarctica, The Cryosphere, 7, 1411–1432, https://doi.org/10.5194/tc-7-1411-2013", "text_with_prefix": "EQC Quality Assessment: \"Gravimetric mass changes from satellite data: assessment of resolution and uncertainty of Greenland ice sheet mass changes for glaciological applications\"\nDataset: satellite-ice-sheet-mass-balance [CDS]\nAspect: uncertainty_q01 | Category: Satellite_ECVs\nSection: Gravimetric mass changes from satellite data: assessment of resolution and uncertainty of Greenland ice sheet mass changes for glaciological applications > ℹ️ If you want to know more > References\n---\n- [[1](https://doi.org/10.1007/s10712-016-9398-7)] Forsberg, R., Sørensen, L.S. and Simonsen, S.B. (2017). Greenland and Antarctica Ice Sheet Mass Changes and Effects on Global Sea Level. Surv. Geophys., 38, 89–104. https://doi.org/10.1007/s10712-016-9398-7\n\n- [[2](https://doi.org/10.3390/geosciences9100415)] Groh, A., Horwath, M., Horvath, A., Meister, R., Sørensen, L.S., Barletta, V.R., Forsberg, R., Wouters, B., Ditmar, P., Ran, J., Klees, R., Su, X., Shang, K., Guo, J., Shum, C.K., Schrama, E., and Shepherd, A. (2019). Evaluating GRACE Mass Change Time Series for the Antarctic and Greenland Ice Sheet, Geosciences, 9(10). https://doi.org/10.3390/geosciences9100415\n\n- [[3](https://earth.gsfc.nasa.gov/cryo/data/polar-altimetry/antarctic-and-greenland-drainage-systems)] Zwally, H., Giovinetto, M., Beckley, M., and Saba, J. (2012). Antarctic and Greenland drainage systems, GSFC cryospheric sciences laboratory. URL: https://earth.gsfc.nasa.gov/cryo/data/polar-altimetry/antarctic-and-greenland-drainage-systems\n\n- [[4](https://essd.copernicus.org/articles/15/1597/2023/)] Otosaka, I. N., Shepherd, A., Ivins, E. R., Schlegel, N.-J., Amory, C., van den Broeke, M. R., Horwath, M., Joughin, I., King, M. D., Krinner, G., Nowicki, S., Payne, A. J., Rignot, E., Scambos, T., Simon, K. M., Smith, B. E., Sørensen, L. S., Velicogna, I., Whitehouse, P. L., A, G., Agosta, C., Ahlstrøm, A. P., Blazquez, A., Colgan, W., Engdahl, M. E., Fettweis, X., Forsberg, R., Gallée, H., Gardner, A., Gilbert, L., Gourmelen, N., Groh, A., Gunter, B. C., Harig, C., Helm, V., Khan, S. A., Kittel, C., Konrad, H., Langen, P. L., Lecavalier, B. S., Liang, C.-C., Loomis, B. D., McMillan, M., Melini, D., Mernild, S. H., Mottram, R., Mouginot, J., Nilsson, J., Noël, B., Pattle, M. E., Peltier, W. R., Pie, N., Roca, M., Sasgen, I., Save, H. V., Seo, K.-W., Scheuchl, B., Schrama, E. J. O., Schröder, L., Simonsen, S. B., Slater, T., Spada, G., Sutterley, T. C., Vishwakarma, B. D., van Wessem, J. M., Wiese, D., van der Wal, W., and Wouters, B. (2023). Mass balance of the Greenland and Antarctic ice sheets from 1992 to 2020, Earth Syst. Sci. Data, 15, 1597–1616, https://doi.org/10.5194/essd-15-1597-2023\n\n- [[5](https://tc.copernicus.org/articles/7/1411/2013/)] Barletta, V. R., Sørensen, L. S., and Forsberg, R. (2013). Scatter of mass changes estimates at basin scale for Greenland and Antarctica, The Cryosphere, 7, 1411–1432, https://doi.org/10.5194/tc-7-1411-2013"} {"chunk_id": "satellite_satellite-ice-sheet-mass-balance_uncertainty_q01__60a2e8258d6d", "report_id": "satellite_satellite-ice-sheet-mass-balance_uncertainty_q01", "dataset_id": "satellite-ice-sheet-mass-balance", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty_q01", "aspect_base": "uncertainty", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Gravimetric mass changes from satellite data: assessment of resolution and uncertainty of Greenland ice sheet mass changes for glaciological applications > ℹ️ If you want to know more > References", "title": "Gravimetric mass changes from satellite data: assessment of resolution and uncertainty of Greenland ice sheet mass changes for glaciological applications", "chunk_index": 15, "token_count": 799, "text_raw": "). Scatter of mass changes estimates at basin scale for Greenland and Antarctica, The Cryosphere, 7, 1411–1432, https://doi.org/10.5194/tc-7-1411-2013\n\n- [[6](https://essd.copernicus.org/articles/12/1367/2020/)] Mankoff, K. D., Solgaard, A., Colgan, W., Ahlstrøm, A. P., Khan, S. A., and Fausto, R. S. (2020). Greenland Ice Sheet solid ice discharge from 1986 through March 2020, Earth Syst. Sci. Data, 12, 1367–1383, https://doi.org/10.5194/essd-12-1367-2020\n\n- [[7](https://tc.copernicus.org/articles/7/469/2013/)] Fettweis, X., Franco, B., Tedesco, M., van Angelen, J. H., Lenaerts, J. T. M., van den Broeke, M. R., and Gallée, H. (2013). Estimating the Greenland ice sheet surface mass balance contribution to future sea level rise using the regional atmospheric climate model MAR, The Cryosphere, 7, 469–489, https://doi.org/10.5194/tc-7-469-2013\n\n- [[8](https://doi.org/10.1126/science.1178176)] van den Broeke, M., Bamber, J., Ettema, J., Rignot, E., Schrama, E., van de Berg, W. J., van Meijgaard, E., Velicogna, I., and Wouters, B. (2009). Partitioning recent Greenland mass loss. Science, 326(5955), 984-986. https://doi.org/10.1126/science.1178176\n\n- [[9](https://doi.org/10.5194/tc-10-1933-2016)] van den Broeke, M., Enderlin, E. M., Howat, I. M., Kuipers Munneke, P., Noël, B. P. Y., van de Berg, W. J., van Meijgaard, E., and Wouters, B. (2016). On the recent contribution of the Greenland ice sheet to sea level change, The Cryosphere, 10, 1933–1946, https://doi.org/10.5194/tc-10-1933-2016\n\n- [[10](https://doi.org/10.1029/2020GL087291)] Velicogna, I., Mohajerani, Y., A, G., Landerer, F., Mouginot, J., Noel, B., Rignot, E., Sutterley, T., van den Broeke, M., van Wessem, M., and Wiese, D. (2020). Continuity of Ice Sheet Mass Loss in Greenland and Antarctica from the GRACE and GRACE Follow-On Missions. Geophys. Res. Lett. 47. https://doi.org/10.1029/2020GL087291", "text_with_prefix": "EQC Quality Assessment: \"Gravimetric mass changes from satellite data: assessment of resolution and uncertainty of Greenland ice sheet mass changes for glaciological applications\"\nDataset: satellite-ice-sheet-mass-balance [CDS]\nAspect: uncertainty_q01 | Category: Satellite_ECVs\nSection: Gravimetric mass changes from satellite data: assessment of resolution and uncertainty of Greenland ice sheet mass changes for glaciological applications > ℹ️ If you want to know more > References\n---\n). Scatter of mass changes estimates at basin scale for Greenland and Antarctica, The Cryosphere, 7, 1411–1432, https://doi.org/10.5194/tc-7-1411-2013\n\n- [[6](https://essd.copernicus.org/articles/12/1367/2020/)] Mankoff, K. D., Solgaard, A., Colgan, W., Ahlstrøm, A. P., Khan, S. A., and Fausto, R. S. (2020). Greenland Ice Sheet solid ice discharge from 1986 through March 2020, Earth Syst. Sci. Data, 12, 1367–1383, https://doi.org/10.5194/essd-12-1367-2020\n\n- [[7](https://tc.copernicus.org/articles/7/469/2013/)] Fettweis, X., Franco, B., Tedesco, M., van Angelen, J. H., Lenaerts, J. T. M., van den Broeke, M. R., and Gallée, H. (2013). Estimating the Greenland ice sheet surface mass balance contribution to future sea level rise using the regional atmospheric climate model MAR, The Cryosphere, 7, 469–489, https://doi.org/10.5194/tc-7-469-2013\n\n- [[8](https://doi.org/10.1126/science.1178176)] van den Broeke, M., Bamber, J., Ettema, J., Rignot, E., Schrama, E., van de Berg, W. J., van Meijgaard, E., Velicogna, I., and Wouters, B. (2009). Partitioning recent Greenland mass loss. Science, 326(5955), 984-986. https://doi.org/10.1126/science.1178176\n\n- [[9](https://doi.org/10.5194/tc-10-1933-2016)] van den Broeke, M., Enderlin, E. M., Howat, I. M., Kuipers Munneke, P., Noël, B. P. Y., van de Berg, W. J., van Meijgaard, E., and Wouters, B. (2016). On the recent contribution of the Greenland ice sheet to sea level change, The Cryosphere, 10, 1933–1946, https://doi.org/10.5194/tc-10-1933-2016\n\n- [[10](https://doi.org/10.1029/2020GL087291)] Velicogna, I., Mohajerani, Y., A, G., Landerer, F., Mouginot, J., Noel, B., Rignot, E., Sutterley, T., van den Broeke, M., van Wessem, M., and Wiese, D. (2020). Continuity of Ice Sheet Mass Loss in Greenland and Antarctica from the GRACE and GRACE Follow-On Missions. Geophys. Res. Lett. 47. https://doi.org/10.1029/2020GL087291"} {"chunk_id": "satellite_satellite-fire-burned-area_climate-and-weather-extremes_q01__e56a8757c3b1", "report_id": "satellite_satellite-fire-burned-area_climate-and-weather-extremes_q01", "dataset_id": "satellite-fire-burned-area", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q01", "aspect_base": "climate-and-weather-extremes", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite fire burned area completeness for seasonal climatology maps > Quality assessment question", "title": "Satellite fire burned area completeness for seasonal climatology maps", "chunk_index": 0, "token_count": 202, "text_raw": "* **How well can we disclose what are the spatial patterns of the total burned area per season over the Iberian Peninsula?**\n\nIn this Use Case we will access the Fire burned area from 2001 to present derived from satellite observations (henceforth, FIRE) data from the Climate Data Store (CDS) of the Copernicus Climate Change Service (C3S) and analyse the spatial patterns of the FIRE reliability over a given Area of Interest (AoI), considering the Nomenclature of Territorial Units for Statistics, at the level 2, which are used for decision monitoring and decision-making purposes (see section 3)’", "text_with_prefix": "EQC Quality Assessment: \"Satellite fire burned area completeness for seasonal climatology maps\"\nDataset: satellite-fire-burned-area [CDS]\nAspect: climate-and-weather-extremes_q01 | Category: Satellite_ECVs\nSection: Satellite fire burned area completeness for seasonal climatology maps > Quality assessment question\n---\n* **How well can we disclose what are the spatial patterns of the total burned area per season over the Iberian Peninsula?**\n\nIn this Use Case we will access the Fire burned area from 2001 to present derived from satellite observations (henceforth, FIRE) data from the Climate Data Store (CDS) of the Copernicus Climate Change Service (C3S) and analyse the spatial patterns of the FIRE reliability over a given Area of Interest (AoI), considering the Nomenclature of Territorial Units for Statistics, at the level 2, which are used for decision monitoring and decision-making purposes (see section 3)’"} {"chunk_id": "satellite_satellite-fire-burned-area_climate-and-weather-extremes_q01__ccf8530e6074", "report_id": "satellite_satellite-fire-burned-area_climate-and-weather-extremes_q01", "dataset_id": "satellite-fire-burned-area", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q01", "aspect_base": "climate-and-weather-extremes", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite fire burned area completeness for seasonal climatology maps > Quality assessment statement", "title": "Satellite fire burned area completeness for seasonal climatology maps", "chunk_index": 1, "token_count": 1063, "text_raw": "These are the key outcomes of this assessment\n\n* The dataset accurately reflects the burned area values observed during significant fire events, such as the summer of 2003 in Portugal, with a high level of precision. The burned area recorded (more than 3500 km² in Centro PT and Alentejo regions) is consintent with the Portuguese national statistics and findings by Lourenço and Luciano (2018). This consistency suggests that the dataset accurately represents extreme fire events.\n\n* The results of this notebook shows that the spatial resolution is sufficient to detect regional differences in burned area, and the temporal resolution is adequate to highlight annual and seasonal variations, ensuring a comprehensive understanding of fire patterns over time.\n\n* The stability of the dataset is evidenced by its consistency with other credible sources, such as Grünig et al., 2023, which supports the findings of significant burned areas in key regions during specific years. The dataset reliably identifies both extreme fire events and regions with consistently low burned areas, such as La Rioja and Murcia.\n\nattachment:b0838746-6694-43a3-8244-486eb1346215.png\n---\nheight: 500px\n---\nThe figure shows the regional climatology of the burned area over the period 2001 to 2019. The averaged burned area, and corresponding spatial pattern, can serve as a baseline for wildfire prevention and monitoring.\n```\n\n## 📋 Methodology\n\n[](code-section-1)\n\n[](code-section-2)\n\n[](code-section-3)\n\n[](code-section-4)\n\n- [](code-section-4.1)\n\n- [](code-section-4.2)\n\n- [](code-section-4.3)\n\n[](code-section-5)\n\n## 📈 Analysis and results\n\n(code-section-1)=\n### 1. Define the AoI, search and download FIRE dataset\nBefore we begin, we must prepare our environment. This includes installing the Application Programming Interface (API) of the CDS and importing the various python libraries that we will need.\n\n#### Install CDS API\nTo install the CDS API, run the following command. We use an exclamation mark to pass the command to the shell (not to the Python interpreter).\nIf you already have the CDS API installed, you can skip or comment this step.\n\n#### Import all the libraries/packages\n\nWe will be working with data in NetCDF format. To best handle this data we will use libraries for working with multidimensional arrays, in particular Xarray. We will also need libraries for plotting and viewing data, in this case we will use Matplotlib and Cartopy.\n\n#### Data overview\n\nTo search for data, visit the CDS website: http://cds.climate.copernicus.eu Here you can search for 'Satellite observations' data using the search bar. The data we need for this tutorial is the Fire burned area from 2001 to present derived from satellite observations. This data provides global (at grid scale) and continents (at pixel scale) information of total burned area (BA), following the Global Climate Observing System (GCOS) convention. The dataset is provided with a high spatial resolution, through the analysis of reflectance changes from medium resolution sensors (Terra MODIS, Sentinel-3 OLCI) supported by thermal information.\n\nThe BA is identified by the date of the first detection of the burned signal in the case of the pixel product, and by the total BA per grid cell in the case of the grid product. Information such as land cover class, confidence level or standard error are also provided.\n\nThe temporal resolution varies per version, from 15 days (at grid scale) to monthly (at pixel scale), and its vertical resolution corresponds to the surface (single level).\n\nDifferent versions are available, offering the first global BA time series at 250 m resolution (FireCCI v5.0cds and FireCCI v5.1cds developed as part of the Fire ECV Climate Change Initiative Project (Fire CCI)). This algorithm was adapted to Sentinel-3 OLCI data to create the C3S v1.0 burned area product, extending the BA database to the present.\n\nHaving selected the correct dataset, we now need to specify what product type, variables, temporal and geographic coverage we are interested in. In this Use case, we will retrieve information of the grid ESA-CCI version. These can all be selected in the “Download data” tab from the CDS. In this tab a form appears in which we will select the following parameters to download, for example:\n\n- Origin: ESA-CCI\n- Sensor: MODIS\n- Variable: Grid\n- Version: 5.1.1cds\n- Region: all\n- Year: 2001 to 2019\n- Month: all\n- Nominal day: 01\n- Format: Zip file (.zip)", "text_with_prefix": "EQC Quality Assessment: \"Satellite fire burned area completeness for seasonal climatology maps\"\nDataset: satellite-fire-burned-area [CDS]\nAspect: climate-and-weather-extremes_q01 | Category: Satellite_ECVs\nSection: Satellite fire burned area completeness for seasonal climatology maps > Quality assessment statement\n---\nThese are the key outcomes of this assessment\n\n* The dataset accurately reflects the burned area values observed during significant fire events, such as the summer of 2003 in Portugal, with a high level of precision. The burned area recorded (more than 3500 km² in Centro PT and Alentejo regions) is consintent with the Portuguese national statistics and findings by Lourenço and Luciano (2018). This consistency suggests that the dataset accurately represents extreme fire events.\n\n* The results of this notebook shows that the spatial resolution is sufficient to detect regional differences in burned area, and the temporal resolution is adequate to highlight annual and seasonal variations, ensuring a comprehensive understanding of fire patterns over time.\n\n* The stability of the dataset is evidenced by its consistency with other credible sources, such as Grünig et al., 2023, which supports the findings of significant burned areas in key regions during specific years. The dataset reliably identifies both extreme fire events and regions with consistently low burned areas, such as La Rioja and Murcia.\n\nattachment:b0838746-6694-43a3-8244-486eb1346215.png\n---\nheight: 500px\n---\nThe figure shows the regional climatology of the burned area over the period 2001 to 2019. The averaged burned area, and corresponding spatial pattern, can serve as a baseline for wildfire prevention and monitoring.\n```\n\n## 📋 Methodology\n\n[](code-section-1)\n\n[](code-section-2)\n\n[](code-section-3)\n\n[](code-section-4)\n\n- [](code-section-4.1)\n\n- [](code-section-4.2)\n\n- [](code-section-4.3)\n\n[](code-section-5)\n\n## 📈 Analysis and results\n\n(code-section-1)=\n### 1. Define the AoI, search and download FIRE dataset\nBefore we begin, we must prepare our environment. This includes installing the Application Programming Interface (API) of the CDS and importing the various python libraries that we will need.\n\n#### Install CDS API\nTo install the CDS API, run the following command. We use an exclamation mark to pass the command to the shell (not to the Python interpreter).\nIf you already have the CDS API installed, you can skip or comment this step.\n\n#### Import all the libraries/packages\n\nWe will be working with data in NetCDF format. To best handle this data we will use libraries for working with multidimensional arrays, in particular Xarray. We will also need libraries for plotting and viewing data, in this case we will use Matplotlib and Cartopy.\n\n#### Data overview\n\nTo search for data, visit the CDS website: http://cds.climate.copernicus.eu Here you can search for 'Satellite observations' data using the search bar. The data we need for this tutorial is the Fire burned area from 2001 to present derived from satellite observations. This data provides global (at grid scale) and continents (at pixel scale) information of total burned area (BA), following the Global Climate Observing System (GCOS) convention. The dataset is provided with a high spatial resolution, through the analysis of reflectance changes from medium resolution sensors (Terra MODIS, Sentinel-3 OLCI) supported by thermal information.\n\nThe BA is identified by the date of the first detection of the burned signal in the case of the pixel product, and by the total BA per grid cell in the case of the grid product. Information such as land cover class, confidence level or standard error are also provided.\n\nThe temporal resolution varies per version, from 15 days (at grid scale) to monthly (at pixel scale), and its vertical resolution corresponds to the surface (single level).\n\nDifferent versions are available, offering the first global BA time series at 250 m resolution (FireCCI v5.0cds and FireCCI v5.1cds developed as part of the Fire ECV Climate Change Initiative Project (Fire CCI)). This algorithm was adapted to Sentinel-3 OLCI data to create the C3S v1.0 burned area product, extending the BA database to the present.\n\nHaving selected the correct dataset, we now need to specify what product type, variables, temporal and geographic coverage we are interested in. In this Use case, we will retrieve information of the grid ESA-CCI version. These can all be selected in the “Download data” tab from the CDS. In this tab a form appears in which we will select the following parameters to download, for example:\n\n- Origin: ESA-CCI\n- Sensor: MODIS\n- Variable: Grid\n- Version: 5.1.1cds\n- Region: all\n- Year: 2001 to 2019\n- Month: all\n- Nominal day: 01\n- Format: Zip file (.zip)"} {"chunk_id": "satellite_satellite-fire-burned-area_climate-and-weather-extremes_q01__1a9195ed1f18", "report_id": "satellite_satellite-fire-burned-area_climate-and-weather-extremes_q01", "dataset_id": "satellite-fire-burned-area", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q01", "aspect_base": "climate-and-weather-extremes", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite fire burned area completeness for seasonal climatology maps > Quality assessment statement", "title": "Satellite fire burned area completeness for seasonal climatology maps", "chunk_index": 2, "token_count": 1080, "text_raw": "- Version: 5.1.1cds\n- Region: all\n- Year: 2001 to 2019\n- Month: all\n- Nominal day: 01\n- Format: Zip file (.zip)\n\nAt the end of the download form, select `Show API request`. This will reveal a block of code, which you can simply copy and paste into a cell of your Jupyter Notebook.\n\nHaving copied the API request to a Jupyter Notebook cell, running it will retrieve and download the data you requested into your local directory. However, before you run it, the terms and conditions of this particular dataset need to have been accepted directly at the CDS website. The option to view and accept these conditions is given at the end of the download form, just above the `Show API request` option. In addition, it is also useful to define the time period and AoI parameters and edit the request accordingly, as exemplified in the cells below.\n\nRegion of interest\nShapefile with regions\n\nDefine request\n\nDownload and regionalize\nReindex using year/month (shift months + 1)\nConvert units from m2 to km2\n\n(code-section-2)=\n### 2. Inspect and view data\n\nNow that we have downloaded the data, we can inspect it. We have requested the data in NetCDF format. This is a commonly used format for array-oriented scientific data. To read and process this data we will make use of the Xarray library. Xarray is an open source project and Python package that makes working with labelled multi-dimensional arrays simple and efficient. We will read the data from our NetCDF file into an xarray.Dataset.\n\nLet's inspect the years to see which have the highest and the lowest values of burned area.\n\nPlot BA time series\n\nFrom these statistics and the corresponding maps, we can conclude that 2003 and 2018 have the highest and the lowest sum burned area values (respectively). We can also see some hotspot regions in the centre of Portugal in August.\n\n(code-section-3)=\n### 3. Define the unit of analysis: Nomenclature of territorial units for statistics, level 2 (NUTS 2)\n\nThe [NUTS](https://ec.europa.eu/eurostat/web/nuts) are a hierarchical system divided into 3 levels. NUTS 1 correspond to major socio-economic regions, NUTS 2 correspond to basic regions for the application of regional policies, and NUTS 3 correspond to small regions for specific diagnoses. Additionally, a NUTS 0 level, usually co-incident with national boundaries is also available. The NUTS legislation is periodically amended; therefore multiple years are available for download.\n\nIn this user question, it is used the most updated version of the NUTS corresponding to version of 2021. In this study case, NUTS 2 will be used, providing the information regarding the main regions/parcels of the Iberian Peninsula.\n\nNote: The sum of burned area and the climatology are calculated per each NUTS 2 region. Generated maps, bar charts and plots will take some minutes to process all the months, years and regions.\n\nMerging BA with NUTS 2 geodataframe\n\n(code-section-4)=\n### 4. Annual Summaries and Summer Climatology Analysis of the Burned Area (JJA of 2001-2019)\n\nThe step below produces the Annual Summaries of BA by representing the Bar Charts and the Maps of the Sum of the Burned Area per NUTS 2, per month/year. In this case, we are analysing the Summer (JJA) months, which is when wildfires tend to be more frequent, in the Iberian Peninsula Region.\n\nIn addition, the Summer Climatology for BA aims to provide an overview of fire activity over a long-term period, such as multiple years or decades.\n\nAccording to the European Environment Agency (EEA (2021) Europe's changing climate hazards — an index-based interactive EEA report, Report no. 15/2021, EEA. DOI: 10.2800/458052) methodology, the following steps are conducted:\n- the sum of Burned Area is calculated for each year within the defined period; \n- the resulting timeseries is average across the number of years.\n\nThis average BA represents the typical magnitude of burned areas during the summer period (June, July and August, from 2001 to 2019).\n\nThe Annual Summaries and Summer Climatology are represented in several ways, such as maps, bar charts and timeseries plots.\n\n(code-section-4.1)=\n#### 4.1. Annual Summaries - Maps and Bar Charts of total Burned Area for the summer season, per NUTS regions per year\n\nPloting maps and bar charts of annual BA by NUTS 2 region, considering the Summer season", "text_with_prefix": "EQC Quality Assessment: \"Satellite fire burned area completeness for seasonal climatology maps\"\nDataset: satellite-fire-burned-area [CDS]\nAspect: climate-and-weather-extremes_q01 | Category: Satellite_ECVs\nSection: Satellite fire burned area completeness for seasonal climatology maps > Quality assessment statement\n---\n- Version: 5.1.1cds\n- Region: all\n- Year: 2001 to 2019\n- Month: all\n- Nominal day: 01\n- Format: Zip file (.zip)\n\nAt the end of the download form, select `Show API request`. This will reveal a block of code, which you can simply copy and paste into a cell of your Jupyter Notebook.\n\nHaving copied the API request to a Jupyter Notebook cell, running it will retrieve and download the data you requested into your local directory. However, before you run it, the terms and conditions of this particular dataset need to have been accepted directly at the CDS website. The option to view and accept these conditions is given at the end of the download form, just above the `Show API request` option. In addition, it is also useful to define the time period and AoI parameters and edit the request accordingly, as exemplified in the cells below.\n\nRegion of interest\nShapefile with regions\n\nDefine request\n\nDownload and regionalize\nReindex using year/month (shift months + 1)\nConvert units from m2 to km2\n\n(code-section-2)=\n### 2. Inspect and view data\n\nNow that we have downloaded the data, we can inspect it. We have requested the data in NetCDF format. This is a commonly used format for array-oriented scientific data. To read and process this data we will make use of the Xarray library. Xarray is an open source project and Python package that makes working with labelled multi-dimensional arrays simple and efficient. We will read the data from our NetCDF file into an xarray.Dataset.\n\nLet's inspect the years to see which have the highest and the lowest values of burned area.\n\nPlot BA time series\n\nFrom these statistics and the corresponding maps, we can conclude that 2003 and 2018 have the highest and the lowest sum burned area values (respectively). We can also see some hotspot regions in the centre of Portugal in August.\n\n(code-section-3)=\n### 3. Define the unit of analysis: Nomenclature of territorial units for statistics, level 2 (NUTS 2)\n\nThe [NUTS](https://ec.europa.eu/eurostat/web/nuts) are a hierarchical system divided into 3 levels. NUTS 1 correspond to major socio-economic regions, NUTS 2 correspond to basic regions for the application of regional policies, and NUTS 3 correspond to small regions for specific diagnoses. Additionally, a NUTS 0 level, usually co-incident with national boundaries is also available. The NUTS legislation is periodically amended; therefore multiple years are available for download.\n\nIn this user question, it is used the most updated version of the NUTS corresponding to version of 2021. In this study case, NUTS 2 will be used, providing the information regarding the main regions/parcels of the Iberian Peninsula.\n\nNote: The sum of burned area and the climatology are calculated per each NUTS 2 region. Generated maps, bar charts and plots will take some minutes to process all the months, years and regions.\n\nMerging BA with NUTS 2 geodataframe\n\n(code-section-4)=\n### 4. Annual Summaries and Summer Climatology Analysis of the Burned Area (JJA of 2001-2019)\n\nThe step below produces the Annual Summaries of BA by representing the Bar Charts and the Maps of the Sum of the Burned Area per NUTS 2, per month/year. In this case, we are analysing the Summer (JJA) months, which is when wildfires tend to be more frequent, in the Iberian Peninsula Region.\n\nIn addition, the Summer Climatology for BA aims to provide an overview of fire activity over a long-term period, such as multiple years or decades.\n\nAccording to the European Environment Agency (EEA (2021) Europe's changing climate hazards — an index-based interactive EEA report, Report no. 15/2021, EEA. DOI: 10.2800/458052) methodology, the following steps are conducted:\n- the sum of Burned Area is calculated for each year within the defined period; \n- the resulting timeseries is average across the number of years.\n\nThis average BA represents the typical magnitude of burned areas during the summer period (June, July and August, from 2001 to 2019).\n\nThe Annual Summaries and Summer Climatology are represented in several ways, such as maps, bar charts and timeseries plots.\n\n(code-section-4.1)=\n#### 4.1. Annual Summaries - Maps and Bar Charts of total Burned Area for the summer season, per NUTS regions per year\n\nPloting maps and bar charts of annual BA by NUTS 2 region, considering the Summer season"} {"chunk_id": "satellite_satellite-fire-burned-area_climate-and-weather-extremes_q01__6b4fd9b48545", "report_id": "satellite_satellite-fire-burned-area_climate-and-weather-extremes_q01", "dataset_id": "satellite-fire-burned-area", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q01", "aspect_base": "climate-and-weather-extremes", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite fire burned area completeness for seasonal climatology maps > Quality assessment statement", "title": "Satellite fire burned area completeness for seasonal climatology maps", "chunk_index": 3, "token_count": 1095, "text_raw": ". Annual Summaries - Maps and Bar Charts of total Burned Area for the summer season, per NUTS regions per year\n\nPloting maps and bar charts of annual BA by NUTS 2 region, considering the Summer season\n\nFrom the above set of plots, we can already distinguish certain years and areas more prone to extreme wildfire events. To better analyse its temporal pattern, let's plot the annual sum as a time series plot - this plot is equivalent to the first one in this notebook, but now we have added the NUTS 2 regions to facilitate our understanding of which regions have greater BA and in which years.\n\n(code-section-4.2)=\n#### 4.2. Annual Summaries - Timeseries of total Burned Area for the summer season, per NUTS regions per year\n\nPloting annual timeseries of BA by NUTS 2 regions, considering the Summer season\n\nHaving looked into this time series, we can clearly see that the highest BA values are reached in certain regions, such as 'Centro (PT)', 'Alentejo' or 'Norte', in Portugal, or 'Andalucía' and 'Extremadura', in Spain. However, we can also see that there is no indication of an increase in BA, in recent years. Now let's plot the time series per region to see this in more detail. We will also use a colour code for month to highlight which are characterized by higher BA values.\n\nList of regions with higher BA\nPloting annual timeseries of BA by NUTS 2 regions and months, considering the Summer season\nprint(region)\n\nFrom these plots, we can now confirm that the month of August is when more BA is detected in the product, which agrees with the known climate seasonality of the region. We will now compute the climatology per season to see typical spatial patterns of the fire BA, together with a bar chart with climate BA values per NUTS 2.\n\n(code-section-4.3)=\n#### 4.3. Seasonal Climatology (2001-2019) - Map and Bar Chart of the average total Burned Area for the summer season, per NUTS regions\n\nWith this final plot we can see what the average BA is, and the corresponding spatial pattern, to serve as a baseline for wildfire prevention and monitoring.\n\n(code-section-5)=\n### 5. Main takeaways\n\n- In particular, during the summer of 2003 the highest value of burned area is observed in Portugal (more specifically, in Centro PT and Alentejo regions), accounting for more than 3500 km², which agrees with known extreme values of burned area occurring in that year (as per the Portuguese national statistics on burned area, available at www.pordata.pt). This results also are consisten with, [Lourenço and Luciano (2018)](https://doi.org/10.4000/MEDITERRANEE.9958), which recorded a burned area for Portugal with values similar with this analysis (more that 4000 km² of burned area); hence, the results presented in this study show that the dataset depicts, in a very clear way, the events described.\n\n- Considering the Annual Summary analysis for the NUTS 2 regions, it can be concluded that Centro PT, Alentejo, Norte, Andalucía and Galicia are the regions that registered the highest values of burned area (this statement it is supported by the study of [Grünig et. al., 2023](https://doi.org/10.1111/GCB.16547), which found a largest average maximum fire size recorded for Portugal and highest average maximum burn severity for Spain), during the Summer months (JJA), from 2001 to 2019 - 2233 km²(2003, in Centro PT, Portugal), 1326 km² (2003, in Alentejo, Portugal), 1301 km² (2005, in Norte, Portugal), 840 km² (2004, in Andalucía, Spain), and 1117 km² (2006, in Galicia, Spain).\n\n- By contrast, several regions show consistently absence or very low values of burned area for the same time period - e.g., La Rioja only had a total of 9 km² of burned area occuring in only 4 out of the 19 years; equivalent pattern can be described for the Region of Murcia, which only had a total of 11 km² of burned area occurring in only 5 out of the 19 years.\n\n## ℹ️ If you want to know more\n\n### Key resources\n\nSome key resources and further reading were linked throughout this assessment.\n\nThe CDS catalogue entry for the data used were:\n\n* Fire burned area from 2001 to present derived from satellite observations:\n\nhttps://cds.climate.copernicus.eu/datasets/satellite-fire-burned-area?tab=overview", "text_with_prefix": "EQC Quality Assessment: \"Satellite fire burned area completeness for seasonal climatology maps\"\nDataset: satellite-fire-burned-area [CDS]\nAspect: climate-and-weather-extremes_q01 | Category: Satellite_ECVs\nSection: Satellite fire burned area completeness for seasonal climatology maps > Quality assessment statement\n---\n. Annual Summaries - Maps and Bar Charts of total Burned Area for the summer season, per NUTS regions per year\n\nPloting maps and bar charts of annual BA by NUTS 2 region, considering the Summer season\n\nFrom the above set of plots, we can already distinguish certain years and areas more prone to extreme wildfire events. To better analyse its temporal pattern, let's plot the annual sum as a time series plot - this plot is equivalent to the first one in this notebook, but now we have added the NUTS 2 regions to facilitate our understanding of which regions have greater BA and in which years.\n\n(code-section-4.2)=\n#### 4.2. Annual Summaries - Timeseries of total Burned Area for the summer season, per NUTS regions per year\n\nPloting annual timeseries of BA by NUTS 2 regions, considering the Summer season\n\nHaving looked into this time series, we can clearly see that the highest BA values are reached in certain regions, such as 'Centro (PT)', 'Alentejo' or 'Norte', in Portugal, or 'Andalucía' and 'Extremadura', in Spain. However, we can also see that there is no indication of an increase in BA, in recent years. Now let's plot the time series per region to see this in more detail. We will also use a colour code for month to highlight which are characterized by higher BA values.\n\nList of regions with higher BA\nPloting annual timeseries of BA by NUTS 2 regions and months, considering the Summer season\nprint(region)\n\nFrom these plots, we can now confirm that the month of August is when more BA is detected in the product, which agrees with the known climate seasonality of the region. We will now compute the climatology per season to see typical spatial patterns of the fire BA, together with a bar chart with climate BA values per NUTS 2.\n\n(code-section-4.3)=\n#### 4.3. Seasonal Climatology (2001-2019) - Map and Bar Chart of the average total Burned Area for the summer season, per NUTS regions\n\nWith this final plot we can see what the average BA is, and the corresponding spatial pattern, to serve as a baseline for wildfire prevention and monitoring.\n\n(code-section-5)=\n### 5. Main takeaways\n\n- In particular, during the summer of 2003 the highest value of burned area is observed in Portugal (more specifically, in Centro PT and Alentejo regions), accounting for more than 3500 km², which agrees with known extreme values of burned area occurring in that year (as per the Portuguese national statistics on burned area, available at www.pordata.pt). This results also are consisten with, [Lourenço and Luciano (2018)](https://doi.org/10.4000/MEDITERRANEE.9958), which recorded a burned area for Portugal with values similar with this analysis (more that 4000 km² of burned area); hence, the results presented in this study show that the dataset depicts, in a very clear way, the events described.\n\n- Considering the Annual Summary analysis for the NUTS 2 regions, it can be concluded that Centro PT, Alentejo, Norte, Andalucía and Galicia are the regions that registered the highest values of burned area (this statement it is supported by the study of [Grünig et. al., 2023](https://doi.org/10.1111/GCB.16547), which found a largest average maximum fire size recorded for Portugal and highest average maximum burn severity for Spain), during the Summer months (JJA), from 2001 to 2019 - 2233 km²(2003, in Centro PT, Portugal), 1326 km² (2003, in Alentejo, Portugal), 1301 km² (2005, in Norte, Portugal), 840 km² (2004, in Andalucía, Spain), and 1117 km² (2006, in Galicia, Spain).\n\n- By contrast, several regions show consistently absence or very low values of burned area for the same time period - e.g., La Rioja only had a total of 9 km² of burned area occuring in only 4 out of the 19 years; equivalent pattern can be described for the Region of Murcia, which only had a total of 11 km² of burned area occurring in only 5 out of the 19 years.\n\n## ℹ️ If you want to know more\n\n### Key resources\n\nSome key resources and further reading were linked throughout this assessment.\n\nThe CDS catalogue entry for the data used were:\n\n* Fire burned area from 2001 to present derived from satellite observations:\n\nhttps://cds.climate.copernicus.eu/datasets/satellite-fire-burned-area?tab=overview"} {"chunk_id": "satellite_satellite-fire-burned-area_climate-and-weather-extremes_q01__832f0628408e", "report_id": "satellite_satellite-fire-burned-area_climate-and-weather-extremes_q01", "dataset_id": "satellite-fire-burned-area", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q01", "aspect_base": "climate-and-weather-extremes", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite fire burned area completeness for seasonal climatology maps > Quality assessment statement", "title": "Satellite fire burned area completeness for seasonal climatology maps", "chunk_index": 4, "token_count": 548, "text_raw": "catalogue entry for the data used were:\n\n* Fire burned area from 2001 to present derived from satellite observations:\n\nhttps://cds.climate.copernicus.eu/datasets/satellite-fire-burned-area?tab=overview\n\nEEA FIRE Climatology Methodology:\n* https://www.eea.europa.eu/publications/europes-changing-climate-hazards-1\n\nFIRE dataset Product User Guide (PUG):\n* https://dast.copernicus-climate.eu/documents/satellite-fire-burned-area/D3.3.14-v1.0_PUGS_CDR_BA-FireCCI_MODIS_v5.1cds_PRODUCTS_v1.0.1.pdf\n\nEurostat NUTS (Nomenclature of territorial units for statistics) regions and definition link:\n* https://ec.europa.eu/eurostat/web/nuts\n\nCode library used\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)\n\n### References\n\n[[1]](https://doi.org/10.1007/s00704-023-04427-y) Bento, Virgílio A., Ana Russo, Inês Vieira, and Célia M. Gouveia. 2023. “Identification of Forest Vulnerability to Droughts in the Iberian Peninsula.” Theoretical and Applied Climatology 152 (1–2): 559–79.\n\n[[2]](https://doi.org/10.4000/MEDITERRANEE.9958) Lourenço, and Luciano. 2018. “Forest Fires in Continental PortugalResult of Profound Alterations in Society and Territorial Consequences”. Méditerranée (Online): 130 | 2018.\n\n[[3]](https://doi.org/10.1111/GCB.16547) Grünig, Marc, Rupert Seidl, and Cornelius Senf. 2023. “Increasing Aridity Causes Larger and More Severe Forest Fires across Europe.” Global Change Biology 29 (6): 1648–59.", "text_with_prefix": "EQC Quality Assessment: \"Satellite fire burned area completeness for seasonal climatology maps\"\nDataset: satellite-fire-burned-area [CDS]\nAspect: climate-and-weather-extremes_q01 | Category: Satellite_ECVs\nSection: Satellite fire burned area completeness for seasonal climatology maps > Quality assessment statement\n---\ncatalogue entry for the data used were:\n\n* Fire burned area from 2001 to present derived from satellite observations:\n\nhttps://cds.climate.copernicus.eu/datasets/satellite-fire-burned-area?tab=overview\n\nEEA FIRE Climatology Methodology:\n* https://www.eea.europa.eu/publications/europes-changing-climate-hazards-1\n\nFIRE dataset Product User Guide (PUG):\n* https://dast.copernicus-climate.eu/documents/satellite-fire-burned-area/D3.3.14-v1.0_PUGS_CDR_BA-FireCCI_MODIS_v5.1cds_PRODUCTS_v1.0.1.pdf\n\nEurostat NUTS (Nomenclature of territorial units for statistics) regions and definition link:\n* https://ec.europa.eu/eurostat/web/nuts\n\nCode library used\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)\n\n### References\n\n[[1]](https://doi.org/10.1007/s00704-023-04427-y) Bento, Virgílio A., Ana Russo, Inês Vieira, and Célia M. Gouveia. 2023. “Identification of Forest Vulnerability to Droughts in the Iberian Peninsula.” Theoretical and Applied Climatology 152 (1–2): 559–79.\n\n[[2]](https://doi.org/10.4000/MEDITERRANEE.9958) Lourenço, and Luciano. 2018. “Forest Fires in Continental PortugalResult of Profound Alterations in Society and Territorial Consequences”. Méditerranée (Online): 130 | 2018.\n\n[[3]](https://doi.org/10.1111/GCB.16547) Grünig, Marc, Rupert Seidl, and Cornelius Senf. 2023. “Increasing Aridity Causes Larger and More Severe Forest Fires across Europe.” Global Change Biology 29 (6): 1648–59."} {"chunk_id": "satellite_satellite-fire-burned-area_trend-assessment_q02__b7d182ead8f3", "report_id": "satellite_satellite-fire-burned-area_trend-assessment_q02", "dataset_id": "satellite-fire-burned-area", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q02", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite fire burned area trends assessment for climate monitoring > Quality assessment question", "title": "Satellite fire burned area trends assessment for climate monitoring", "chunk_index": 0, "token_count": 194, "text_raw": "* **How well can we disclose what are the trends and quality of the total burned area over the Iberian Peninsula?**\n\nIn this Use Case we will access the Fire burned area from 2001 to present derived from satellite observations (henceforth, FIRE) data from the Climate Data Store (CDS) of the Copernicus Climate Change Service (C3S) and analyse the spatial patterns of the FIRE reliability over a given Area of Interest (AoI), considering the Nomenclature of Territorial Units for Statistics, at the level 2, which are used for decision monitoring and decision-making purposes (see section 3).", "text_with_prefix": "EQC Quality Assessment: \"Satellite fire burned area trends assessment for climate monitoring\"\nDataset: satellite-fire-burned-area [CDS]\nAspect: trend-assessment_q02 | Category: Satellite_ECVs\nSection: Satellite fire burned area trends assessment for climate monitoring > Quality assessment question\n---\n* **How well can we disclose what are the trends and quality of the total burned area over the Iberian Peninsula?**\n\nIn this Use Case we will access the Fire burned area from 2001 to present derived from satellite observations (henceforth, FIRE) data from the Climate Data Store (CDS) of the Copernicus Climate Change Service (C3S) and analyse the spatial patterns of the FIRE reliability over a given Area of Interest (AoI), considering the Nomenclature of Territorial Units for Statistics, at the level 2, which are used for decision monitoring and decision-making purposes (see section 3)."} {"chunk_id": "satellite_satellite-fire-burned-area_trend-assessment_q02__8353c9d1e828", "report_id": "satellite_satellite-fire-burned-area_trend-assessment_q02", "dataset_id": "satellite-fire-burned-area", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q02", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite fire burned area trends assessment for climate monitoring > Quality assessment statement", "title": "Satellite fire burned area trends assessment for climate monitoring", "chunk_index": 1, "token_count": 1049, "text_raw": "These are the key outcomes of this assessment\n\n* The dataset shows strong spatial completeness by covering key NUTS 2 regions and temporal completeness over an 18-year period. However, this study focuses only on the summer; hence, significant fire events outside of this season are not accounted for.\n\n* The spatial resolution is sufficient to detect temporal and regional variations. For example, the dataset accurately reflects the burned area values observed during significant fire events, such as the summer of 2003 in Portugal. The burned area recorded (more than 200 km² in Centro PT region) is consistent with the Portuguese national statistics and findings by [[2]](https://doi.org/10.4000/MEDITERRANEE.9958) and [[10]](https://doi.org/10.3390/cli12090143).\n\n* Furthermore, the trend analysis indicates the absence of significant trends across the studied areas. These results are consistent those described in the literature, by [[8]](https://doi.org/10.1038/s41598-019-50281-2), where the authors showed non-significant changes of the Burned Area in Portugal (June to October, from 1980 to 2017).\n\n* These results are also consistent across different product versions, since a good global agreement has been described between FireCCI51, MCD64A1 c6 and GFED4 (all based on MODIS information), particularly starting in 2003 when both MODIS Terra and Aqua data were available. ([[9]](https://climate.esa.int/en/projects/fire/key-documents/); page 45)\n\n## 📋 Methodology\n\n**[](code-section-1)**\n\n**[](code-section-2)**\n\n* **[](code-section-2.1)**\n\n* **[](code-section-2.2)**\n\n**[](code-section-3)**\n\n## 📈 Analysis and results\n\n(code-section-1)=\n### 1. Data overview, download and NUTS regions definition\n\n#### Import all the libraries/packages\n\nWe will be working with data in NetCDF format. To best handle this data we will use libraries for working with multidimensional arrays, in particular Xarray. We will also need libraries for plotting and viewing data, in this case we will use Matplotlib and Cartopy.\n\n#### Data overview\n\nTo search for data, visit the CDS website: http://cds.climate.copernicus.eu Here you can search for 'Satellite observations' data using the search bar. The data we need for this tutorial is the Fire burned area from 2001 to present derived from satellite observations. This data provides global (at grid scale) and continents (at pixel scale) information of total burned area (BA), following the Global Climate Observing System (GCOS) convention. The dataset is provided with a high spatial resolution, through the analysis of reflectance changes from medium resolution sensors (Terra MODIS, Sentinel-3 OLCI) supported by thermal information.\n\nThe BA is identified by the date of the first detection of the burned signal in the case of the pixel product, and by the total BA per grid cell in the case of the grid product. Information such as land cover class, confidence level or standard error are also provided.\n\nThe BA products can be on pixel scale and grid scale, based on resolution and coverage. At the grid scale, data is provided globally with a horizontal resolution of 0.25° latitude by 0.25° longitude, offering a coarser spatial resolution that is suitable for broad, global analyses.\n\nThis grid product data is available every 15 days in version 5.0cds and monthly in versions 5.1.1cds and 1.1.\n\nConversely, at the pixel scale, data is more detailed, covering continents with a finer horizontal resolution of 250m or 300m, and is aggregated on a monthly basis. This pixel product provides higher spatial resolution suitable for more localized, detailed studies.\n\nDifferent versions are available, offering the first global BA time series at 250 m resolution (FireCCI v5.0cds and FireCCI v5.1cds developed as part of the Fire ECV Climate Change Initiative Project (Fire CCI)). This algorithm was adapted to Sentinel-3 OLCI data to create the C3S v1.1 burned area product, extending the BA database to the present.\n\nHaving selected the correct dataset, we now need to specify what product type, variables, temporal and geographic coverage we are interested in. In this Use case, we will retrieve information of the grid ESA-CCI version. These can all be selected in the `Download data` tab from the CDS. In this tab a form appears in which we will select the following parameters to download, for example:", "text_with_prefix": "EQC Quality Assessment: \"Satellite fire burned area trends assessment for climate monitoring\"\nDataset: satellite-fire-burned-area [CDS]\nAspect: trend-assessment_q02 | Category: Satellite_ECVs\nSection: Satellite fire burned area trends assessment for climate monitoring > Quality assessment statement\n---\nThese are the key outcomes of this assessment\n\n* The dataset shows strong spatial completeness by covering key NUTS 2 regions and temporal completeness over an 18-year period. However, this study focuses only on the summer; hence, significant fire events outside of this season are not accounted for.\n\n* The spatial resolution is sufficient to detect temporal and regional variations. For example, the dataset accurately reflects the burned area values observed during significant fire events, such as the summer of 2003 in Portugal. The burned area recorded (more than 200 km² in Centro PT region) is consistent with the Portuguese national statistics and findings by [[2]](https://doi.org/10.4000/MEDITERRANEE.9958) and [[10]](https://doi.org/10.3390/cli12090143).\n\n* Furthermore, the trend analysis indicates the absence of significant trends across the studied areas. These results are consistent those described in the literature, by [[8]](https://doi.org/10.1038/s41598-019-50281-2), where the authors showed non-significant changes of the Burned Area in Portugal (June to October, from 1980 to 2017).\n\n* These results are also consistent across different product versions, since a good global agreement has been described between FireCCI51, MCD64A1 c6 and GFED4 (all based on MODIS information), particularly starting in 2003 when both MODIS Terra and Aqua data were available. ([[9]](https://climate.esa.int/en/projects/fire/key-documents/); page 45)\n\n## 📋 Methodology\n\n**[](code-section-1)**\n\n**[](code-section-2)**\n\n* **[](code-section-2.1)**\n\n* **[](code-section-2.2)**\n\n**[](code-section-3)**\n\n## 📈 Analysis and results\n\n(code-section-1)=\n### 1. Data overview, download and NUTS regions definition\n\n#### Import all the libraries/packages\n\nWe will be working with data in NetCDF format. To best handle this data we will use libraries for working with multidimensional arrays, in particular Xarray. We will also need libraries for plotting and viewing data, in this case we will use Matplotlib and Cartopy.\n\n#### Data overview\n\nTo search for data, visit the CDS website: http://cds.climate.copernicus.eu Here you can search for 'Satellite observations' data using the search bar. The data we need for this tutorial is the Fire burned area from 2001 to present derived from satellite observations. This data provides global (at grid scale) and continents (at pixel scale) information of total burned area (BA), following the Global Climate Observing System (GCOS) convention. The dataset is provided with a high spatial resolution, through the analysis of reflectance changes from medium resolution sensors (Terra MODIS, Sentinel-3 OLCI) supported by thermal information.\n\nThe BA is identified by the date of the first detection of the burned signal in the case of the pixel product, and by the total BA per grid cell in the case of the grid product. Information such as land cover class, confidence level or standard error are also provided.\n\nThe BA products can be on pixel scale and grid scale, based on resolution and coverage. At the grid scale, data is provided globally with a horizontal resolution of 0.25° latitude by 0.25° longitude, offering a coarser spatial resolution that is suitable for broad, global analyses.\n\nThis grid product data is available every 15 days in version 5.0cds and monthly in versions 5.1.1cds and 1.1.\n\nConversely, at the pixel scale, data is more detailed, covering continents with a finer horizontal resolution of 250m or 300m, and is aggregated on a monthly basis. This pixel product provides higher spatial resolution suitable for more localized, detailed studies.\n\nDifferent versions are available, offering the first global BA time series at 250 m resolution (FireCCI v5.0cds and FireCCI v5.1cds developed as part of the Fire ECV Climate Change Initiative Project (Fire CCI)). This algorithm was adapted to Sentinel-3 OLCI data to create the C3S v1.1 burned area product, extending the BA database to the present.\n\nHaving selected the correct dataset, we now need to specify what product type, variables, temporal and geographic coverage we are interested in. In this Use case, we will retrieve information of the grid ESA-CCI version. These can all be selected in the `Download data` tab from the CDS. In this tab a form appears in which we will select the following parameters to download, for example:"} {"chunk_id": "satellite_satellite-fire-burned-area_trend-assessment_q02__75818ee10d2d", "report_id": "satellite_satellite-fire-burned-area_trend-assessment_q02", "dataset_id": "satellite-fire-burned-area", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q02", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite fire burned area trends assessment for climate monitoring > Quality assessment statement", "title": "Satellite fire burned area trends assessment for climate monitoring", "chunk_index": 2, "token_count": 1087, "text_raw": "we will retrieve information of the grid ESA-CCI version. These can all be selected in the `Download data` tab from the CDS. In this tab a form appears in which we will select the following parameters to download, for example:\n\n- Origin: ESA-CCI\n- Sensor: MODIS\n- Variable: Grid\n- Version: 5.1.1cds\n- Region: all\n- Year: 2001 to 2019\n- Month: all\n- Nominal day: 01\n- Format: Zip file (.zip)\n\nAt the end of the download form, select `Show API request`. This will reveal a block of code, which you can simply copy and paste into a cell of your Jupyter Notebook-\n\nHaving copied the API request to a Jupyter Notebook cell, running it will retrieve and download the data you requested into your local directory. However, before you run it, the terms and conditions of this particular dataset need to have been accepted directly at the CDS website. The option to view and accept these conditions is given at the end of the download form, just above the `Show API request` option. In addition, it is also useful to define the time period and AoI parameters and edit the request accordingly, as exemplified in the cells below.\n\nRegion of interest\nShapefile with regions\n\nDefine request\n\nDownload and regionalize\nReindex using year/month (shift months + 1)\nConvert units from m2 to km2\n\n```text\n100%|██████████| 19/19 [00:02<00:00, 7.97it/s]\n```\n\n#### Define the unit of analysis: Nomenclature of territorial units for statistics, level 2 (NUTS 2)\n\nThe [NUTS](https://ec.europa.eu/eurostat/web/nuts) are a hierarchical system divided into 3 levels. NUTS 1 correspond to major socio-economic regions, NUTS 2 correspond to basic regions for the application of regional policies, and NUTS 3 correspond to small regions for specific diagnoses. Additionally, a NUTS 0 level, usually co-incident with national boundaries is also available. The NUTS legislation is periodically amended; therefore multiple years are available for download.\n\nIn this user question, NUTS 2 will be used, providing the information regarding the main regions/parcels of the Iberian Peninsula.\n\nNote: The sum of burned area and the climatology are calculated per each NUTS 2 region. Generated maps, bar charts and plots will take some minutes to process all the months, years and regions.\n\nMerging BA with NUTS 2 geodataframe\n\n(code-section-2)=\n### 2. Trend Analysis and Slope Calculation\n#### Mann-Kendall Test and Theil-Senn slope\n\nThe Theil-Sen (1968) trend analysis, a non-parametric method, calculates the median slope of all data point pairs. It's a robust technique for trend identification, especially in the presence of outliers and noise. The Theil-Sen estimator offers several advantages, such as simplicity, resistance to extreme values, and statistical validity. This makes it a preferred choice for trend analysis in various applications \nThe Sen slope estimator is found to be a powerful tool to develop the linear relationships. Sen’s slope has the advantage over the slope of regression, in the sense that gross data series errors and outliers do not have much effect.\n\nThe Mann-Kendall test (Kendall, 1975; Mann, 1945) is a non-parametric test that determines whether there is a statistically significant upward or downward trend (monotonic trend) over time. Kendall's Tau (τ) is a measure of the strength and direction of monotonic trends in time series data. It is often used in environmental studies, hydrology, climatology, and various other fields to assess trends in data where the relationship may not be linear and where non-parametric methods are preferred and is often referred to as Kendall's Tau or Kendall's Rank Correlation Coefficient.\n\nThe τ value can be either positive or negative. It quantifies the strength and direction of the trend in the data.\n\n* A positive tau (τ > 0) indicates an increasing or upward trend.\n* A negative tau (τ < 0) indicates a decreasing or downward trend.\n* A tau value close to zero (|τ| ≈ 0) suggests no significant trend or a random pattern in the data.\n\nIn addition to τ value, the Mann-Kendall test provides a p-value that indicates whether the observed trend is statistically significant. A low p-value (typically less than a chosen significance level, e.g., 0.05) suggests that the trend is statistically significant.\n\nThese methods were therefore adopted below to estimate trends and their statistical significance.\n\n(code-section-2.1)=\n### 2.1. Annual Trend Analysis and Slope Calculation for specific hotspot regions", "text_with_prefix": "EQC Quality Assessment: \"Satellite fire burned area trends assessment for climate monitoring\"\nDataset: satellite-fire-burned-area [CDS]\nAspect: trend-assessment_q02 | Category: Satellite_ECVs\nSection: Satellite fire burned area trends assessment for climate monitoring > Quality assessment statement\n---\nwe will retrieve information of the grid ESA-CCI version. These can all be selected in the `Download data` tab from the CDS. In this tab a form appears in which we will select the following parameters to download, for example:\n\n- Origin: ESA-CCI\n- Sensor: MODIS\n- Variable: Grid\n- Version: 5.1.1cds\n- Region: all\n- Year: 2001 to 2019\n- Month: all\n- Nominal day: 01\n- Format: Zip file (.zip)\n\nAt the end of the download form, select `Show API request`. This will reveal a block of code, which you can simply copy and paste into a cell of your Jupyter Notebook-\n\nHaving copied the API request to a Jupyter Notebook cell, running it will retrieve and download the data you requested into your local directory. However, before you run it, the terms and conditions of this particular dataset need to have been accepted directly at the CDS website. The option to view and accept these conditions is given at the end of the download form, just above the `Show API request` option. In addition, it is also useful to define the time period and AoI parameters and edit the request accordingly, as exemplified in the cells below.\n\nRegion of interest\nShapefile with regions\n\nDefine request\n\nDownload and regionalize\nReindex using year/month (shift months + 1)\nConvert units from m2 to km2\n\n```text\n100%|██████████| 19/19 [00:02<00:00, 7.97it/s]\n```\n\n#### Define the unit of analysis: Nomenclature of territorial units for statistics, level 2 (NUTS 2)\n\nThe [NUTS](https://ec.europa.eu/eurostat/web/nuts) are a hierarchical system divided into 3 levels. NUTS 1 correspond to major socio-economic regions, NUTS 2 correspond to basic regions for the application of regional policies, and NUTS 3 correspond to small regions for specific diagnoses. Additionally, a NUTS 0 level, usually co-incident with national boundaries is also available. The NUTS legislation is periodically amended; therefore multiple years are available for download.\n\nIn this user question, NUTS 2 will be used, providing the information regarding the main regions/parcels of the Iberian Peninsula.\n\nNote: The sum of burned area and the climatology are calculated per each NUTS 2 region. Generated maps, bar charts and plots will take some minutes to process all the months, years and regions.\n\nMerging BA with NUTS 2 geodataframe\n\n(code-section-2)=\n### 2. Trend Analysis and Slope Calculation\n#### Mann-Kendall Test and Theil-Senn slope\n\nThe Theil-Sen (1968) trend analysis, a non-parametric method, calculates the median slope of all data point pairs. It's a robust technique for trend identification, especially in the presence of outliers and noise. The Theil-Sen estimator offers several advantages, such as simplicity, resistance to extreme values, and statistical validity. This makes it a preferred choice for trend analysis in various applications \nThe Sen slope estimator is found to be a powerful tool to develop the linear relationships. Sen’s slope has the advantage over the slope of regression, in the sense that gross data series errors and outliers do not have much effect.\n\nThe Mann-Kendall test (Kendall, 1975; Mann, 1945) is a non-parametric test that determines whether there is a statistically significant upward or downward trend (monotonic trend) over time. Kendall's Tau (τ) is a measure of the strength and direction of monotonic trends in time series data. It is often used in environmental studies, hydrology, climatology, and various other fields to assess trends in data where the relationship may not be linear and where non-parametric methods are preferred and is often referred to as Kendall's Tau or Kendall's Rank Correlation Coefficient.\n\nThe τ value can be either positive or negative. It quantifies the strength and direction of the trend in the data.\n\n* A positive tau (τ > 0) indicates an increasing or upward trend.\n* A negative tau (τ < 0) indicates a decreasing or downward trend.\n* A tau value close to zero (|τ| ≈ 0) suggests no significant trend or a random pattern in the data.\n\nIn addition to τ value, the Mann-Kendall test provides a p-value that indicates whether the observed trend is statistically significant. A low p-value (typically less than a chosen significance level, e.g., 0.05) suggests that the trend is statistically significant.\n\nThese methods were therefore adopted below to estimate trends and their statistical significance.\n\n(code-section-2.1)=\n### 2.1. Annual Trend Analysis and Slope Calculation for specific hotspot regions"} {"chunk_id": "satellite_satellite-fire-burned-area_trend-assessment_q02__0b3203145d45", "report_id": "satellite_satellite-fire-burned-area_trend-assessment_q02", "dataset_id": "satellite-fire-burned-area", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q02", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite fire burned area trends assessment for climate monitoring > Quality assessment statement", "title": "Satellite fire burned area trends assessment for climate monitoring", "chunk_index": 3, "token_count": 1097, "text_raw": ") suggests that the trend is statistically significant.\n\nThese methods were therefore adopted below to estimate trends and their statistical significance.\n\n(code-section-2.1)=\n### 2.1. Annual Trend Analysis and Slope Calculation for specific hotspot regions\n\nTo illustrate the burned area trends, time series of the total burned area over four regions are shown (Galicia (ES), Andalucía (ES), Norte (PT) and Centro (PT)). For each time series, the Theil-Sen trend line is plotted and the values of its slope and intercept are provided, together with τ and the p-value, both returned by the Mann-Kendall test\n\nComputation and definition of Mean-Kendall and Theil-Sen Statistics\n\nFunction for ploting trends\n\nThe results show no significant trends, and a very high interannual variability. In general, these results largely agree with those of [[6]](https://doi.org/10.3390/FIRE4040074), although they used different periods, performed a global analysis and used a more sophisticated trend analysis scheme. They showed that the CCI datasets exhibit much less significant trends when compared to longer term datasets (such as those based in AVHRR observations), mainly due to the short time span of the timeseries.\n\n(code-section-2.2)=\n### 2.2. Yearly Trend Maps by NUTS regions.\n\nNext, a regional view of these results is provided in map form, where the trends and their significance were calculated for each NUTS region.\n\nCreate a proxy artist for the legend handle\nTurn off the axis\nShow the legend with the proxy artist\n\nRegarding this map, it can be stated that the same conclusions apply to all NUTS regions over the Iberian Peninsula, as none of them exhibits significant burned area trends over the considered period, likely because the trends may not be robust due to the short time span of the time series. As discussed in sections 3.2.1 and 3.2.2 of [[6]](https://doi.org/10.3390/FIRE4040074), longer time series are crucial for detecting significant trends in burned area, as shorter periods are often subject to greater variability and noise.\n\nThis aligns with findings from [[5]](https://doi.org/10.1088/2515-7620/AB25D2), which highlight that satellite datasets with limited temporal coverage can lead to inconsistencies and non-significant trends due to year-to-year fluctuations and sensor changes. Therefore, to obtain reliable and significant trends, it is recommended to utilize longer datasets that span several decades.\n\n(code-section-3)=\n### 3. Main takeaways\n\n- The trend analysis across the studied NUTS 2 regions (Galicia, Andalucía, Norte, and Centro) for the summer months (June, July, August) from 2001 to 2019 revealed no significant trends in the burned area. This lack of significance suggests that the observed variations in burned area are largely due to interannual variability rather than a consistent temporal trend.\n\n- The results align with findings from previous studies (e.g., [[6]](https://doi.org/10.3390/FIRE4040074)) that emphasize the importance of longer time series for trend detection. The short 18-year span of the dataset used in this analysis, particularly focused on the summer months, likely contributes to the absence of significant trends. Longer datasets would reduce the influence of year-to-year variability and provide more robust trend analysis.\n\n- The absence of significant trends in the Iberian Peninsula's burned area is consistent with other research, such as [[8]](https://doi.org/10.1038/s41598-019-50281-2), which also found non-significant changes in burned areas in Portugal over a similar period (1980-2017). This reinforces the notion that short-term datasets are insufficient for capturing long-term trends in fire activity.\n\n- Despite the lack of significant trends, the analysis demonstrates the reliability of the satellite-based dataset in capturing regional variations in burned area, particularly during significant fire events (e.g., the 2003 fires in Portugal). However, the non-significant trends suggest that more comprehensive datasets with broader temporal coverage are necessary for definitive conclusions.\n\n- To achieve a more accurate and significant understanding of trends in burned areas, future studies should utilize longer datasets that span several decades. This would help mitigate the impact of short-term fluctuations and provide a clearer picture of fire activity trends over time.\n\n## ℹ️ If you want to know more\n\n### Key resources\n\nSome key resources and further reading were linked throughout this assessment.\n\nThe CDS catalogue entry for the data used were:\n\n* Fire burned area from 2001 to present derived from satellite observations:\n\nhttps://cds.climate.copernicus.eu/datasets/satellite-fire-burned-area?tab=overview\n\nEEA FIRE Climatology Methodology:\n* https://www.eea.europa.eu/publications/europes-changing-climate-hazards-1", "text_with_prefix": "EQC Quality Assessment: \"Satellite fire burned area trends assessment for climate monitoring\"\nDataset: satellite-fire-burned-area [CDS]\nAspect: trend-assessment_q02 | Category: Satellite_ECVs\nSection: Satellite fire burned area trends assessment for climate monitoring > Quality assessment statement\n---\n) suggests that the trend is statistically significant.\n\nThese methods were therefore adopted below to estimate trends and their statistical significance.\n\n(code-section-2.1)=\n### 2.1. Annual Trend Analysis and Slope Calculation for specific hotspot regions\n\nTo illustrate the burned area trends, time series of the total burned area over four regions are shown (Galicia (ES), Andalucía (ES), Norte (PT) and Centro (PT)). For each time series, the Theil-Sen trend line is plotted and the values of its slope and intercept are provided, together with τ and the p-value, both returned by the Mann-Kendall test\n\nComputation and definition of Mean-Kendall and Theil-Sen Statistics\n\nFunction for ploting trends\n\nThe results show no significant trends, and a very high interannual variability. In general, these results largely agree with those of [[6]](https://doi.org/10.3390/FIRE4040074), although they used different periods, performed a global analysis and used a more sophisticated trend analysis scheme. They showed that the CCI datasets exhibit much less significant trends when compared to longer term datasets (such as those based in AVHRR observations), mainly due to the short time span of the timeseries.\n\n(code-section-2.2)=\n### 2.2. Yearly Trend Maps by NUTS regions.\n\nNext, a regional view of these results is provided in map form, where the trends and their significance were calculated for each NUTS region.\n\nCreate a proxy artist for the legend handle\nTurn off the axis\nShow the legend with the proxy artist\n\nRegarding this map, it can be stated that the same conclusions apply to all NUTS regions over the Iberian Peninsula, as none of them exhibits significant burned area trends over the considered period, likely because the trends may not be robust due to the short time span of the time series. As discussed in sections 3.2.1 and 3.2.2 of [[6]](https://doi.org/10.3390/FIRE4040074), longer time series are crucial for detecting significant trends in burned area, as shorter periods are often subject to greater variability and noise.\n\nThis aligns with findings from [[5]](https://doi.org/10.1088/2515-7620/AB25D2), which highlight that satellite datasets with limited temporal coverage can lead to inconsistencies and non-significant trends due to year-to-year fluctuations and sensor changes. Therefore, to obtain reliable and significant trends, it is recommended to utilize longer datasets that span several decades.\n\n(code-section-3)=\n### 3. Main takeaways\n\n- The trend analysis across the studied NUTS 2 regions (Galicia, Andalucía, Norte, and Centro) for the summer months (June, July, August) from 2001 to 2019 revealed no significant trends in the burned area. This lack of significance suggests that the observed variations in burned area are largely due to interannual variability rather than a consistent temporal trend.\n\n- The results align with findings from previous studies (e.g., [[6]](https://doi.org/10.3390/FIRE4040074)) that emphasize the importance of longer time series for trend detection. The short 18-year span of the dataset used in this analysis, particularly focused on the summer months, likely contributes to the absence of significant trends. Longer datasets would reduce the influence of year-to-year variability and provide more robust trend analysis.\n\n- The absence of significant trends in the Iberian Peninsula's burned area is consistent with other research, such as [[8]](https://doi.org/10.1038/s41598-019-50281-2), which also found non-significant changes in burned areas in Portugal over a similar period (1980-2017). This reinforces the notion that short-term datasets are insufficient for capturing long-term trends in fire activity.\n\n- Despite the lack of significant trends, the analysis demonstrates the reliability of the satellite-based dataset in capturing regional variations in burned area, particularly during significant fire events (e.g., the 2003 fires in Portugal). However, the non-significant trends suggest that more comprehensive datasets with broader temporal coverage are necessary for definitive conclusions.\n\n- To achieve a more accurate and significant understanding of trends in burned areas, future studies should utilize longer datasets that span several decades. This would help mitigate the impact of short-term fluctuations and provide a clearer picture of fire activity trends over time.\n\n## ℹ️ If you want to know more\n\n### Key resources\n\nSome key resources and further reading were linked throughout this assessment.\n\nThe CDS catalogue entry for the data used were:\n\n* Fire burned area from 2001 to present derived from satellite observations:\n\nhttps://cds.climate.copernicus.eu/datasets/satellite-fire-burned-area?tab=overview\n\nEEA FIRE Climatology Methodology:\n* https://www.eea.europa.eu/publications/europes-changing-climate-hazards-1"} {"chunk_id": "satellite_satellite-fire-burned-area_trend-assessment_q02__3c808355b410", "report_id": "satellite_satellite-fire-burned-area_trend-assessment_q02", "dataset_id": "satellite-fire-burned-area", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q02", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite fire burned area trends assessment for climate monitoring > Quality assessment statement", "title": "Satellite fire burned area trends assessment for climate monitoring", "chunk_index": 4, "token_count": 1111, "text_raw": "ernicus.eu/datasets/satellite-fire-burned-area?tab=overview\n\nEEA FIRE Climatology Methodology:\n* https://www.eea.europa.eu/publications/europes-changing-climate-hazards-1\n\nFIRE dataset Product User Guide (PUG):\n* https://dast.copernicus-climate.eu/documents/satellite-fire-burned-area/D3.3.14-v1.0_PUGS_CDR_BA-FireCCI_MODIS_v5.1cds_PRODUCTS_v1.0.1.pdf\n\nEurostat NUTS (Nomenclature of territorial units for statistics) regions and definition link:\n* https://ec.europa.eu/eurostat/web/nuts\n\nCode library used\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)\n\n### References\n\n[[1]](https://doi.org/10.1007/s00704-023-04427-y) Bento, Virgílio A., Ana Russo, Inês Vieira, and Célia M. Gouveia. 2023. “Identification of Forest Vulnerability to Droughts in the Iberian Peninsula.” Theoretical and Applied Climatology 152 (1–2): 559–79.\n\n[[2]](https://doi.org/10.4000/MEDITERRANEE.9958) Lourenço, and Luciano. 2018. “Forest Fires in Continental PortugalResult of Profound Alterations in Society and Territorial Consequences”. no. 130 (September).\n\n[[3]](https://doi.org/10.1111/GCB.16547) Grünig, Marc, Rupert Seidl, and Cornelius Senf. 2023. “Increasing Aridity Causes Larger and More Severe Forest Fires across Europe.” Global Change Biology 29 (6): 1648–59.\n\n[[4]](https://doi.org/10.26682/csjuod.2020.23.2.41) Aswad, Fawaz & Yousif, Ali & Ibrahim, Sayran. (2020). \"Trend Analysis Using Mann-Kendall and Sen’s Slope Estimator Test for Annual and Monthly Rainfall for Sinjar District, Iraq.\"\n\n[[5]](https://doi.org/10.1088/2515-7620/AB25D2) Forkel, Matthias, Wouter Dorigo, Gitta Lasslop, Emilio Chuvieco, Stijn Hantson, Angelika Heil, Irene Teubner, Kirsten Thonicke, and Sandy P. Harrison. 2019. “Recent Global and Regional Trends in Burned Area and Their Compensating Environmental Controls.” Environmental Research Communications 1 (5): 051005.\n\n[[6]](https://doi.org/10.3390/FIRE4040074) Otón, Gonzalo, José Miguel C. Pereira, João M.N. Silva, and Emilio Chuvieco. 2021. “Analysis of Trends in the FireCCI Global Long Term Burned Area Product (1982–2018).” Fire 2021, Vol. 4, Page 74 4 (4): 74.\n\n[[7]](https://doi.org/10.1007/s10668-020-00842-7) Rasul, Azad, Gaylan R.Faqe Ibrahim, Hasan M. Hameed, and Kevin Tansey. 2021. “A Trend of Increasing Burned Areas in Iraq from 2001 to 2019.” Environment, Development and Sustainability 23 (4): 5739–55.\n\n[[8]](https://doi.org/10.1038/s41598-019-50281-2) Turco, M., Jerez, S., Augusto, S. et al. Climate drivers of the 2017 devastating fires in Portugal. Sci Rep 9, 13886 (2019).\n\n[[9]](https://climate.esa.int/en/projects/fire/key-documents/) Stroppiana D., Sali M., Boschetti M., Busetto L., Ranghetti L., Franquesa M., Lizundia-Loiola J., Pettinari M.L. (2022) ESA CCI ECV Fire Disturbance: D4.1 Product Validation and Intercomparison Report, version 2.1. Available at: https://climate.esa.int/en/projects/fire/key-documents/\n\n[[10]](https://doi.org/10.3390/cli12090143) DaCamara, C.C. The Signature of Climate in Annual Burned Area in Portugal. Climate 2024, 12, 143. https://doi.org/10.3390/cli12090143", "text_with_prefix": "EQC Quality Assessment: \"Satellite fire burned area trends assessment for climate monitoring\"\nDataset: satellite-fire-burned-area [CDS]\nAspect: trend-assessment_q02 | Category: Satellite_ECVs\nSection: Satellite fire burned area trends assessment for climate monitoring > Quality assessment statement\n---\nernicus.eu/datasets/satellite-fire-burned-area?tab=overview\n\nEEA FIRE Climatology Methodology:\n* https://www.eea.europa.eu/publications/europes-changing-climate-hazards-1\n\nFIRE dataset Product User Guide (PUG):\n* https://dast.copernicus-climate.eu/documents/satellite-fire-burned-area/D3.3.14-v1.0_PUGS_CDR_BA-FireCCI_MODIS_v5.1cds_PRODUCTS_v1.0.1.pdf\n\nEurostat NUTS (Nomenclature of territorial units for statistics) regions and definition link:\n* https://ec.europa.eu/eurostat/web/nuts\n\nCode library used\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)\n\n### References\n\n[[1]](https://doi.org/10.1007/s00704-023-04427-y) Bento, Virgílio A., Ana Russo, Inês Vieira, and Célia M. Gouveia. 2023. “Identification of Forest Vulnerability to Droughts in the Iberian Peninsula.” Theoretical and Applied Climatology 152 (1–2): 559–79.\n\n[[2]](https://doi.org/10.4000/MEDITERRANEE.9958) Lourenço, and Luciano. 2018. “Forest Fires in Continental PortugalResult of Profound Alterations in Society and Territorial Consequences”. no. 130 (September).\n\n[[3]](https://doi.org/10.1111/GCB.16547) Grünig, Marc, Rupert Seidl, and Cornelius Senf. 2023. “Increasing Aridity Causes Larger and More Severe Forest Fires across Europe.” Global Change Biology 29 (6): 1648–59.\n\n[[4]](https://doi.org/10.26682/csjuod.2020.23.2.41) Aswad, Fawaz & Yousif, Ali & Ibrahim, Sayran. (2020). \"Trend Analysis Using Mann-Kendall and Sen’s Slope Estimator Test for Annual and Monthly Rainfall for Sinjar District, Iraq.\"\n\n[[5]](https://doi.org/10.1088/2515-7620/AB25D2) Forkel, Matthias, Wouter Dorigo, Gitta Lasslop, Emilio Chuvieco, Stijn Hantson, Angelika Heil, Irene Teubner, Kirsten Thonicke, and Sandy P. Harrison. 2019. “Recent Global and Regional Trends in Burned Area and Their Compensating Environmental Controls.” Environmental Research Communications 1 (5): 051005.\n\n[[6]](https://doi.org/10.3390/FIRE4040074) Otón, Gonzalo, José Miguel C. Pereira, João M.N. Silva, and Emilio Chuvieco. 2021. “Analysis of Trends in the FireCCI Global Long Term Burned Area Product (1982–2018).” Fire 2021, Vol. 4, Page 74 4 (4): 74.\n\n[[7]](https://doi.org/10.1007/s10668-020-00842-7) Rasul, Azad, Gaylan R.Faqe Ibrahim, Hasan M. Hameed, and Kevin Tansey. 2021. “A Trend of Increasing Burned Areas in Iraq from 2001 to 2019.” Environment, Development and Sustainability 23 (4): 5739–55.\n\n[[8]](https://doi.org/10.1038/s41598-019-50281-2) Turco, M., Jerez, S., Augusto, S. et al. Climate drivers of the 2017 devastating fires in Portugal. Sci Rep 9, 13886 (2019).\n\n[[9]](https://climate.esa.int/en/projects/fire/key-documents/) Stroppiana D., Sali M., Boschetti M., Busetto L., Ranghetti L., Franquesa M., Lizundia-Loiola J., Pettinari M.L. (2022) ESA CCI ECV Fire Disturbance: D4.1 Product Validation and Intercomparison Report, version 2.1. Available at: https://climate.esa.int/en/projects/fire/key-documents/\n\n[[10]](https://doi.org/10.3390/cli12090143) DaCamara, C.C. The Signature of Climate in Annual Burned Area in Portugal. Climate 2024, 12, 143. https://doi.org/10.3390/cli12090143"} {"chunk_id": "satellite_satellite-lai-fapar_extremes-detection_q02__f1f08d1f27ab", "report_id": "satellite_satellite-lai-fapar_extremes-detection_q02", "dataset_id": "satellite-lai-fapar", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extremes-detection_q02", "aspect_base": "extremes-detection", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Impact of Drought on the Leaf Area Index in Yunnan Province, China > Quality assessment question(s)", "title": "Impact of Drought on the Leaf Area Index in Yunnan Province, China", "chunk_index": 0, "token_count": 439, "text_raw": "* **Does version 3 of the C3S LAI dataset (derived from SPOT satellite imagery) provide sufficient temporal and spatial completeness to capture the impact of the 2009–2010 drought on vegetation in Yunnan Province?**\n* **How does the choice of relevant quality flags affect product spatial and temporal completeness?**\n\nThe Leaf Area Index (LAI) and Fraction of Absorbed Photosynthetically Active Radiation (fAPAR) dataset, provided by the Climate Data Store (C3S), offers 10-daily gridded observations from 1981 to the present. It includes effective LAI values derived from multiple satellite sensors across different product versions. Effective LAI is defined as half the total surface area of photosynthetically active plant elements per unit of horizontal ground area. It relates to true LAI through a canopy-dependent structure factor. As a key biophysical parameter, LAI is widely used to assess vegetation status, monitor ecosystem dynamics, and inform environmental and agricultural decision-making.\n\nThis quality assessment uses version 3 of the LAI product, selected because it covers the study period from 2007 to 2013. During this time, data were obtained from the System Pour l'Observation de la Terre (SPOT) satellite, Vegetation (VGT) sensor, with a spatial resolution of 1 km and a 10-day temporal resolution. The dataset is evaluated over Yunnan Province, China, with particular attention to the 2009-2010 drought.\n\nThe objectives of this assessment are: first, to assess whether the product's spatial and temporal completeness is sufficient to capture drought-related changes in LAI, which serves as a proxy for vegetation health; and second, to evaluate how different choices of quality flags influence the analysis.", "text_with_prefix": "EQC Quality Assessment: \"Impact of Drought on the Leaf Area Index in Yunnan Province, China\"\nDataset: satellite-lai-fapar [CDS]\nAspect: extremes-detection_q02 | Category: Satellite_ECVs\nSection: Impact of Drought on the Leaf Area Index in Yunnan Province, China > Quality assessment question(s)\n---\n* **Does version 3 of the C3S LAI dataset (derived from SPOT satellite imagery) provide sufficient temporal and spatial completeness to capture the impact of the 2009–2010 drought on vegetation in Yunnan Province?**\n* **How does the choice of relevant quality flags affect product spatial and temporal completeness?**\n\nThe Leaf Area Index (LAI) and Fraction of Absorbed Photosynthetically Active Radiation (fAPAR) dataset, provided by the Climate Data Store (C3S), offers 10-daily gridded observations from 1981 to the present. It includes effective LAI values derived from multiple satellite sensors across different product versions. Effective LAI is defined as half the total surface area of photosynthetically active plant elements per unit of horizontal ground area. It relates to true LAI through a canopy-dependent structure factor. As a key biophysical parameter, LAI is widely used to assess vegetation status, monitor ecosystem dynamics, and inform environmental and agricultural decision-making.\n\nThis quality assessment uses version 3 of the LAI product, selected because it covers the study period from 2007 to 2013. During this time, data were obtained from the System Pour l'Observation de la Terre (SPOT) satellite, Vegetation (VGT) sensor, with a spatial resolution of 1 km and a 10-day temporal resolution. The dataset is evaluated over Yunnan Province, China, with particular attention to the 2009-2010 drought.\n\nThe objectives of this assessment are: first, to assess whether the product's spatial and temporal completeness is sufficient to capture drought-related changes in LAI, which serves as a proxy for vegetation health; and second, to evaluate how different choices of quality flags influence the analysis."} {"chunk_id": "satellite_satellite-lai-fapar_extremes-detection_q02__75a7646a238c", "report_id": "satellite_satellite-lai-fapar_extremes-detection_q02", "dataset_id": "satellite-lai-fapar", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extremes-detection_q02", "aspect_base": "extremes-detection", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Impact of Drought on the Leaf Area Index in Yunnan Province, China > Quality assessment statement", "title": "Impact of Drought on the Leaf Area Index in Yunnan Province, China", "chunk_index": 1, "token_count": 415, "text_raw": "These are the key outcomes of this assessment\n\n* Version 3 of the C3S LAI dataset provides sufficient temporal and spatial completeness to detect the impact of the 2009–2010 drought on vegetation in Yunnan Province when an appropriate quality-filtering strategy is applied. \n* Applying a relaxed quality filter (bitmask 0x9C1) significantly increases LAI data availability during the wet season in Yunnan, maintaining over 84% completeness. However, this comes at the cost of increased uncertainty, as it includes pixels affected by albedo-related issues.\n* The conservative filter (bitmask 0xFC1), which excludes all flagged data, severely limits completeness during wet months, particularly from June to August, and especially in densely vegetated areas such as evergreen and mixed forests. However, this filter can be applied during the dry season with high completeness, while maintaining higher standards of data reliability.\n* While the relaxed and conservative filters yield comparable aggregated anomalies during dry-season conditions, particularly during the peak of the drought (e.g. March 2010), greater caution is warranted during wet, high-LAI periods, where retrieval uncertainties increase and filtering choices have a stronger influence on the aggregated signal.\n* The standardized LAI anomaly in Yunnan’s Tropic of Cancer zone was predominantly negative in March 2010, with extensive negative values also observed in the adjacent months, indicating that this period represents the strongest drought-related impact on LAI. This pattern is mostly consistent with the observations reported by Zhang et al. (2024) [[1]](https://doi.org/10.1038/s41598-024-58068-w).\n```", "text_with_prefix": "EQC Quality Assessment: \"Impact of Drought on the Leaf Area Index in Yunnan Province, China\"\nDataset: satellite-lai-fapar [CDS]\nAspect: extremes-detection_q02 | Category: Satellite_ECVs\nSection: Impact of Drought on the Leaf Area Index in Yunnan Province, China > Quality assessment statement\n---\nThese are the key outcomes of this assessment\n\n* Version 3 of the C3S LAI dataset provides sufficient temporal and spatial completeness to detect the impact of the 2009–2010 drought on vegetation in Yunnan Province when an appropriate quality-filtering strategy is applied. \n* Applying a relaxed quality filter (bitmask 0x9C1) significantly increases LAI data availability during the wet season in Yunnan, maintaining over 84% completeness. However, this comes at the cost of increased uncertainty, as it includes pixels affected by albedo-related issues.\n* The conservative filter (bitmask 0xFC1), which excludes all flagged data, severely limits completeness during wet months, particularly from June to August, and especially in densely vegetated areas such as evergreen and mixed forests. However, this filter can be applied during the dry season with high completeness, while maintaining higher standards of data reliability.\n* While the relaxed and conservative filters yield comparable aggregated anomalies during dry-season conditions, particularly during the peak of the drought (e.g. March 2010), greater caution is warranted during wet, high-LAI periods, where retrieval uncertainties increase and filtering choices have a stronger influence on the aggregated signal.\n* The standardized LAI anomaly in Yunnan’s Tropic of Cancer zone was predominantly negative in March 2010, with extensive negative values also observed in the adjacent months, indicating that this period represents the strongest drought-related impact on LAI. This pattern is mostly consistent with the observations reported by Zhang et al. (2024) [[1]](https://doi.org/10.1038/s41598-024-58068-w).\n```"} {"chunk_id": "satellite_satellite-lai-fapar_extremes-detection_q02__7ca35664f983", "report_id": "satellite_satellite-lai-fapar_extremes-detection_q02", "dataset_id": "satellite-lai-fapar", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extremes-detection_q02", "aspect_base": "extremes-detection", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Impact of Drought on the Leaf Area Index in Yunnan Province, China > Methodology", "title": "Impact of Drought on the Leaf Area Index in Yunnan Province, China", "chunk_index": 2, "token_count": 580, "text_raw": "The LAI variable from version 3 of the C3S product was downloaded for the period 2007-2013 and clipped to Yunnan Province, China. Two different quality filters were applied to the data: bitmask 0xFC1 (hereafter referred to as the conservative filter) and bitmask 0x9C1 (hereafter referred to as the relaxed filter). For each month in the study period, the percentage of completeness (the fraction of pixels that are not missing) was calculated under both filters for comparison. A date exhibiting a large difference in completeness between filters was selected, and the corresponding LAI map was visually compared with a land cover map to identify whether specific land cover types or zones were disproportionately affected by missing data. Monthly mean LAI over the entire province was then calculated and plotted for each year to assess whether the 2009-2010 drought is evident as lower LAI relative to other years. This analysis was performed for both filters to allow comparison. Finally, standardized LAI anomalies were calculated for Yunnan's Tropic of Cancer zone, and the results were compared with those reported by Zhang et al. (2024) [[1]](https://doi.org/10.1038/s41598-024-58068-w).\n\nThe analysis and results are organised in the following steps, which are detailed in the sections below:\n\n**[](section-1)** \n* Download LAI data from the SPOT satellite (VGT sensor) for the period 2007-2013.\n\n**[](section-2)** \n* Apply two distinct quality-masking approaches to filter the data. \n* Plot the percentage of valid LAI values over time for each approach. \n* Map LAI across Yunnan to compare the results from the two quality-masking methods. \n* Analyse the outcomes in relation to the dataset Documentation.\n\n**[](section-3)** \n* Calculate the spatial and monthly mean LAI, and plot the values for each year over the months of the year. Compare the results for the relaxed and the conservative filters.\n* Calculate the spatial and monthly standard deviation of LAI, and plot the values for the year 2010. Compare the results for the relaxed and the conservative filters.\n* Map LAI anomalies in Yunnan's Tropic of Cancer zone for the drought period from August 2009 to October 2010, relative to the reference years 2007-2008 and 2012-2013.", "text_with_prefix": "EQC Quality Assessment: \"Impact of Drought on the Leaf Area Index in Yunnan Province, China\"\nDataset: satellite-lai-fapar [CDS]\nAspect: extremes-detection_q02 | Category: Satellite_ECVs\nSection: Impact of Drought on the Leaf Area Index in Yunnan Province, China > Methodology\n---\nThe LAI variable from version 3 of the C3S product was downloaded for the period 2007-2013 and clipped to Yunnan Province, China. Two different quality filters were applied to the data: bitmask 0xFC1 (hereafter referred to as the conservative filter) and bitmask 0x9C1 (hereafter referred to as the relaxed filter). For each month in the study period, the percentage of completeness (the fraction of pixels that are not missing) was calculated under both filters for comparison. A date exhibiting a large difference in completeness between filters was selected, and the corresponding LAI map was visually compared with a land cover map to identify whether specific land cover types or zones were disproportionately affected by missing data. Monthly mean LAI over the entire province was then calculated and plotted for each year to assess whether the 2009-2010 drought is evident as lower LAI relative to other years. This analysis was performed for both filters to allow comparison. Finally, standardized LAI anomalies were calculated for Yunnan's Tropic of Cancer zone, and the results were compared with those reported by Zhang et al. (2024) [[1]](https://doi.org/10.1038/s41598-024-58068-w).\n\nThe analysis and results are organised in the following steps, which are detailed in the sections below:\n\n**[](section-1)** \n* Download LAI data from the SPOT satellite (VGT sensor) for the period 2007-2013.\n\n**[](section-2)** \n* Apply two distinct quality-masking approaches to filter the data. \n* Plot the percentage of valid LAI values over time for each approach. \n* Map LAI across Yunnan to compare the results from the two quality-masking methods. \n* Analyse the outcomes in relation to the dataset Documentation.\n\n**[](section-3)** \n* Calculate the spatial and monthly mean LAI, and plot the values for each year over the months of the year. Compare the results for the relaxed and the conservative filters.\n* Calculate the spatial and monthly standard deviation of LAI, and plot the values for the year 2010. Compare the results for the relaxed and the conservative filters.\n* Map LAI anomalies in Yunnan's Tropic of Cancer zone for the drought period from August 2009 to October 2010, relative to the reference years 2007-2008 and 2012-2013."} {"chunk_id": "satellite_satellite-lai-fapar_extremes-detection_q02__135fe9965d17", "report_id": "satellite_satellite-lai-fapar_extremes-detection_q02", "dataset_id": "satellite-lai-fapar", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extremes-detection_q02", "aspect_base": "extremes-detection", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Impact of Drought on the Leaf Area Index in Yunnan Province, China > Analysis and results > 1. Request and download data > Download and transform", "title": "Impact of Drought on the Leaf Area Index in Yunnan Province, China", "chunk_index": 3, "token_count": 490, "text_raw": "In this step, the dataset was spatially subsetted using a shapefile of Yunnan, an inland province in southwestern China. The shapefile was accessed from the GADM Database of Global Administrative Areas (version 4.1) via the UC Davis GeoData repository [[2]](https://gadm.org).\n\nTo filter out unreliable data, two distinct quality-masking strategies were applied:\n\n* **Conservative:** Retains only the highest-quality retrievals by applying bitmask 0xFC1. This mask excludes data flagged with any of the following bits: 0, 6, 7, 8, 9, 10, and 11, corresponding to missing input data, untrusted retrievals, unusable observations, inconsistent inputs, high albedo uncertainty (both snow and non-snow), and unreliable uncertainty estimates.\n* **Relaxed filter:** Permits retrievals with high albedo uncertainty (considered still usable), using bitmask 0x9C1. This mask excludes a smaller subset of flags: bits 0, 6, 7, 8, and 11, while allowing data flagged only for high albedo uncertainty in non-snow and snow conditions (bits 9 and 10).\n\nThese bitmasks are defined in Table 2, with flag definitions detailed in Table 7 of the [Product User Guide and Specification PUGS](https://dast.copernicus-climate.eu/documents/satellite-lai-fapar/D3.3.9-v3.0_PUGS_CDR_LAI_FAPAR_MULTI_SENSOR_v3.0_PRODUCTS_v1.0.1.pdf).\n\n```text\n100%|██████████| 84/84 [00:32<00:00, 2.58it/s]\n100%|██████████| 84/84 [00:33<00:00, 2.51it/s]\n```\n\n(section-2)=", "text_with_prefix": "EQC Quality Assessment: \"Impact of Drought on the Leaf Area Index in Yunnan Province, China\"\nDataset: satellite-lai-fapar [CDS]\nAspect: extremes-detection_q02 | Category: Satellite_ECVs\nSection: Impact of Drought on the Leaf Area Index in Yunnan Province, China > Analysis and results > 1. Request and download data > Download and transform\n---\nIn this step, the dataset was spatially subsetted using a shapefile of Yunnan, an inland province in southwestern China. The shapefile was accessed from the GADM Database of Global Administrative Areas (version 4.1) via the UC Davis GeoData repository [[2]](https://gadm.org).\n\nTo filter out unreliable data, two distinct quality-masking strategies were applied:\n\n* **Conservative:** Retains only the highest-quality retrievals by applying bitmask 0xFC1. This mask excludes data flagged with any of the following bits: 0, 6, 7, 8, 9, 10, and 11, corresponding to missing input data, untrusted retrievals, unusable observations, inconsistent inputs, high albedo uncertainty (both snow and non-snow), and unreliable uncertainty estimates.\n* **Relaxed filter:** Permits retrievals with high albedo uncertainty (considered still usable), using bitmask 0x9C1. This mask excludes a smaller subset of flags: bits 0, 6, 7, 8, and 11, while allowing data flagged only for high albedo uncertainty in non-snow and snow conditions (bits 9 and 10).\n\nThese bitmasks are defined in Table 2, with flag definitions detailed in Table 7 of the [Product User Guide and Specification PUGS](https://dast.copernicus-climate.eu/documents/satellite-lai-fapar/D3.3.9-v3.0_PUGS_CDR_LAI_FAPAR_MULTI_SENSOR_v3.0_PRODUCTS_v1.0.1.pdf).\n\n```text\n100%|██████████| 84/84 [00:32<00:00, 2.58it/s]\n100%|██████████| 84/84 [00:33<00:00, 2.51it/s]\n```\n\n(section-2)="} {"chunk_id": "satellite_satellite-lai-fapar_extremes-detection_q02__c930c3a93d95", "report_id": "satellite_satellite-lai-fapar_extremes-detection_q02", "dataset_id": "satellite-lai-fapar", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extremes-detection_q02", "aspect_base": "extremes-detection", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Impact of Drought on the Leaf Area Index in Yunnan Province, China > Analysis and results > 2. Completeness analysis > Filter data", "title": "Impact of Drought on the Leaf Area Index in Yunnan Province, China", "chunk_index": 4, "token_count": 1030, "text_raw": "For each time step, the percentage of valid LAI data points within the Yunnan region is computed, and the results are plotted over time. This operation is repeated using both the conservative and relaxed filters.\n\nThe goal of this comparison is to assess how different treatments of uncertainty flags affect data completeness over Yunnan, particularly during periods of increased retrieval uncertainty such as the rainy season.\n\n*Figure 1. Percentage of valid LAI values within the Yunnan region using the conservative mask (0xFC1) and the relaxed mask (0x9C1). Percentages are shown over time, with points color-coded by month. Note that the y-axis scale differs between the two panels.*\n\nUsing the conservative filter, the percentage of valid LAI data in Yunnan drops significantly between June and August, reaching levels below 20%. Adjacent months such as May and September also show reduced data completeness (see *Figure 1*, top pannel). This pattern corresponds to Yunnan’s marked wet and dry seasons [[3]](https://doi.org/10.1016/j.chnaes.2013.09.004), where the rainy season leads to a higher frequency of missing observations and albedo-related uncertainties, mainly due to cloud cover and wet canopy conditions.\n\nIn contrast, when applying the relaxed filter, the percentage of valid LAI data remains consistently above 84% throughout the period studied (2007–2013) (see *Figure 1*, bottom panel). This improvement is due to the relaxed mask allowing pixels flagged only for high albedo uncertainty in both non-snow and snow conditions (bits 9 and 10). In this context, bit 9 often relates to reflectance disturbances caused by clouds or wet vegetation, which are prevalent in Yunnan during the rainy season. As shown in Figure 15 of the [PQAR](https://dast.copernicus-climate.eu/documents/satellite-lai-fapar/D2.3.9-v3.0_PQAR_CDR_LAI_FAPAR_MULTI_SENSOR_v3.0_PRODUCTS_v1.1.pdf), the study area is frequently affected by the activation of bit 9, especially during wet months. It can also be observed that the dry months at the end of the year (November–December) exhibit higher completeness for the relaxed filter, but the difference between wet and dry seasons is significantly less pronounced than under the conservative filter.\n\nIt is also instructive to investigate the spatial pattern of missing LAI values. *Figure 2* shows LAI maps for August 10 from 2007 to 2013, comparing the conservative and relaxed filters. August 10 was chosen to exemplify differences in data completeness depending on the filter used beacuse it falls within the wet season and has low completeness under the conservative filter, as shown in *Figure 1*. The results highlight a strong contrast between the two approaches: the conservative filter produces highly incomplete maps across much of Yunnan in all years, whereas the relaxed filter provides nearly complete spatial coverage.\n\nIn some years, the conservative filter shows higher completeness in eastern Yunnan, where cropland, woody savannas, and grasslands are more prevalent than elsewhere in the province. Missing data under the conservative filter occur throughout Yunnan, particularly in high-LAI regions such as the mixed and evergreen forests in the west and southwest (see *Figure 3*).\n\n1. Build list of dates (August 10 of each year)\n2. Create figure and axes\nTitles and formatting\nShared colorbar\n3. Update function for animation\nEnsure latitude runs south → north\nUpdate array content\n4. Create animation\n\n1. Updated color list:\n0–0.2: white\n0.2–0.4: light yellow → 0.6: yellow → 0.8: deep yellow-green\n0.8–2.0: greens (dense vegetation)\n\n*Figure 2. Animated comparison of LAI maps on August 10 from 2007 to 2013 under the conservative (left) and relaxed (right) filters.*\n\n![land_cover.png](attachment:b1a763fc-0793-47fb-86e7-5c22b44c66d7.png)\n\n*Figure 3. Land cover map of Yunnan based on MODIS Land Surface Type data product -MCD12Q1. Figure reproduced from Kim et al. (2017) [[4]](https://doi.org/10.1007/s11629-016-3971-x).*\n\n(section-3)=", "text_with_prefix": "EQC Quality Assessment: \"Impact of Drought on the Leaf Area Index in Yunnan Province, China\"\nDataset: satellite-lai-fapar [CDS]\nAspect: extremes-detection_q02 | Category: Satellite_ECVs\nSection: Impact of Drought on the Leaf Area Index in Yunnan Province, China > Analysis and results > 2. Completeness analysis > Filter data\n---\nFor each time step, the percentage of valid LAI data points within the Yunnan region is computed, and the results are plotted over time. This operation is repeated using both the conservative and relaxed filters.\n\nThe goal of this comparison is to assess how different treatments of uncertainty flags affect data completeness over Yunnan, particularly during periods of increased retrieval uncertainty such as the rainy season.\n\n*Figure 1. Percentage of valid LAI values within the Yunnan region using the conservative mask (0xFC1) and the relaxed mask (0x9C1). Percentages are shown over time, with points color-coded by month. Note that the y-axis scale differs between the two panels.*\n\nUsing the conservative filter, the percentage of valid LAI data in Yunnan drops significantly between June and August, reaching levels below 20%. Adjacent months such as May and September also show reduced data completeness (see *Figure 1*, top pannel). This pattern corresponds to Yunnan’s marked wet and dry seasons [[3]](https://doi.org/10.1016/j.chnaes.2013.09.004), where the rainy season leads to a higher frequency of missing observations and albedo-related uncertainties, mainly due to cloud cover and wet canopy conditions.\n\nIn contrast, when applying the relaxed filter, the percentage of valid LAI data remains consistently above 84% throughout the period studied (2007–2013) (see *Figure 1*, bottom panel). This improvement is due to the relaxed mask allowing pixels flagged only for high albedo uncertainty in both non-snow and snow conditions (bits 9 and 10). In this context, bit 9 often relates to reflectance disturbances caused by clouds or wet vegetation, which are prevalent in Yunnan during the rainy season. As shown in Figure 15 of the [PQAR](https://dast.copernicus-climate.eu/documents/satellite-lai-fapar/D2.3.9-v3.0_PQAR_CDR_LAI_FAPAR_MULTI_SENSOR_v3.0_PRODUCTS_v1.1.pdf), the study area is frequently affected by the activation of bit 9, especially during wet months. It can also be observed that the dry months at the end of the year (November–December) exhibit higher completeness for the relaxed filter, but the difference between wet and dry seasons is significantly less pronounced than under the conservative filter.\n\nIt is also instructive to investigate the spatial pattern of missing LAI values. *Figure 2* shows LAI maps for August 10 from 2007 to 2013, comparing the conservative and relaxed filters. August 10 was chosen to exemplify differences in data completeness depending on the filter used beacuse it falls within the wet season and has low completeness under the conservative filter, as shown in *Figure 1*. The results highlight a strong contrast between the two approaches: the conservative filter produces highly incomplete maps across much of Yunnan in all years, whereas the relaxed filter provides nearly complete spatial coverage.\n\nIn some years, the conservative filter shows higher completeness in eastern Yunnan, where cropland, woody savannas, and grasslands are more prevalent than elsewhere in the province. Missing data under the conservative filter occur throughout Yunnan, particularly in high-LAI regions such as the mixed and evergreen forests in the west and southwest (see *Figure 3*).\n\n1. Build list of dates (August 10 of each year)\n2. Create figure and axes\nTitles and formatting\nShared colorbar\n3. Update function for animation\nEnsure latitude runs south → north\nUpdate array content\n4. Create animation\n\n1. Updated color list:\n0–0.2: white\n0.2–0.4: light yellow → 0.6: yellow → 0.8: deep yellow-green\n0.8–2.0: greens (dense vegetation)\n\n*Figure 2. Animated comparison of LAI maps on August 10 from 2007 to 2013 under the conservative (left) and relaxed (right) filters.*\n\n![land_cover.png](attachment:b1a763fc-0793-47fb-86e7-5c22b44c66d7.png)\n\n*Figure 3. Land cover map of Yunnan based on MODIS Land Surface Type data product -MCD12Q1. Figure reproduced from Kim et al. (2017) [[4]](https://doi.org/10.1007/s11629-016-3971-x).*\n\n(section-3)="} {"chunk_id": "satellite_satellite-lai-fapar_extremes-detection_q02__0ba3d250b8a9", "report_id": "satellite_satellite-lai-fapar_extremes-detection_q02", "dataset_id": "satellite-lai-fapar", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extremes-detection_q02", "aspect_base": "extremes-detection", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Impact of Drought on the Leaf Area Index in Yunnan Province, China > Analysis and results > 3. Drought effect on LAI > Spatial and monthly mean", "title": "Impact of Drought on the Leaf Area Index in Yunnan Province, China", "chunk_index": 5, "token_count": 1046, "text_raw": "Yunnan Province is positioned at the intersection of two major monsoon systems, the Indian monsoon and the East Asian monsoon, as well as the Qinghai-Tibet Plateau [[3]](https://doi.org/10.1016/j.chnaes.2013.09.004). These monsoons bring heavy rainfall during a distinct wet season, but due to the region’s varied topography, precipitation is unevenly distributed, often leading to water shortages during the dry season. Natural vegetation and wetlands help buffer this imbalance by storing and gradually releasing water. However, ongoing droughts have revealed the degradation of these ecosystems. Climate change is expected to intensify both the frequency and severity of droughts across Southeast Asia, including Yunnan [[4]](https://doi.org/10.1007/s11629-016-3971-x).\n\nYunnan has historically faced frequent droughts, with severe events occurring every two to three years between 1950 and 2014. The drought from autumn 2009 to spring 2010 was the most extreme since meteorological records began in 1959. It caused widespread ecological and agricultural disruption—drying up rivers and reservoirs, triggering large-scale crop failures, leading to the death of rare plant species, and altering water quality in plateau lakes [[1]](https://doi.org/10.1038/s41598-024-58068-w).\n\nDrought impacts on vegetation—such as reduced leaf growth, early senescence, and plant mortality—can be detected at regional scales using satellite-derived Leaf Area Index (LAI) data [[4]](https://doi.org/10.1007/s11629-016-3971-x). In the specific case of Yunnan, commonly used products include GLASS LAI (derived from MODIS surface reflectance) [[4]](https://doi.org/10.1007/s11629-016-3971-x) and MODIS LAI [[1]](https://doi.org/10.1038/s41598-024-58068-w), both of which are valuable for monitoring vegetation responses to drought.\n\nThe spatial and monthly mean LAI over Yunnan was calculated for each year from 2007 to 2013 to assess whether values in 2009 and 2010 were lower than in other years. These means were derived using both the relaxed and conservative filters to test whether the results support the earlier observations from *Figures 2* and *3*, namely, that the conservative filter tends to filter out high-LAI areas during the wet season, thereby lowering mean values compared with the relaxed filter.\n\n*Figure 4. Monthly mean LAI from 2007–2013 under conservative (left) and relaxed (right) filters, plotted by calendar month with consistent year-wise color coding.*\n\nUnder the relaxed filter (right panel of *Figure 4*), the values for 2009 appear to fall within the typical range for that time of year compared with other non-drought years. In contrast, many months in 2010 show visibly lower values than the corresponding months in other years. This pattern may reflect the delayed impact of the drought, with vegetation stress becoming more pronounced over time. Moreover, because the spatial mean represents an average across a large and ecologically diverse region, localized drought effects in 2009 may have been masked in the overall mean.\n\nAs expected, the conservative filter produces mean LAI values similar to those from the relaxed filter in months with high data completeness (November to April). However, in months with low completeness, the conservative filter yields noticeably lower mean LAI values. In 2010, mean LAI values are still reduced relative to other years, though the effect is observed in fewer months than when using the relaxed filter.\n\nThe monthly mean LAI for 2010 was computed by aggregating dekadal LAI retrievals spatially over the study region and temporally to monthly resolution. The associated retrieval uncertainty (LAI_ERR), provided as one standard deviation (1σ) under a Gaussian error assumption, was averaged over the same spatial and temporal domains and displayed as ±1σ (~68%) and ±2σ (~95%) bands around the monthly mean LAI (see *Figure 5*).\n\nIt is important to note that, because the uncertainty bands are derived from the mean of LAI_ERR values, they represent the typical retrieval uncertainty of individual observations within each month, rather than the formal uncertainty of the aggregated monthly mean LAI. In other words, the bands describe the magnitude of per-pixel retrieval error and should not be interpreted as confidence intervals of the spatially averaged mean.", "text_with_prefix": "EQC Quality Assessment: \"Impact of Drought on the Leaf Area Index in Yunnan Province, China\"\nDataset: satellite-lai-fapar [CDS]\nAspect: extremes-detection_q02 | Category: Satellite_ECVs\nSection: Impact of Drought on the Leaf Area Index in Yunnan Province, China > Analysis and results > 3. Drought effect on LAI > Spatial and monthly mean\n---\nYunnan Province is positioned at the intersection of two major monsoon systems, the Indian monsoon and the East Asian monsoon, as well as the Qinghai-Tibet Plateau [[3]](https://doi.org/10.1016/j.chnaes.2013.09.004). These monsoons bring heavy rainfall during a distinct wet season, but due to the region’s varied topography, precipitation is unevenly distributed, often leading to water shortages during the dry season. Natural vegetation and wetlands help buffer this imbalance by storing and gradually releasing water. However, ongoing droughts have revealed the degradation of these ecosystems. Climate change is expected to intensify both the frequency and severity of droughts across Southeast Asia, including Yunnan [[4]](https://doi.org/10.1007/s11629-016-3971-x).\n\nYunnan has historically faced frequent droughts, with severe events occurring every two to three years between 1950 and 2014. The drought from autumn 2009 to spring 2010 was the most extreme since meteorological records began in 1959. It caused widespread ecological and agricultural disruption—drying up rivers and reservoirs, triggering large-scale crop failures, leading to the death of rare plant species, and altering water quality in plateau lakes [[1]](https://doi.org/10.1038/s41598-024-58068-w).\n\nDrought impacts on vegetation—such as reduced leaf growth, early senescence, and plant mortality—can be detected at regional scales using satellite-derived Leaf Area Index (LAI) data [[4]](https://doi.org/10.1007/s11629-016-3971-x). In the specific case of Yunnan, commonly used products include GLASS LAI (derived from MODIS surface reflectance) [[4]](https://doi.org/10.1007/s11629-016-3971-x) and MODIS LAI [[1]](https://doi.org/10.1038/s41598-024-58068-w), both of which are valuable for monitoring vegetation responses to drought.\n\nThe spatial and monthly mean LAI over Yunnan was calculated for each year from 2007 to 2013 to assess whether values in 2009 and 2010 were lower than in other years. These means were derived using both the relaxed and conservative filters to test whether the results support the earlier observations from *Figures 2* and *3*, namely, that the conservative filter tends to filter out high-LAI areas during the wet season, thereby lowering mean values compared with the relaxed filter.\n\n*Figure 4. Monthly mean LAI from 2007–2013 under conservative (left) and relaxed (right) filters, plotted by calendar month with consistent year-wise color coding.*\n\nUnder the relaxed filter (right panel of *Figure 4*), the values for 2009 appear to fall within the typical range for that time of year compared with other non-drought years. In contrast, many months in 2010 show visibly lower values than the corresponding months in other years. This pattern may reflect the delayed impact of the drought, with vegetation stress becoming more pronounced over time. Moreover, because the spatial mean represents an average across a large and ecologically diverse region, localized drought effects in 2009 may have been masked in the overall mean.\n\nAs expected, the conservative filter produces mean LAI values similar to those from the relaxed filter in months with high data completeness (November to April). However, in months with low completeness, the conservative filter yields noticeably lower mean LAI values. In 2010, mean LAI values are still reduced relative to other years, though the effect is observed in fewer months than when using the relaxed filter.\n\nThe monthly mean LAI for 2010 was computed by aggregating dekadal LAI retrievals spatially over the study region and temporally to monthly resolution. The associated retrieval uncertainty (LAI_ERR), provided as one standard deviation (1σ) under a Gaussian error assumption, was averaged over the same spatial and temporal domains and displayed as ±1σ (~68%) and ±2σ (~95%) bands around the monthly mean LAI (see *Figure 5*).\n\nIt is important to note that, because the uncertainty bands are derived from the mean of LAI_ERR values, they represent the typical retrieval uncertainty of individual observations within each month, rather than the formal uncertainty of the aggregated monthly mean LAI. In other words, the bands describe the magnitude of per-pixel retrieval error and should not be interpreted as confidence intervals of the spatially averaged mean."} {"chunk_id": "satellite_satellite-lai-fapar_extremes-detection_q02__d8ac2a25a9ad", "report_id": "satellite_satellite-lai-fapar_extremes-detection_q02", "dataset_id": "satellite-lai-fapar", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extremes-detection_q02", "aspect_base": "extremes-detection", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Impact of Drought on the Leaf Area Index in Yunnan Province, China > Analysis and results > 3. Drought effect on LAI > Spatial and monthly mean", "title": "Impact of Drought on the Leaf Area Index in Yunnan Province, China", "chunk_index": 6, "token_count": 1131, "text_raw": "within each month, rather than the formal uncertainty of the aggregated monthly mean LAI. In other words, the bands describe the magnitude of per-pixel retrieval error and should not be interpreted as confidence intervals of the spatially averaged mean.\n\nThe uncertainty envelopes are relatively wide compared to the magnitude of the mean LAI values. This indicates that retrieval uncertainty is of comparable magnitude to the signal itself, particularly under low-LAI conditions. During the dry months, both filtering strategies yield similar uncertainty levels, suggesting comparable retrieval performance. In contrast, during the wet season (June–August), the relaxed filter exhibits substantially wider uncertainty bands, with moderately elevated uncertainty in May and September. This pattern suggests that under dense canopy conditions, where radiative transfer inversion becomes less constrained, the relaxed filtering approach retains more retrievals with higher associated uncertainty. The wider bands therefore reflect increased retrieval imprecision rather than increased spatial variability.\n\nBuild legend entries for this filter\n----- Custom legend -----\n\n*Figure 5. Monthly and spatial mean of LAI and ±1σ (~68%) and ±2σ (~95%) (bands) over Yunnan province in 2010 under conservative (top) and relaxed (bottom) filters.*\n\nBecause the retrieval uncertainty (LAI_ERR) was found to be large relative to the magnitude of the monthly mean LAI, a simple arithmetic mean would assign equal weight to observations with very different reliability. To account for this heteroscedastic uncertainty, we therefore computed inverse-variance weighted monthly and spatial means for 2010. Let $L_{f,t,x}$ denote LAI for filter $f$, dekadal time step $t$, and pixel $x$, and let $\\sigma_{f,t,x}$ denote the corresponding LAI_ERR (1σ retrieval uncertainty). For all finite observations with $\\sigma_{f,t,x} > 0$, we define the weight as\n\n$$\nw_{f,t,x} = \\frac{1}{\\sigma_{f,t,x}^{2}},\n$$\n\nand set the weight to zero otherwise. Monthly means were first computed per pixel using a weighted average over all dekadal observations within month $m$,\n\n$$\n\\mu_{f,m,x} =\n\\frac{\\sum_{t \\in m} w_{f,t,x} \\, L_{f,t,x}}\n{\\sum_{t \\in m} w_{f,t,x}}.\n$$\n\nAssuming independent Gaussian errors and that LAI_ERR represents the standard deviation of each observation, the propagated uncertainty of this weighted monthly mean is\n\n$$\n\\sigma_{f,m,x}^{(\\mathrm{month})}\n=\\sqrt{\\frac{1}{\\sum_{t \\in m} w_{f,t,x}}}.\n$$\n\nTo obtain a single regional value per month, these monthly pixel means were then aggregated spatially using inverse-variance weights based on their propagated uncertainty,\n\n$$\nW_{f,m,x}\n=\\frac{1}{\\left(\\sigma_{f,m,x}^{(\\mathrm{month})}\\right)^2},\n$$\n\nyielding the spatially weighted monthly mean\n\n$$\n\\mu_{f,m}\n=\\frac{\\sum_{x} W_{f,m,x} \\, \\mu_{f,m,x}}\n{\\sum_{x} W_{f,m,x}}.\n$$\n\nThis approach reduces the influence of observations with large retrieval uncertainty and yields a minimum-variance linear estimator of the monthly regional mean under the assumptions of independent Gaussian errors.\n\nmonthly per-pixel\nspatial per-month\n\n*Figure 6. Monthly inverse-variance weighted spatial mean LAI for 2010.*\n\n--- Inputs ---\nSimple (your existing monthly mean LAI from monthly_mean)\nWeighted (from your inverse-variance workflow; already computed earlier)\nIf you have `mu` already, use it; otherwise uncomment the next line:\nmu = monthly_spatial_weighted.compute()\n--- Plot: 2 columns ---\nLeft: simple mean\nRight: weighted mean\n\n*Figure 7. Monthly spatial mean LAI for 2010 computed using (left) simple arithmetic averaging and (right) inverse-variance weighting. Lines represent conservative and relaxed filters.*\n\nIn *Figure 6*, the propagated uncertainty band (±1σ) are not visibly distinguishable from the mean curves. This indicates that the formal uncertainty of the spatially aggregated monthly mean LAI is very small relative to the magnitude of the mean itself. Under the assumption of independent Gaussian errors, the inverse-variance weighting leads to a strong reduction of uncertainty when aggregating across a large number of observations. Consequently, the standard deviation of the regional mean becomes negligible at the scale of the figure.\n\n*Figure 7* compares the simple (unweighted) and inverse-variance weighted monthly means. From May to October, the relaxed filter yields consistently higher LAI values than the conservative filter, with the largest differences occurring during the wet season. When inverse-variance weighting is applied, the monthly mean LAI decreases for both filtering schemes across all months. This indicates that observations with higher retrieval uncertainty tend to be associated with higher LAI values and are down-weighted in the weighted aggregation.", "text_with_prefix": "EQC Quality Assessment: \"Impact of Drought on the Leaf Area Index in Yunnan Province, China\"\nDataset: satellite-lai-fapar [CDS]\nAspect: extremes-detection_q02 | Category: Satellite_ECVs\nSection: Impact of Drought on the Leaf Area Index in Yunnan Province, China > Analysis and results > 3. Drought effect on LAI > Spatial and monthly mean\n---\nwithin each month, rather than the formal uncertainty of the aggregated monthly mean LAI. In other words, the bands describe the magnitude of per-pixel retrieval error and should not be interpreted as confidence intervals of the spatially averaged mean.\n\nThe uncertainty envelopes are relatively wide compared to the magnitude of the mean LAI values. This indicates that retrieval uncertainty is of comparable magnitude to the signal itself, particularly under low-LAI conditions. During the dry months, both filtering strategies yield similar uncertainty levels, suggesting comparable retrieval performance. In contrast, during the wet season (June–August), the relaxed filter exhibits substantially wider uncertainty bands, with moderately elevated uncertainty in May and September. This pattern suggests that under dense canopy conditions, where radiative transfer inversion becomes less constrained, the relaxed filtering approach retains more retrievals with higher associated uncertainty. The wider bands therefore reflect increased retrieval imprecision rather than increased spatial variability.\n\nBuild legend entries for this filter\n----- Custom legend -----\n\n*Figure 5. Monthly and spatial mean of LAI and ±1σ (~68%) and ±2σ (~95%) (bands) over Yunnan province in 2010 under conservative (top) and relaxed (bottom) filters.*\n\nBecause the retrieval uncertainty (LAI_ERR) was found to be large relative to the magnitude of the monthly mean LAI, a simple arithmetic mean would assign equal weight to observations with very different reliability. To account for this heteroscedastic uncertainty, we therefore computed inverse-variance weighted monthly and spatial means for 2010. Let $L_{f,t,x}$ denote LAI for filter $f$, dekadal time step $t$, and pixel $x$, and let $\\sigma_{f,t,x}$ denote the corresponding LAI_ERR (1σ retrieval uncertainty). For all finite observations with $\\sigma_{f,t,x} > 0$, we define the weight as\n\n$$\nw_{f,t,x} = \\frac{1}{\\sigma_{f,t,x}^{2}},\n$$\n\nand set the weight to zero otherwise. Monthly means were first computed per pixel using a weighted average over all dekadal observations within month $m$,\n\n$$\n\\mu_{f,m,x} =\n\\frac{\\sum_{t \\in m} w_{f,t,x} \\, L_{f,t,x}}\n{\\sum_{t \\in m} w_{f,t,x}}.\n$$\n\nAssuming independent Gaussian errors and that LAI_ERR represents the standard deviation of each observation, the propagated uncertainty of this weighted monthly mean is\n\n$$\n\\sigma_{f,m,x}^{(\\mathrm{month})}\n=\\sqrt{\\frac{1}{\\sum_{t \\in m} w_{f,t,x}}}.\n$$\n\nTo obtain a single regional value per month, these monthly pixel means were then aggregated spatially using inverse-variance weights based on their propagated uncertainty,\n\n$$\nW_{f,m,x}\n=\\frac{1}{\\left(\\sigma_{f,m,x}^{(\\mathrm{month})}\\right)^2},\n$$\n\nyielding the spatially weighted monthly mean\n\n$$\n\\mu_{f,m}\n=\\frac{\\sum_{x} W_{f,m,x} \\, \\mu_{f,m,x}}\n{\\sum_{x} W_{f,m,x}}.\n$$\n\nThis approach reduces the influence of observations with large retrieval uncertainty and yields a minimum-variance linear estimator of the monthly regional mean under the assumptions of independent Gaussian errors.\n\nmonthly per-pixel\nspatial per-month\n\n*Figure 6. Monthly inverse-variance weighted spatial mean LAI for 2010.*\n\n--- Inputs ---\nSimple (your existing monthly mean LAI from monthly_mean)\nWeighted (from your inverse-variance workflow; already computed earlier)\nIf you have `mu` already, use it; otherwise uncomment the next line:\nmu = monthly_spatial_weighted.compute()\n--- Plot: 2 columns ---\nLeft: simple mean\nRight: weighted mean\n\n*Figure 7. Monthly spatial mean LAI for 2010 computed using (left) simple arithmetic averaging and (right) inverse-variance weighting. Lines represent conservative and relaxed filters.*\n\nIn *Figure 6*, the propagated uncertainty band (±1σ) are not visibly distinguishable from the mean curves. This indicates that the formal uncertainty of the spatially aggregated monthly mean LAI is very small relative to the magnitude of the mean itself. Under the assumption of independent Gaussian errors, the inverse-variance weighting leads to a strong reduction of uncertainty when aggregating across a large number of observations. Consequently, the standard deviation of the regional mean becomes negligible at the scale of the figure.\n\n*Figure 7* compares the simple (unweighted) and inverse-variance weighted monthly means. From May to October, the relaxed filter yields consistently higher LAI values than the conservative filter, with the largest differences occurring during the wet season. When inverse-variance weighting is applied, the monthly mean LAI decreases for both filtering schemes across all months. This indicates that observations with higher retrieval uncertainty tend to be associated with higher LAI values and are down-weighted in the weighted aggregation."} {"chunk_id": "satellite_satellite-lai-fapar_extremes-detection_q02__d51c42c7915b", "report_id": "satellite_satellite-lai-fapar_extremes-detection_q02", "dataset_id": "satellite-lai-fapar", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extremes-detection_q02", "aspect_base": "extremes-detection", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Impact of Drought on the Leaf Area Index in Yunnan Province, China > Analysis and results > 3. Drought effect on LAI > Spatial and monthly mean", "title": "Impact of Drought on the Leaf Area Index in Yunnan Province, China", "chunk_index": 7, "token_count": 309, "text_raw": "When inverse-variance weighting is applied, the monthly mean LAI decreases for both filtering schemes across all months. This indicates that observations with higher retrieval uncertainty tend to be associated with higher LAI values and are down-weighted in the weighted aggregation.\n\nThe effect is most pronounced during the wet months, particularly in September, where the discrepancy between simple and weighted means is largest. This suggests that retrieval uncertainty increases under dense canopy conditions and that the relaxed filtering scheme retains a greater proportion of high-uncertainty observations during peak vegetation periods. As a result, inverse-variance weighting exerts a stronger downward adjustment in these months, reducing the influence of less reliable retrievals. During the dry months (March and April), when mean LAI values are lowest, the inverse-variance weighted means remain lower than the simple means, but the differences are substantially smaller than those observed during the wet season.\n\nOverall, the comparison demonstrates that accounting for heteroscedastic retrieval uncertainty meaningfully alters the magnitude of the seasonal LAI cycle, particularly during periods of high vegetation density, while leaving the general seasonal pattern intact.", "text_with_prefix": "EQC Quality Assessment: \"Impact of Drought on the Leaf Area Index in Yunnan Province, China\"\nDataset: satellite-lai-fapar [CDS]\nAspect: extremes-detection_q02 | Category: Satellite_ECVs\nSection: Impact of Drought on the Leaf Area Index in Yunnan Province, China > Analysis and results > 3. Drought effect on LAI > Spatial and monthly mean\n---\nWhen inverse-variance weighting is applied, the monthly mean LAI decreases for both filtering schemes across all months. This indicates that observations with higher retrieval uncertainty tend to be associated with higher LAI values and are down-weighted in the weighted aggregation.\n\nThe effect is most pronounced during the wet months, particularly in September, where the discrepancy between simple and weighted means is largest. This suggests that retrieval uncertainty increases under dense canopy conditions and that the relaxed filtering scheme retains a greater proportion of high-uncertainty observations during peak vegetation periods. As a result, inverse-variance weighting exerts a stronger downward adjustment in these months, reducing the influence of less reliable retrievals. During the dry months (March and April), when mean LAI values are lowest, the inverse-variance weighted means remain lower than the simple means, but the differences are substantially smaller than those observed during the wet season.\n\nOverall, the comparison demonstrates that accounting for heteroscedastic retrieval uncertainty meaningfully alters the magnitude of the seasonal LAI cycle, particularly during periods of high vegetation density, while leaving the general seasonal pattern intact."} {"chunk_id": "satellite_satellite-lai-fapar_extremes-detection_q02__91241cb497ff", "report_id": "satellite_satellite-lai-fapar_extremes-detection_q02", "dataset_id": "satellite-lai-fapar", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extremes-detection_q02", "aspect_base": "extremes-detection", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Impact of Drought on the Leaf Area Index in Yunnan Province, China > Analysis and results > 3. Drought effect on LAI > Standardised LAI anomaly", "title": "Impact of Drought on the Leaf Area Index in Yunnan Province, China", "chunk_index": 8, "token_count": 1067, "text_raw": "Zhang et al. (2024) [[1]](https://doi.org/10.1038/s41598-024-58068-w) investigated the effects of drought on vegetation in the Tropic of Cancer zone of Yunnan. To achieve this, they used various vegetation health indicators, including the LAI, which was obtained from the MOD15A2H product (layer name: LAI_500m). They calculated the monthly LAI anomaly for the drought period from August 2009 to October 2010. The reference periods 2007–2008 and 2012–2013 were chosen as the baseline to compute the LAI anomalies because they experienced minimal drought events and are adjacent to the drought years under investigation (2009–2010). This proximity helps control for long-term changes, such as those driven by human activity, that might otherwise skew the analysis [[1]](https://doi.org/10.1038/s41598-024-58068-w). For example, northwestern Yunnan was a key region in the Grain-for-Green Programme, which may have significantly influenced vegetation dynamics [[5]](https://doi.org/10.1016/j.scitotenv.2016.09.026).\n\nIn this assessment, we applied the same approach as Zhang et al. (2024) to the C3S LAI product with the relaxed filter, using the same reference period (2007–2008 and 2012–2013) to compute the standardised LAI anomaly:\n\n$\\text{LAI anomaly} = \\frac{\\text{LAI}_t - \\overline{\\text{LAI}}}{\\delta_{\\text{LAI}}}$\n\nwhere:\n\n$\\text{LAI}_t$ = LAI 2010 for a specific month for that pixel\n\n$\\overline{\\text{LAI}}$ = Mean LAI 2007-2008 and 2012-2013 for that same month for that pixel\n\n$\\delta_{\\text{LAI}}$ = Standard deviation LAI 2007-2008 and 2012-2013 for that same month for that pixel\n\nExtract LAI data\n--- Step 1. Clip to geographic range (Tropic of Cancer region in Yunnan) ---\n--- Step 2. Define normal period (2007–2008 and 2012–2013) ---\n--- Step 3. Compute monthly climatology (mean + std) ---\n--- Step 4. Select target years (2009 and 2010) ---\n--- Step 5. Compute monthly means for each year ---\n--- Step 6. Calculate standardized monthly anomaly (SMA) ---\nAvoid inf/-inf\n\n--- Custom colormap ---\n\n--- Subset anomalies to Aug 2009 – Oct 2010 ---\nTime labels for subplot titles\n--- Set up 3×4 subplot grid ---\n--- Plot each month ---\nSet to the Tropic of Cancer rectangle\n--- Date label under each subplot ---\n--- Gridlines ---\n--- Longitude ticks only for top row ---\n--- Latitude ticks for left column only ---\n--- Shared colorbar at bottom ---\n--- Main title ---\n\n*Figure 6. Monthly standardized LAI anomalies during the drought period from August 2009 to October 2010, calculated using the C3S LAI product (relaxed quality filter).*\n\n![Zhang2024.jpg](attachment:21b489ba-8fcd-4b82-80eb-f14234020195.jpg)\n\n*Figure 7. Monthly standardized LAI anomalies during the drought period from August 2009 to October 2010, based on the MOD15A2H product. Source: Zhang et al. (2024) [[1]](https://doi.org/10.1038/s41598-024-58068-w)*\n\nAccording to the Standardized Precipitation Evapotranspiration Index (SPEI) calculated by Zhang et al. (2024) [[1]](https://doi.org/10.1038/s41598-024-58068-w), the drought fully developed across the study area between October 2009 and February 2010. Because LAI has a lagged response to drought, the results obtained with the C3S product (*Figure 6*) are consistent: March 2010 shows the largest areas with negative anomalies, with adjacent months also displaying widespread negative values. These findings align with Zhang et al. (2024) [[1]](https://doi.org/10.1038/s41598-024-58068-w), who reported that the period of greatest vegetation stress due to the cumulative effects of drought occurred between March and May 2010, corresponding to spring.", "text_with_prefix": "EQC Quality Assessment: \"Impact of Drought on the Leaf Area Index in Yunnan Province, China\"\nDataset: satellite-lai-fapar [CDS]\nAspect: extremes-detection_q02 | Category: Satellite_ECVs\nSection: Impact of Drought on the Leaf Area Index in Yunnan Province, China > Analysis and results > 3. Drought effect on LAI > Standardised LAI anomaly\n---\nZhang et al. (2024) [[1]](https://doi.org/10.1038/s41598-024-58068-w) investigated the effects of drought on vegetation in the Tropic of Cancer zone of Yunnan. To achieve this, they used various vegetation health indicators, including the LAI, which was obtained from the MOD15A2H product (layer name: LAI_500m). They calculated the monthly LAI anomaly for the drought period from August 2009 to October 2010. The reference periods 2007–2008 and 2012–2013 were chosen as the baseline to compute the LAI anomalies because they experienced minimal drought events and are adjacent to the drought years under investigation (2009–2010). This proximity helps control for long-term changes, such as those driven by human activity, that might otherwise skew the analysis [[1]](https://doi.org/10.1038/s41598-024-58068-w). For example, northwestern Yunnan was a key region in the Grain-for-Green Programme, which may have significantly influenced vegetation dynamics [[5]](https://doi.org/10.1016/j.scitotenv.2016.09.026).\n\nIn this assessment, we applied the same approach as Zhang et al. (2024) to the C3S LAI product with the relaxed filter, using the same reference period (2007–2008 and 2012–2013) to compute the standardised LAI anomaly:\n\n$\\text{LAI anomaly} = \\frac{\\text{LAI}_t - \\overline{\\text{LAI}}}{\\delta_{\\text{LAI}}}$\n\nwhere:\n\n$\\text{LAI}_t$ = LAI 2010 for a specific month for that pixel\n\n$\\overline{\\text{LAI}}$ = Mean LAI 2007-2008 and 2012-2013 for that same month for that pixel\n\n$\\delta_{\\text{LAI}}$ = Standard deviation LAI 2007-2008 and 2012-2013 for that same month for that pixel\n\nExtract LAI data\n--- Step 1. Clip to geographic range (Tropic of Cancer region in Yunnan) ---\n--- Step 2. Define normal period (2007–2008 and 2012–2013) ---\n--- Step 3. Compute monthly climatology (mean + std) ---\n--- Step 4. Select target years (2009 and 2010) ---\n--- Step 5. Compute monthly means for each year ---\n--- Step 6. Calculate standardized monthly anomaly (SMA) ---\nAvoid inf/-inf\n\n--- Custom colormap ---\n\n--- Subset anomalies to Aug 2009 – Oct 2010 ---\nTime labels for subplot titles\n--- Set up 3×4 subplot grid ---\n--- Plot each month ---\nSet to the Tropic of Cancer rectangle\n--- Date label under each subplot ---\n--- Gridlines ---\n--- Longitude ticks only for top row ---\n--- Latitude ticks for left column only ---\n--- Shared colorbar at bottom ---\n--- Main title ---\n\n*Figure 6. Monthly standardized LAI anomalies during the drought period from August 2009 to October 2010, calculated using the C3S LAI product (relaxed quality filter).*\n\n![Zhang2024.jpg](attachment:21b489ba-8fcd-4b82-80eb-f14234020195.jpg)\n\n*Figure 7. Monthly standardized LAI anomalies during the drought period from August 2009 to October 2010, based on the MOD15A2H product. Source: Zhang et al. (2024) [[1]](https://doi.org/10.1038/s41598-024-58068-w)*\n\nAccording to the Standardized Precipitation Evapotranspiration Index (SPEI) calculated by Zhang et al. (2024) [[1]](https://doi.org/10.1038/s41598-024-58068-w), the drought fully developed across the study area between October 2009 and February 2010. Because LAI has a lagged response to drought, the results obtained with the C3S product (*Figure 6*) are consistent: March 2010 shows the largest areas with negative anomalies, with adjacent months also displaying widespread negative values. These findings align with Zhang et al. (2024) [[1]](https://doi.org/10.1038/s41598-024-58068-w), who reported that the period of greatest vegetation stress due to the cumulative effects of drought occurred between March and May 2010, corresponding to spring."} {"chunk_id": "satellite_satellite-lai-fapar_extremes-detection_q02__78682a3d44a5", "report_id": "satellite_satellite-lai-fapar_extremes-detection_q02", "dataset_id": "satellite-lai-fapar", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extremes-detection_q02", "aspect_base": "extremes-detection", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Impact of Drought on the Leaf Area Index in Yunnan Province, China > Analysis and results > 3. Drought effect on LAI > Standardised LAI anomaly", "title": "Impact of Drought on the Leaf Area Index in Yunnan Province, China", "chunk_index": 9, "token_count": 467, "text_raw": "https://doi.org/10.1038/s41598-024-58068-w), who reported that the period of greatest vegetation stress due to the cumulative effects of drought occurred between March and May 2010, corresponding to spring.\n\nSome differences can be observed between the *Figure 6* (C3S) and *Figure 7* (MODIS). C3S LAI anomalies then to show more extremes (values <-2 or 2>) than the anomamlies calculated using MODIS. For example, it can be noted that the anomalies in March 2010 using the C3S were even more negative than the ones using MODIS. Also, the C3S data show positive anomalies in the western part of the study area during October and November 2009, as well as January and May 2010—patterns that are absent in the MODIS results. Another difference is that, according to MODIS, vegetation LAI had largely returned to normal conditions by October 2010, whereas the C3S data still indicate scattered pixels with negative anomalies.\n\nAlthough LAI responds more slowly to drought because it reflects cumulative canopy biomass rather than immediate physiological stress, it tends to recover relatively quickly once water availability improves, since canopy structure is often not severely damaged in the short term. In contrast, gross primary productivity (GPP) responds directly to water stress and requires more time to recover, as plants must restore physiological functions such as stomatal conductance and photosynthetic activity. This makes GPP a more sensitive indicator of drought impacts and a better tool for identifying the regions most affected [[1]](https://doi.org/10.1038/s41598-024-58068-w). Nevertheless, LAI remains valuable for assessing longer-term structural changes in vegetation and for linking drought effects to phenological dynamics and canopy development.", "text_with_prefix": "EQC Quality Assessment: \"Impact of Drought on the Leaf Area Index in Yunnan Province, China\"\nDataset: satellite-lai-fapar [CDS]\nAspect: extremes-detection_q02 | Category: Satellite_ECVs\nSection: Impact of Drought on the Leaf Area Index in Yunnan Province, China > Analysis and results > 3. Drought effect on LAI > Standardised LAI anomaly\n---\nhttps://doi.org/10.1038/s41598-024-58068-w), who reported that the period of greatest vegetation stress due to the cumulative effects of drought occurred between March and May 2010, corresponding to spring.\n\nSome differences can be observed between the *Figure 6* (C3S) and *Figure 7* (MODIS). C3S LAI anomalies then to show more extremes (values <-2 or 2>) than the anomamlies calculated using MODIS. For example, it can be noted that the anomalies in March 2010 using the C3S were even more negative than the ones using MODIS. Also, the C3S data show positive anomalies in the western part of the study area during October and November 2009, as well as January and May 2010—patterns that are absent in the MODIS results. Another difference is that, according to MODIS, vegetation LAI had largely returned to normal conditions by October 2010, whereas the C3S data still indicate scattered pixels with negative anomalies.\n\nAlthough LAI responds more slowly to drought because it reflects cumulative canopy biomass rather than immediate physiological stress, it tends to recover relatively quickly once water availability improves, since canopy structure is often not severely damaged in the short term. In contrast, gross primary productivity (GPP) responds directly to water stress and requires more time to recover, as plants must restore physiological functions such as stomatal conductance and photosynthetic activity. This makes GPP a more sensitive indicator of drought impacts and a better tool for identifying the regions most affected [[1]](https://doi.org/10.1038/s41598-024-58068-w). Nevertheless, LAI remains valuable for assessing longer-term structural changes in vegetation and for linking drought effects to phenological dynamics and canopy development."} {"chunk_id": "satellite_satellite-lai-fapar_extremes-detection_q02__a04db1c7ad24", "report_id": "satellite_satellite-lai-fapar_extremes-detection_q02", "dataset_id": "satellite-lai-fapar", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extremes-detection_q02", "aspect_base": "extremes-detection", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Impact of Drought on the Leaf Area Index in Yunnan Province, China > ℹ️ If you want to know more", "title": "Impact of Drought on the Leaf Area Index in Yunnan Province, China", "chunk_index": 10, "token_count": 290, "text_raw": "* Zhu, H. & Tan, Y. (2022). Flora and Vegetation of Yunnan, Southwestern China: Diversity, Origin and Evolution. Diversity, 14(5), 340. [](https://doi.org/10.3390/d14050340)\n* Qian, L.-S., Chen, J.-H., Deng, T., & Sun, H. (2020). Plant diversity in Yunnan: Current status and future directions. Plant Diversity, 42(4), 281–291. [](https://doi.org/10.1016/j.pld.2020.07.006)\n* Lan, T., & Yan, X. (2024). Analysis of drought characteristics and causes in Yunnan Province in the last 60 years (1961–2020). Journal of Hydrometeorology, 25, 177–190. [](https://doi.org/10.1175/JHM-D-23-0092.1)", "text_with_prefix": "EQC Quality Assessment: \"Impact of Drought on the Leaf Area Index in Yunnan Province, China\"\nDataset: satellite-lai-fapar [CDS]\nAspect: extremes-detection_q02 | Category: Satellite_ECVs\nSection: Impact of Drought on the Leaf Area Index in Yunnan Province, China > ℹ️ If you want to know more\n---\n* Zhu, H. & Tan, Y. (2022). Flora and Vegetation of Yunnan, Southwestern China: Diversity, Origin and Evolution. Diversity, 14(5), 340. [](https://doi.org/10.3390/d14050340)\n* Qian, L.-S., Chen, J.-H., Deng, T., & Sun, H. (2020). Plant diversity in Yunnan: Current status and future directions. Plant Diversity, 42(4), 281–291. [](https://doi.org/10.1016/j.pld.2020.07.006)\n* Lan, T., & Yan, X. (2024). Analysis of drought characteristics and causes in Yunnan Province in the last 60 years (1961–2020). Journal of Hydrometeorology, 25, 177–190. [](https://doi.org/10.1175/JHM-D-23-0092.1)"} {"chunk_id": "satellite_satellite-lai-fapar_extremes-detection_q02__0c1d1c1ef008", "report_id": "satellite_satellite-lai-fapar_extremes-detection_q02", "dataset_id": "satellite-lai-fapar", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extremes-detection_q02", "aspect_base": "extremes-detection", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Impact of Drought on the Leaf Area Index in Yunnan Province, China > ℹ️ If you want to know more > Key resources", "title": "Impact of Drought on the Leaf Area Index in Yunnan Province, China", "chunk_index": 11, "token_count": 478, "text_raw": "Code libraries used:\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)\n\nDocumentation:\n* [Product User Guide and Specification (PUGS) - Multi-sensor CDR LAI and fAPAR v3.0](https://dast.copernicus-climate.eu/documents/satellite-lai-fapar/D3.3.9-v3.0_PUGS_CDR_LAI_FAPAR_MULTI_SENSOR_v3.0_PRODUCTS_v1.0.1.pdf)\n* [Algorithm Theoretical Basis Document (ATBD) - Multi-sensor CDR LAI and fAPAR v3.0](https://dast.copernicus-climate.eu/documents/satellite-lai-fapar/D1.4.4-v3.0_ATBD_CDR_LAI_FAPAR_MULTI_SENSOR_v3.0_PRODUCTS_v1.0.1.pdf)\n* [Product Quality Assurance Document (PQAD) - Multi-sensor LAI and fAPAR v3.0](https://dast.copernicus-climate.eu/documents/satellite-lai-fapar/D2.2.9-v3.0_PQAD_CDR_LAI_FAPAR_MULTI_SENSOR_v3.0_PRODUCTS_v1.1.pdf)\n* [Product Quality Assessment Report (PQAR) - Multi-sensor LAI and fAPAR v3.0](https://dast.copernicus-climate.eu/documents/satellite-lai-fapar/D2.3.9-v3.0_PQAR_CDR_LAI_FAPAR_MULTI_SENSOR_v3.0_PRODUCTS_v1.1.pdf)", "text_with_prefix": "EQC Quality Assessment: \"Impact of Drought on the Leaf Area Index in Yunnan Province, China\"\nDataset: satellite-lai-fapar [CDS]\nAspect: extremes-detection_q02 | Category: Satellite_ECVs\nSection: Impact of Drought on the Leaf Area Index in Yunnan Province, China > ℹ️ If you want to know more > Key resources\n---\nCode libraries used:\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)\n\nDocumentation:\n* [Product User Guide and Specification (PUGS) - Multi-sensor CDR LAI and fAPAR v3.0](https://dast.copernicus-climate.eu/documents/satellite-lai-fapar/D3.3.9-v3.0_PUGS_CDR_LAI_FAPAR_MULTI_SENSOR_v3.0_PRODUCTS_v1.0.1.pdf)\n* [Algorithm Theoretical Basis Document (ATBD) - Multi-sensor CDR LAI and fAPAR v3.0](https://dast.copernicus-climate.eu/documents/satellite-lai-fapar/D1.4.4-v3.0_ATBD_CDR_LAI_FAPAR_MULTI_SENSOR_v3.0_PRODUCTS_v1.0.1.pdf)\n* [Product Quality Assurance Document (PQAD) - Multi-sensor LAI and fAPAR v3.0](https://dast.copernicus-climate.eu/documents/satellite-lai-fapar/D2.2.9-v3.0_PQAD_CDR_LAI_FAPAR_MULTI_SENSOR_v3.0_PRODUCTS_v1.1.pdf)\n* [Product Quality Assessment Report (PQAR) - Multi-sensor LAI and fAPAR v3.0](https://dast.copernicus-climate.eu/documents/satellite-lai-fapar/D2.3.9-v3.0_PQAR_CDR_LAI_FAPAR_MULTI_SENSOR_v3.0_PRODUCTS_v1.1.pdf)"} {"chunk_id": "satellite_satellite-lai-fapar_extremes-detection_q02__00c0e78bec54", "report_id": "satellite_satellite-lai-fapar_extremes-detection_q02", "dataset_id": "satellite-lai-fapar", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extremes-detection_q02", "aspect_base": "extremes-detection", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Impact of Drought on the Leaf Area Index in Yunnan Province, China > ℹ️ If you want to know more > References", "title": "Impact of Drought on the Leaf Area Index in Yunnan Province, China", "chunk_index": 12, "token_count": 495, "text_raw": "[[1]](https://doi.org/10.1038/s41598-024-58068-w) Zhang, Y., Gu, T., He, S., Cheng, F., Wang, J., Ye, H., Zhang, Y., Su, H., & Li, Q. (2024). Extreme drought along the tropic of cancer (Yunnan section) and its impact on vegetation. Sci Rep 14, 7508.\n\n[[2]](https://gadm.org) GADM. (2023). GADM database of global administrative areas (version 4.1). Data accessed via UC Davis GeoData repository.\n\n[[3]](https://doi.org/10.1016/j.chnaes.2013.09.004) Yu, W., Shao, M., Ren, M., Zhou, H., Jiang, Z., & Li, D. (2013). Analysis on spatial and temporal characteristics of drought in Yunnan Province. Acta Ecologica Sinica, 33(6), 317–324.\n\n[[4]](https://doi.org/10.1007/s11629-016-3971-x) Kim, K., Wang, M.-c., Ranjitkar, S., Liu, S.-h., Xu, J.-c., & Zomer, R. J. (2017). Using leaf area index (LAI) to assess vegetation response to drought in Yunnan province of China. J. Mt. Sci. 14, 1863–1872.\n\n[[5]](https://doi.org/10.1016/j.scitotenv.2016.09.026) Wang, J., Peng, J., Zhao, M., Liu, Y., & Chen, Y. (2017). Significant trade-off for the impact of Grain-for-Green Programme on ecosystem services in North-western Yunnan, China. Science of The Total Environment, 574, 57–64.", "text_with_prefix": "EQC Quality Assessment: \"Impact of Drought on the Leaf Area Index in Yunnan Province, China\"\nDataset: satellite-lai-fapar [CDS]\nAspect: extremes-detection_q02 | Category: Satellite_ECVs\nSection: Impact of Drought on the Leaf Area Index in Yunnan Province, China > ℹ️ If you want to know more > References\n---\n[[1]](https://doi.org/10.1038/s41598-024-58068-w) Zhang, Y., Gu, T., He, S., Cheng, F., Wang, J., Ye, H., Zhang, Y., Su, H., & Li, Q. (2024). Extreme drought along the tropic of cancer (Yunnan section) and its impact on vegetation. Sci Rep 14, 7508.\n\n[[2]](https://gadm.org) GADM. (2023). GADM database of global administrative areas (version 4.1). Data accessed via UC Davis GeoData repository.\n\n[[3]](https://doi.org/10.1016/j.chnaes.2013.09.004) Yu, W., Shao, M., Ren, M., Zhou, H., Jiang, Z., & Li, D. (2013). Analysis on spatial and temporal characteristics of drought in Yunnan Province. Acta Ecologica Sinica, 33(6), 317–324.\n\n[[4]](https://doi.org/10.1007/s11629-016-3971-x) Kim, K., Wang, M.-c., Ranjitkar, S., Liu, S.-h., Xu, J.-c., & Zomer, R. J. (2017). Using leaf area index (LAI) to assess vegetation response to drought in Yunnan province of China. J. Mt. Sci. 14, 1863–1872.\n\n[[5]](https://doi.org/10.1016/j.scitotenv.2016.09.026) Wang, J., Peng, J., Zhao, M., Liu, Y., & Chen, Y. (2017). Significant trade-off for the impact of Grain-for-Green Programme on ecosystem services in North-western Yunnan, China. Science of The Total Environment, 574, 57–64."} {"chunk_id": "satellite_satellite-land-cover_completeness_q03__549a0a3b4e11", "report_id": "satellite_satellite-land-cover_completeness_q03", "dataset_id": "satellite-land-cover", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q03", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Land Cover trend assessment for Spatial Planning and Land Management > Quality assessment question", "title": "Satellite Land Cover trend assessment for Spatial Planning and Land Management", "chunk_index": 0, "token_count": 292, "text_raw": "* **Is the dataset suitable for the analysis of afforestation/deforestation trends in the Iberian Peninsula?**\n\nLand Cover data is an invaluable resource for a wide range of fields, from climate change research to forest management. Land Cover products that provide historical timelines enable scientists, policymakers, and planners to understand and analyse the transformation of land cover over recent decades ([EUROSTAT,2022](https://ec.europa.eu/eurostat/statistics-explained/index.php?title=Forests,_forestry_and_logging); [New EU Forest Strategy for 2030](https://commission.europa.eu/document/cf3294e1-8358-4c93-8de4-3e1503b95201_en)).\n\nThis notebook will access the ***Land cover classification gridded maps from 1992 to present derived from satellite observations*** (henceforth, LC) data from the Climate Data Store (CDS) of the Copernicus Climate Change Service (C3S), and analyse the spatial patterns of a specific LC type over a given Area of Interest (AoI) and time.", "text_with_prefix": "EQC Quality Assessment: \"Satellite Land Cover trend assessment for Spatial Planning and Land Management\"\nDataset: satellite-land-cover [CDS]\nAspect: completeness_q03 | Category: Satellite_ECVs\nSection: Satellite Land Cover trend assessment for Spatial Planning and Land Management > Quality assessment question\n---\n* **Is the dataset suitable for the analysis of afforestation/deforestation trends in the Iberian Peninsula?**\n\nLand Cover data is an invaluable resource for a wide range of fields, from climate change research to forest management. Land Cover products that provide historical timelines enable scientists, policymakers, and planners to understand and analyse the transformation of land cover over recent decades ([EUROSTAT,2022](https://ec.europa.eu/eurostat/statistics-explained/index.php?title=Forests,_forestry_and_logging); [New EU Forest Strategy for 2030](https://commission.europa.eu/document/cf3294e1-8358-4c93-8de4-3e1503b95201_en)).\n\nThis notebook will access the ***Land cover classification gridded maps from 1992 to present derived from satellite observations*** (henceforth, LC) data from the Climate Data Store (CDS) of the Copernicus Climate Change Service (C3S), and analyse the spatial patterns of a specific LC type over a given Area of Interest (AoI) and time."} {"chunk_id": "satellite_satellite-land-cover_completeness_q03__04076f38703a", "report_id": "satellite_satellite-land-cover_completeness_q03", "dataset_id": "satellite-land-cover", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q03", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Land Cover trend assessment for Spatial Planning and Land Management > Quality assessment statement", "title": "Satellite Land Cover trend assessment for Spatial Planning and Land Management", "chunk_index": 1, "token_count": 241, "text_raw": "These are the key outcomes of this assessment\n\nThe dataset maintains strong temporal continuity, with annual updates ensuring a smooth and reliable representation of forest land cover changes over time. While breakpoints were identified, they generally did not indicate major disruptions, reinforcing the dataset’s stability for long-term trend analysis.\n\nThe presence of breakpoints does not necessarily indicate abrupt landscape shifts but rather highlights the sensitivity of detection methods to gradual changes. This suggests that while breakpoints can help refine analysis, their impact on overall trends remains limited, emphasising the dataset’s resilience to minor variations.\n\nFor the specific land cover type analysed, the dataset exhibits a consistent ability to capture underlying trends. The similarity in results across segmented and total trends suggests that the data structure is well-calibrated, minimising distortions that could arise from classification inconsistencies or methodological biases.\n\n```\n\n![Forest_Map_Series.png](attachment:Forest_Map_Series.png)", "text_with_prefix": "EQC Quality Assessment: \"Satellite Land Cover trend assessment for Spatial Planning and Land Management\"\nDataset: satellite-land-cover [CDS]\nAspect: completeness_q03 | Category: Satellite_ECVs\nSection: Satellite Land Cover trend assessment for Spatial Planning and Land Management > Quality assessment statement\n---\nThese are the key outcomes of this assessment\n\nThe dataset maintains strong temporal continuity, with annual updates ensuring a smooth and reliable representation of forest land cover changes over time. While breakpoints were identified, they generally did not indicate major disruptions, reinforcing the dataset’s stability for long-term trend analysis.\n\nThe presence of breakpoints does not necessarily indicate abrupt landscape shifts but rather highlights the sensitivity of detection methods to gradual changes. This suggests that while breakpoints can help refine analysis, their impact on overall trends remains limited, emphasising the dataset’s resilience to minor variations.\n\nFor the specific land cover type analysed, the dataset exhibits a consistent ability to capture underlying trends. The similarity in results across segmented and total trends suggests that the data structure is well-calibrated, minimising distortions that could arise from classification inconsistencies or methodological biases.\n\n```\n\n![Forest_Map_Series.png](attachment:Forest_Map_Series.png)"} {"chunk_id": "satellite_satellite-land-cover_completeness_q03__54567dd939fb", "report_id": "satellite_satellite-land-cover_completeness_q03", "dataset_id": "satellite-land-cover", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q03", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Land Cover trend assessment for Spatial Planning and Land Management > Methodology", "title": "Satellite Land Cover trend assessment for Spatial Planning and Land Management", "chunk_index": 2, "token_count": 111, "text_raw": "**This Use Case comprises the following steps:**\n\n**[](code-section-1)**\n\n**[](code-section-2)**\n\n**[](code-section-3)**\n\n**[](code-section-4)**\n\n**[](code-section-5)**", "text_with_prefix": "EQC Quality Assessment: \"Satellite Land Cover trend assessment for Spatial Planning and Land Management\"\nDataset: satellite-land-cover [CDS]\nAspect: completeness_q03 | Category: Satellite_ECVs\nSection: Satellite Land Cover trend assessment for Spatial Planning and Land Management > Methodology\n---\n**This Use Case comprises the following steps:**\n\n**[](code-section-1)**\n\n**[](code-section-2)**\n\n**[](code-section-3)**\n\n**[](code-section-4)**\n\n**[](code-section-5)**"} {"chunk_id": "satellite_satellite-land-cover_completeness_q03__48c1d910b09f", "report_id": "satellite_satellite-land-cover_completeness_q03", "dataset_id": "satellite-land-cover", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q03", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 1. Define the AoI, search and download LC data", "title": "Satellite Land Cover trend assessment for Spatial Planning and Land Management", "chunk_index": 3, "token_count": 112, "text_raw": "Before we begin we must prepare our environment. This includes installing the Application Programming Interface (API) of the CDS, and importing the various python libraries that we will need.", "text_with_prefix": "EQC Quality Assessment: \"Satellite Land Cover trend assessment for Spatial Planning and Land Management\"\nDataset: satellite-land-cover [CDS]\nAspect: completeness_q03 | Category: Satellite_ECVs\nSection: Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 1. Define the AoI, search and download LC data\n---\nBefore we begin we must prepare our environment. This includes installing the Application Programming Interface (API) of the CDS, and importing the various python libraries that we will need."} {"chunk_id": "satellite_satellite-land-cover_completeness_q03__b25f89bf9201", "report_id": "satellite_satellite-land-cover_completeness_q03", "dataset_id": "satellite-land-cover", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q03", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 1. Define the AoI, search and download LC data > Install CDS API", "title": "Satellite Land Cover trend assessment for Spatial Planning and Land Management", "chunk_index": 4, "token_count": 967, "text_raw": "To install the CDS API, run the following command. We use an exclamation mark to pass the command to the shell (not to the Python interpreter).\nIf you already have the CDS API installed, you can skip or comment this step.\n\n```text\nDefaulting to user installation because normal site-packages is not writeable\nRequirement already satisfied: cdsapi in /data/common/miniforge3/envs/wp5/lib/python3.12/site-packages (0.7.7)\nRequirement already satisfied: ecmwf-datastores-client>=0.4.0 in /data/common/miniforge3/envs/wp5/lib/python3.12/site-packages (from cdsapi) (0.4.1)\nRequirement already satisfied: requests>=2.5.0 in /data/common/miniforge3/envs/wp5/lib/python3.12/site-packages (from cdsapi) (2.32.5)\nRequirement already satisfied: tqdm in /data/common/miniforge3/envs/wp5/lib/python3.12/site-packages (from cdsapi) (4.67.1)\nRequirement already satisfied: attrs in /data/common/miniforge3/envs/wp5/lib/python3.12/site-packages (from ecmwf-datastores-client>=0.4.0->cdsapi) (25.4.0)\nRequirement already satisfied: multiurl>=0.3.7 in /data/common/miniforge3/envs/wp5/lib/python3.12/site-packages (from ecmwf-datastores-client>=0.4.0->cdsapi) (0.3.7)\nRequirement already satisfied: typing-extensions in /data/common/miniforge3/envs/wp5/lib/python3.12/site-packages (from ecmwf-datastores-client>=0.4.0->cdsapi) (4.15.0)\nRequirement already satisfied: pytz in /data/common/miniforge3/envs/wp5/lib/python3.12/site-packages (from multiurl>=0.3.7->ecmwf-datastores-client>=0.4.0->cdsapi) (2025.2)\nRequirement already satisfied: python-dateutil in /data/common/miniforge3/envs/wp5/lib/python3.12/site-packages (from multiurl>=0.3.7->ecmwf-datastores-client>=0.4.0->cdsapi) (2.9.0.post0)\nRequirement already satisfied: charset_normalizer<4,>=2 in /data/common/miniforge3/envs/wp5/lib/python3.12/site-packages (from requests>=2.5.0->cdsapi) (3.4.4)\nRequirement already satisfied: idna<4,>=2.5 in /data/common/miniforge3/envs/wp5/lib/python3.12/site-packages (from requests>=2.5.0->cdsapi) (3.11)\nRequirement already satisfied: urllib3<3,>=1.21.1 in /data/common/miniforge3/envs/wp5/lib/python3.12/site-packages (from requests>=2.5.0->cdsapi) (2.6.0)\nRequirement already satisfied: certifi>=2017.4.17 in /data/common/miniforge3/envs/wp5/lib/python3.12/site-packages (from requests>=2.5.0->cdsapi) (2025.11.12)\nRequirement already satisfied: six>=1.5 in /data/common/miniforge3/envs/wp5/lib/python3.12/site-packages (from python-dateutil->multiurl>=0.3.7->ecmwf-datastores-client>=0.4.0->cdsapi) (1.17.0)\n```\n\n##### Import all the libraries/packages\n\nWe will be working with data in NetCDF format. To best handle this type of data we will use libraries for working with multidimensional arrays, in particular Xarray. \nWe will also need libraries for plotting and viewing data.\n\nImport Standard Libraries\nImport Numerical & Statistical Libraries\nimport ruptures as rpt\nImport Geospatial Libraries\nImport Visualization Libraries\nImport External Tools\nSet Matplotlib Style\nSet the CDSAPI location\n\n##### Data Overview", "text_with_prefix": "EQC Quality Assessment: \"Satellite Land Cover trend assessment for Spatial Planning and Land Management\"\nDataset: satellite-land-cover [CDS]\nAspect: completeness_q03 | Category: Satellite_ECVs\nSection: Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 1. Define the AoI, search and download LC data > Install CDS API\n---\nTo install the CDS API, run the following command. We use an exclamation mark to pass the command to the shell (not to the Python interpreter).\nIf you already have the CDS API installed, you can skip or comment this step.\n\n```text\nDefaulting to user installation because normal site-packages is not writeable\nRequirement already satisfied: cdsapi in /data/common/miniforge3/envs/wp5/lib/python3.12/site-packages (0.7.7)\nRequirement already satisfied: ecmwf-datastores-client>=0.4.0 in /data/common/miniforge3/envs/wp5/lib/python3.12/site-packages (from cdsapi) (0.4.1)\nRequirement already satisfied: requests>=2.5.0 in /data/common/miniforge3/envs/wp5/lib/python3.12/site-packages (from cdsapi) (2.32.5)\nRequirement already satisfied: tqdm in /data/common/miniforge3/envs/wp5/lib/python3.12/site-packages (from cdsapi) (4.67.1)\nRequirement already satisfied: attrs in /data/common/miniforge3/envs/wp5/lib/python3.12/site-packages (from ecmwf-datastores-client>=0.4.0->cdsapi) (25.4.0)\nRequirement already satisfied: multiurl>=0.3.7 in /data/common/miniforge3/envs/wp5/lib/python3.12/site-packages (from ecmwf-datastores-client>=0.4.0->cdsapi) (0.3.7)\nRequirement already satisfied: typing-extensions in /data/common/miniforge3/envs/wp5/lib/python3.12/site-packages (from ecmwf-datastores-client>=0.4.0->cdsapi) (4.15.0)\nRequirement already satisfied: pytz in /data/common/miniforge3/envs/wp5/lib/python3.12/site-packages (from multiurl>=0.3.7->ecmwf-datastores-client>=0.4.0->cdsapi) (2025.2)\nRequirement already satisfied: python-dateutil in /data/common/miniforge3/envs/wp5/lib/python3.12/site-packages (from multiurl>=0.3.7->ecmwf-datastores-client>=0.4.0->cdsapi) (2.9.0.post0)\nRequirement already satisfied: charset_normalizer<4,>=2 in /data/common/miniforge3/envs/wp5/lib/python3.12/site-packages (from requests>=2.5.0->cdsapi) (3.4.4)\nRequirement already satisfied: idna<4,>=2.5 in /data/common/miniforge3/envs/wp5/lib/python3.12/site-packages (from requests>=2.5.0->cdsapi) (3.11)\nRequirement already satisfied: urllib3<3,>=1.21.1 in /data/common/miniforge3/envs/wp5/lib/python3.12/site-packages (from requests>=2.5.0->cdsapi) (2.6.0)\nRequirement already satisfied: certifi>=2017.4.17 in /data/common/miniforge3/envs/wp5/lib/python3.12/site-packages (from requests>=2.5.0->cdsapi) (2025.11.12)\nRequirement already satisfied: six>=1.5 in /data/common/miniforge3/envs/wp5/lib/python3.12/site-packages (from python-dateutil->multiurl>=0.3.7->ecmwf-datastores-client>=0.4.0->cdsapi) (1.17.0)\n```\n\n##### Import all the libraries/packages\n\nWe will be working with data in NetCDF format. To best handle this type of data we will use libraries for working with multidimensional arrays, in particular Xarray. \nWe will also need libraries for plotting and viewing data.\n\nImport Standard Libraries\nImport Numerical & Statistical Libraries\nimport ruptures as rpt\nImport Geospatial Libraries\nImport Visualization Libraries\nImport External Tools\nSet Matplotlib Style\nSet the CDSAPI location\n\n##### Data Overview"} {"chunk_id": "satellite_satellite-land-cover_completeness_q03__5ec8d7ded53e", "report_id": "satellite_satellite-land-cover_completeness_q03", "dataset_id": "satellite-land-cover", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q03", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 1. Define the AoI, search and download LC data > Install CDS API", "title": "Satellite Land Cover trend assessment for Spatial Planning and Land Management", "chunk_index": 5, "token_count": 775, "text_raw": "Libraries\nImport Numerical & Statistical Libraries\nimport ruptures as rpt\nImport Geospatial Libraries\nImport Visualization Libraries\nImport External Tools\nSet Matplotlib Style\nSet the CDSAPI location\n\n##### Data Overview\n\nTo search for data, visit the CDS website: http://cds.climate.copernicus.eu Here you can search for 'Satellite observations' using the search bar. The data we need for this tutorial is the ***Land cover classification gridded maps from 1992 to present derived from satellite observations***. This catalogue entry provides global Land Cover Classification (LCC) maps with a very high spatial resolution, with a L4 processing level, on an annual basis with a one-year delay, following the [Global Climate Observing System (GCOS) convention requirements](https://library.wmo.int/records/item/58111-the-2022-gcos-ecvs-requirements-gcos-245). LULC maps correspond to a global classification scheme, encompassing 22 classes.\n\nThe dataset consists of 2 versions (v2.0.7 produced by the European Space Agency (ESA) Climate Change Initiative (CCI) and v2.1.1 produced by Copernicus Climate Change Service (C3S)).\n\nData specifications for this use case:\n* **Years:** 1992 to 2022\n* **Version:** v2.0.7 before 1992 and v2.1.1 after 2016\n* **Format:** Zip files\n\nAt the end of the download form, select “**Show API request**”. This will reveal a block of code, which you can simply copy and paste into a cell of your Jupyter Notebook. Having copied the API request, running it will retrieve and download the data you requested into your local directory. However, before you run it, the **terms and conditions** of this particular dataset need to have been accepted directly at the CDS website. The option to view and accept these conditions is given at the end of the download form, just above the “**Show API request**” option. In addition, it is also useful to define the time period and AoI parameters and edit the request accordingly, as exemplified in the cells below.\n\nYears to download\nList of requests to retrieve data\n\nDownload the dataset\n\n```text\n100%|██████████| 32/32 [00:00<00:00, 88.34it/s] \n/data/common/miniforge3/envs/wp5/lib/python3.12/site-packages/earthkit/data/readers/netcdf/fieldlist.py:202: FutureWarning: In a future version of xarray the default value for data_vars will change from data_vars='all' to data_vars=None. This is likely to lead to different results when multiple datasets have matching variables with overlapping values. To opt in to new defaults and get rid of these warnings now use `set_options(use_new_combine_kwarg_defaults=True) or set data_vars explicitly.\n return xr.open_mfdataset(\n/data/wp5/.tmp/ipykernel_11804/1703755849.py:7: DeprecationWarning: dropping variables using `drop` is deprecated; use drop_vars.\n ds = ds.assign_coords(year=ds[\"time\"].dt.year).swap_dims(time=\"year\").drop(\"time\")\n```\n\nInspect the database", "text_with_prefix": "EQC Quality Assessment: \"Satellite Land Cover trend assessment for Spatial Planning and Land Management\"\nDataset: satellite-land-cover [CDS]\nAspect: completeness_q03 | Category: Satellite_ECVs\nSection: Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 1. Define the AoI, search and download LC data > Install CDS API\n---\nLibraries\nImport Numerical & Statistical Libraries\nimport ruptures as rpt\nImport Geospatial Libraries\nImport Visualization Libraries\nImport External Tools\nSet Matplotlib Style\nSet the CDSAPI location\n\n##### Data Overview\n\nTo search for data, visit the CDS website: http://cds.climate.copernicus.eu Here you can search for 'Satellite observations' using the search bar. The data we need for this tutorial is the ***Land cover classification gridded maps from 1992 to present derived from satellite observations***. This catalogue entry provides global Land Cover Classification (LCC) maps with a very high spatial resolution, with a L4 processing level, on an annual basis with a one-year delay, following the [Global Climate Observing System (GCOS) convention requirements](https://library.wmo.int/records/item/58111-the-2022-gcos-ecvs-requirements-gcos-245). LULC maps correspond to a global classification scheme, encompassing 22 classes.\n\nThe dataset consists of 2 versions (v2.0.7 produced by the European Space Agency (ESA) Climate Change Initiative (CCI) and v2.1.1 produced by Copernicus Climate Change Service (C3S)).\n\nData specifications for this use case:\n* **Years:** 1992 to 2022\n* **Version:** v2.0.7 before 1992 and v2.1.1 after 2016\n* **Format:** Zip files\n\nAt the end of the download form, select “**Show API request**”. This will reveal a block of code, which you can simply copy and paste into a cell of your Jupyter Notebook. Having copied the API request, running it will retrieve and download the data you requested into your local directory. However, before you run it, the **terms and conditions** of this particular dataset need to have been accepted directly at the CDS website. The option to view and accept these conditions is given at the end of the download form, just above the “**Show API request**” option. In addition, it is also useful to define the time period and AoI parameters and edit the request accordingly, as exemplified in the cells below.\n\nYears to download\nList of requests to retrieve data\n\nDownload the dataset\n\n```text\n100%|██████████| 32/32 [00:00<00:00, 88.34it/s] \n/data/common/miniforge3/envs/wp5/lib/python3.12/site-packages/earthkit/data/readers/netcdf/fieldlist.py:202: FutureWarning: In a future version of xarray the default value for data_vars will change from data_vars='all' to data_vars=None. This is likely to lead to different results when multiple datasets have matching variables with overlapping values. To opt in to new defaults and get rid of these warnings now use `set_options(use_new_combine_kwarg_defaults=True) or set data_vars explicitly.\n return xr.open_mfdataset(\n/data/wp5/.tmp/ipykernel_11804/1703755849.py:7: DeprecationWarning: dropping variables using `drop` is deprecated; use drop_vars.\n ds = ds.assign_coords(year=ds[\"time\"].dt.year).swap_dims(time=\"year\").drop(\"time\")\n```\n\nInspect the database"} {"chunk_id": "satellite_satellite-land-cover_completeness_q03__0085afed905c", "report_id": "satellite_satellite-land-cover_completeness_q03", "dataset_id": "satellite-land-cover", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q03", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 1. Define the AoI, search and download LC data > Install CDS API", "title": "Satellite Land Cover trend assessment for Spatial Planning and Land Management", "chunk_index": 6, "token_count": 792, "text_raw": "_vars.\n ds = ds.assign_coords(year=ds[\"time\"].dt.year).swap_dims(time=\"year\").drop(\"time\")\n```\n\nInspect the database\n\n```text\n Size: 7GB\nDimensions: (year: 31, latitude: 3600, longitude: 5040, bounds: 2)\nCoordinates:\n * year (year) int64 248B 1992 1993 1994 ... 2020 2021 2022\n * latitude (latitude) float64 29kB 45.0 45.0 44.99 ... 35.0 35.0\n * longitude (longitude) float64 40kB -9.999 -9.996 ... 3.996 3.999\n lat_bounds (latitude, bounds, longitude) float64 290MB dask.array\n lon_bounds (longitude, bounds) float64 81kB dask.array\n time_bounds (year, bounds, longitude) datetime64[ns] 2MB dask.array\nDimensions without coordinates: bounds\nData variables:\n lccs_class (year, latitude, longitude) uint8 562MB dask.array\n processed_flag (year, latitude, longitude) float32 2GB dask.array\n current_pixel_state (year, latitude, longitude) float32 2GB dask.array\n observation_count (year, latitude, longitude) uint16 1GB dask.array\n change_count (year, latitude, longitude) uint8 562MB dask.array\n crs (year, longitude) int32 625kB dask.array\nAttributes: (12/38)\n id: ESACCI-LC-L4-LCCS-Map-300m-P1Y-1992-v2.0.7cds\n title: Land Cover Map of ESA CCI brokered by CDS\n summary: This dataset characterizes the land cover of ...\n type: ESACCI-LC-L4-LCCS-Map-300m-P1Y\n project: Climate Change Initiative - European Space Ag...\n references: http://www.esa-landcover-cci.org/\n ... ...\n geospatial_lon_max: 180\n spatial_resolution: 300m\n geospatial_lat_units: degrees_north\n geospatial_lat_resolution: 0.002778\n geospatial_lon_units: degrees_east\n geospatial_lon_resolution: 0.002778\n```\n\n(code-section-2)=", "text_with_prefix": "EQC Quality Assessment: \"Satellite Land Cover trend assessment for Spatial Planning and Land Management\"\nDataset: satellite-land-cover [CDS]\nAspect: completeness_q03 | Category: Satellite_ECVs\nSection: Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 1. Define the AoI, search and download LC data > Install CDS API\n---\n_vars.\n ds = ds.assign_coords(year=ds[\"time\"].dt.year).swap_dims(time=\"year\").drop(\"time\")\n```\n\nInspect the database\n\n```text\n Size: 7GB\nDimensions: (year: 31, latitude: 3600, longitude: 5040, bounds: 2)\nCoordinates:\n * year (year) int64 248B 1992 1993 1994 ... 2020 2021 2022\n * latitude (latitude) float64 29kB 45.0 45.0 44.99 ... 35.0 35.0\n * longitude (longitude) float64 40kB -9.999 -9.996 ... 3.996 3.999\n lat_bounds (latitude, bounds, longitude) float64 290MB dask.array\n lon_bounds (longitude, bounds) float64 81kB dask.array\n time_bounds (year, bounds, longitude) datetime64[ns] 2MB dask.array\nDimensions without coordinates: bounds\nData variables:\n lccs_class (year, latitude, longitude) uint8 562MB dask.array\n processed_flag (year, latitude, longitude) float32 2GB dask.array\n current_pixel_state (year, latitude, longitude) float32 2GB dask.array\n observation_count (year, latitude, longitude) uint16 1GB dask.array\n change_count (year, latitude, longitude) uint8 562MB dask.array\n crs (year, longitude) int32 625kB dask.array\nAttributes: (12/38)\n id: ESACCI-LC-L4-LCCS-Map-300m-P1Y-1992-v2.0.7cds\n title: Land Cover Map of ESA CCI brokered by CDS\n summary: This dataset characterizes the land cover of ...\n type: ESACCI-LC-L4-LCCS-Map-300m-P1Y\n project: Climate Change Initiative - European Space Ag...\n references: http://www.esa-landcover-cci.org/\n ... ...\n geospatial_lon_max: 180\n spatial_resolution: 300m\n geospatial_lat_units: degrees_north\n geospatial_lat_resolution: 0.002778\n geospatial_lon_units: degrees_east\n geospatial_lon_resolution: 0.002778\n```\n\n(code-section-2)="} {"chunk_id": "satellite_satellite-land-cover_completeness_q03__9b04aad35d3a", "report_id": "satellite_satellite-land-cover_completeness_q03", "dataset_id": "satellite-land-cover", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q03", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 2. Inspect and view data for the defined AoI (Iberian Peninsula) > Compute Land Cover classes area coverage for each NUTS 2 region", "title": "Satellite Land Cover trend assessment for Spatial Planning and Land Management", "chunk_index": 7, "token_count": 236, "text_raw": "To identify changes in LC patterns, by regions, NUTS 2 will be used, providing the information regarding the main regions of Iberian Peninsula.\n\nThe NUTS are a hierarchical system divided into 3 levels (https://ec.europa.eu/eurostat/web/nuts). NUTS 1 correspond to major socio-economic regions, NUTS 2 correspond to basic regions for the application of regional policies, and NUTS 3 correspond to small regions for specific diagnoses. Additionally a NUTS 0 level, usually co-incident with national boundaries is also available. The NUTS legislation is periodically amended; therefore multiple years are available for download.", "text_with_prefix": "EQC Quality Assessment: \"Satellite Land Cover trend assessment for Spatial Planning and Land Management\"\nDataset: satellite-land-cover [CDS]\nAspect: completeness_q03 | Category: Satellite_ECVs\nSection: Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 2. Inspect and view data for the defined AoI (Iberian Peninsula) > Compute Land Cover classes area coverage for each NUTS 2 region\n---\nTo identify changes in LC patterns, by regions, NUTS 2 will be used, providing the information regarding the main regions of Iberian Peninsula.\n\nThe NUTS are a hierarchical system divided into 3 levels (https://ec.europa.eu/eurostat/web/nuts). NUTS 1 correspond to major socio-economic regions, NUTS 2 correspond to basic regions for the application of regional policies, and NUTS 3 correspond to small regions for specific diagnoses. Additionally a NUTS 0 level, usually co-incident with national boundaries is also available. The NUTS legislation is periodically amended; therefore multiple years are available for download."} {"chunk_id": "satellite_satellite-land-cover_completeness_q03__d5833c1feeb2", "report_id": "satellite_satellite-land-cover_completeness_q03", "dataset_id": "satellite-land-cover", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q03", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 2. Inspect and view data for the defined AoI (Iberian Peninsula) > Mask regions", "title": "Satellite Land Cover trend assessment for Spatial Planning and Land Management", "chunk_index": 8, "token_count": 171, "text_raw": "First, we need to establish the geometry of the NUTS region (level 2) in order to make the corresponding statistics.\n\nConfigure Dask\nDefine CRS and bounding box for Iberian Peninsula\nLoad and filter GeoDataFrame\nEnsure dataset CRS is set\nEnsure dataset coordinates overlap with the filtered regions\nSubset dataset to valid ranges\nCheck subset dataset dimensions\nCreate the regionmask\nCreate a 2D mask", "text_with_prefix": "EQC Quality Assessment: \"Satellite Land Cover trend assessment for Spatial Planning and Land Management\"\nDataset: satellite-land-cover [CDS]\nAspect: completeness_q03 | Category: Satellite_ECVs\nSection: Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 2. Inspect and view data for the defined AoI (Iberian Peninsula) > Mask regions\n---\nFirst, we need to establish the geometry of the NUTS region (level 2) in order to make the corresponding statistics.\n\nConfigure Dask\nDefine CRS and bounding box for Iberian Peninsula\nLoad and filter GeoDataFrame\nEnsure dataset CRS is set\nEnsure dataset coordinates overlap with the filtered regions\nSubset dataset to valid ranges\nCheck subset dataset dimensions\nCreate the regionmask\nCreate a 2D mask"} {"chunk_id": "satellite_satellite-land-cover_completeness_q03__8999d5ced943", "report_id": "satellite_satellite-land-cover_completeness_q03", "dataset_id": "satellite-land-cover", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q03", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 2. Inspect and view data for the defined AoI (Iberian Peninsula) > Compute cell area", "title": "Satellite Land Cover trend assessment for Spatial Planning and Land Management", "chunk_index": 9, "token_count": 176, "text_raw": "Then, we can calculate the area of each pixel taking into consideration the curvature of the earth (i.e., weighted by the cosine of Latitude).\n\nScaling factor for conversion (constant longitude resolution)\nUse the latitude values directly from the dataset\nCalculate the difference between consecutive latitude values\nConvert latitude differences to kilometres\nCompute the grid cell area for each latitude\nAssign attributes to the grid cell area\nAdd the grid cell area as a coordinate to the dataset", "text_with_prefix": "EQC Quality Assessment: \"Satellite Land Cover trend assessment for Spatial Planning and Land Management\"\nDataset: satellite-land-cover [CDS]\nAspect: completeness_q03 | Category: Satellite_ECVs\nSection: Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 2. Inspect and view data for the defined AoI (Iberian Peninsula) > Compute cell area\n---\nThen, we can calculate the area of each pixel taking into consideration the curvature of the earth (i.e., weighted by the cosine of Latitude).\n\nScaling factor for conversion (constant longitude resolution)\nUse the latitude values directly from the dataset\nCalculate the difference between consecutive latitude values\nConvert latitude differences to kilometres\nCompute the grid cell area for each latitude\nAssign attributes to the grid cell area\nAdd the grid cell area as a coordinate to the dataset"} {"chunk_id": "satellite_satellite-land-cover_completeness_q03__f57562b4671b", "report_id": "satellite_satellite-land-cover_completeness_q03", "dataset_id": "satellite-land-cover", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q03", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 2. Inspect and view data for the defined AoI (Iberian Peninsula) > Select Forest Classes and Prepare Dataset", "title": "Satellite Land Cover trend assessment for Spatial Planning and Land Management", "chunk_index": 10, "token_count": 490, "text_raw": "To analyse forest cover, we selected all land cover (LC) classes corresponding to forest types from the LC dataset. The classes are described in the [(C3S, LC - Product User Guide and Specification (2024)](https://dast.copernicus-climate.eu/documents/satellite-land-cover/WP2-FDDP-LC-2021-2022-SENTINEL3-300m-v2.1.1_PUGS_v1.1_final.pdf).\n\nThe following class codes were included:\n\n- **50** – Tree cover, broadleaved, evergreen \n- **60** – Tree cover, broadleaved, deciduous, closed \n- **61** – Tree cover, broadleaved, deciduous, open \n- **62** – Tree cover, broadleaved, deciduous \n- **70** – Tree cover, needleleaved, evergreen \n- **71** – Tree cover, needleleaved, evergreen, closed \n- **72** – Tree cover, needleleaved, evergreen, open \n- **80** – Tree cover, needleleaved, deciduous \n- **81** – Tree cover, needleleaved, deciduous, closed \n- **82** – Tree cover, needleleaved, deciduous, open \n- **90** – Tree cover, mixed leaf type \n- **100** – Mosaic tree and shrub / herbaceous cover \n- **160** – Tree cover, flooded, fresh or brackish water \n- **170** – Tree cover, flooded, saline water\n\nAll these classes were aggregated to represent **forest cover** in the analysis.\n\nDefine forest classes\nCreate a mask for forested areas\nMasked forest area using cell area\nStack latitude and longitude into a single dimension\nCreate a stacked mask and align dimensions\nAttach the stacked mask to the dataset\nGroup by regions and compute forested area for each year", "text_with_prefix": "EQC Quality Assessment: \"Satellite Land Cover trend assessment for Spatial Planning and Land Management\"\nDataset: satellite-land-cover [CDS]\nAspect: completeness_q03 | Category: Satellite_ECVs\nSection: Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 2. Inspect and view data for the defined AoI (Iberian Peninsula) > Select Forest Classes and Prepare Dataset\n---\nTo analyse forest cover, we selected all land cover (LC) classes corresponding to forest types from the LC dataset. The classes are described in the [(C3S, LC - Product User Guide and Specification (2024)](https://dast.copernicus-climate.eu/documents/satellite-land-cover/WP2-FDDP-LC-2021-2022-SENTINEL3-300m-v2.1.1_PUGS_v1.1_final.pdf).\n\nThe following class codes were included:\n\n- **50** – Tree cover, broadleaved, evergreen \n- **60** – Tree cover, broadleaved, deciduous, closed \n- **61** – Tree cover, broadleaved, deciduous, open \n- **62** – Tree cover, broadleaved, deciduous \n- **70** – Tree cover, needleleaved, evergreen \n- **71** – Tree cover, needleleaved, evergreen, closed \n- **72** – Tree cover, needleleaved, evergreen, open \n- **80** – Tree cover, needleleaved, deciduous \n- **81** – Tree cover, needleleaved, deciduous, closed \n- **82** – Tree cover, needleleaved, deciduous, open \n- **90** – Tree cover, mixed leaf type \n- **100** – Mosaic tree and shrub / herbaceous cover \n- **160** – Tree cover, flooded, fresh or brackish water \n- **170** – Tree cover, flooded, saline water\n\nAll these classes were aggregated to represent **forest cover** in the analysis.\n\nDefine forest classes\nCreate a mask for forested areas\nMasked forest area using cell area\nStack latitude and longitude into a single dimension\nCreate a stacked mask and align dimensions\nAttach the stacked mask to the dataset\nGroup by regions and compute forested area for each year"} {"chunk_id": "satellite_satellite-land-cover_completeness_q03__9d879bb943ea", "report_id": "satellite_satellite-land-cover_completeness_q03", "dataset_id": "satellite-land-cover", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q03", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 2. Inspect and view data for the defined AoI (Iberian Peninsula) > Compute forest area per region and year", "title": "Satellite Land Cover trend assessment for Spatial Planning and Land Management", "chunk_index": 11, "token_count": 129, "text_raw": "Ensure valid region indexing and naming\nSave growth data by year\nAdd results to the list\nConvert results to a DataFrame\nAdd geometry to results_df\nEnsure results_df is a GeoDataFrame", "text_with_prefix": "EQC Quality Assessment: \"Satellite Land Cover trend assessment for Spatial Planning and Land Management\"\nDataset: satellite-land-cover [CDS]\nAspect: completeness_q03 | Category: Satellite_ECVs\nSection: Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 2. Inspect and view data for the defined AoI (Iberian Peninsula) > Compute forest area per region and year\n---\nEnsure valid region indexing and naming\nSave growth data by year\nAdd results to the list\nConvert results to a DataFrame\nAdd geometry to results_df\nEnsure results_df is a GeoDataFrame"} {"chunk_id": "satellite_satellite-land-cover_completeness_q03__4c38cbca0488", "report_id": "satellite_satellite-land-cover_completeness_q03", "dataset_id": "satellite-land-cover", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q03", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 2. Inspect and view data for the defined AoI (Iberian Peninsula) > Map forest percentage coverage over-time by NUTS regions in the AoI", "title": "Satellite Land Cover trend assessment for Spatial Planning and Land Management", "chunk_index": 12, "token_count": 216, "text_raw": "Reproject the GeoDataFrame to a projected CRS (EPSG:3035 is a good choice for Europe)\nCalculate the total area of each region (in square kilometers)\nEnsure 'Year' is converted to string if needed\nCalculate Forest Percentage\nPivot the data to have years as columns\nExtract years for plotting\nNormalize color scale across all maps based on percentage coverage\nPlot settings\nPlot each year in a separate subplot\nHide unused subplots\nStep 8: Adjust layout\nAllocate space for the color bar\nCreate a single ScalarMappable object for the common colorbar", "text_with_prefix": "EQC Quality Assessment: \"Satellite Land Cover trend assessment for Spatial Planning and Land Management\"\nDataset: satellite-land-cover [CDS]\nAspect: completeness_q03 | Category: Satellite_ECVs\nSection: Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 2. Inspect and view data for the defined AoI (Iberian Peninsula) > Map forest percentage coverage over-time by NUTS regions in the AoI\n---\nReproject the GeoDataFrame to a projected CRS (EPSG:3035 is a good choice for Europe)\nCalculate the total area of each region (in square kilometers)\nEnsure 'Year' is converted to string if needed\nCalculate Forest Percentage\nPivot the data to have years as columns\nExtract years for plotting\nNormalize color scale across all maps based on percentage coverage\nPlot settings\nPlot each year in a separate subplot\nHide unused subplots\nStep 8: Adjust layout\nAllocate space for the color bar\nCreate a single ScalarMappable object for the common colorbar"} {"chunk_id": "satellite_satellite-land-cover_completeness_q03__9064bab08900", "report_id": "satellite_satellite-land-cover_completeness_q03", "dataset_id": "satellite-land-cover", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q03", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 2. Inspect and view data for the defined AoI (Iberian Peninsula) > Map Analysis", "title": "Satellite Land Cover trend assessment for Spatial Planning and Land Management", "chunk_index": 13, "token_count": 277, "text_raw": "- Over the 28-year period, all regions maintain a consistent area of forest.\n\n- Regions such as Principado de Asturias, Cantabria, País Vasco, Galicia, and Centro (PT) consistently exhibit the highest levels of forest coverage.\n\n- Northern Spain (Galicia, Asturias, Cantabria, País Vasco) is often affected by negative fluctuations, particularly in 1998–2000 and 2003–2004. The classification models may have struggled with complex topography and mixed vegetation types [(Zhao et al., 2023)](https://doi.org/10.3390/rs15092285).\n\n- Southern Portugal & Spain (Alentejo, Algarve, Andalucía), unlike northern Spain, experiences more positive fluctuations in area, particularly in 1998-2004.\n\n- From 2006–2007 onward the maps show more stable estimates.", "text_with_prefix": "EQC Quality Assessment: \"Satellite Land Cover trend assessment for Spatial Planning and Land Management\"\nDataset: satellite-land-cover [CDS]\nAspect: completeness_q03 | Category: Satellite_ECVs\nSection: Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 2. Inspect and view data for the defined AoI (Iberian Peninsula) > Map Analysis\n---\n- Over the 28-year period, all regions maintain a consistent area of forest.\n\n- Regions such as Principado de Asturias, Cantabria, País Vasco, Galicia, and Centro (PT) consistently exhibit the highest levels of forest coverage.\n\n- Northern Spain (Galicia, Asturias, Cantabria, País Vasco) is often affected by negative fluctuations, particularly in 1998–2000 and 2003–2004. The classification models may have struggled with complex topography and mixed vegetation types [(Zhao et al., 2023)](https://doi.org/10.3390/rs15092285).\n\n- Southern Portugal & Spain (Alentejo, Algarve, Andalucía), unlike northern Spain, experiences more positive fluctuations in area, particularly in 1998-2004.\n\n- From 2006–2007 onward the maps show more stable estimates."} {"chunk_id": "satellite_satellite-land-cover_completeness_q03__98beb8572da4", "report_id": "satellite_satellite-land-cover_completeness_q03", "dataset_id": "satellite-land-cover", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q03", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 2. Inspect and view data for the defined AoI (Iberian Peninsula) > Potential Drivers and Methodological Considerations in Afforestation/Deforestation Trends in the Iberian Peninsula (1992–2022)", "title": "Satellite Land Cover trend assessment for Spatial Planning and Land Management", "chunk_index": 14, "token_count": 939, "text_raw": "Forest dynamics in the Iberian Peninsula over the last three decades have been shaped by a combination of **natural events**, **human activities**, and **policy interventions**. Before analysing trends in a land use and land cover (LULC) time series, it is crucial to account for abrupt shifts that may distort trend estimates. These shifts, often referred to as breakpoints, can result from:\n\n**Natural Events**\n\n- Wildfires: The Iberian Peninsula experiences frequent and intense wildfires, particularly in **Galicia, Catalonia, and central Portugal**. Major fire seasons occurred in **2003, 2005, and 2017**, causing the destruction of vast areas of forest [(Mato et al., 2014)](https://api.semanticscholar.org/CorpusID:265869582). Wildfires have become a key driver of land cover change, especially in fire-prone regions of Portugal and Spain [(Silva et al., 2011)](https://doi.org/10.1016/j.landurbplan.2011.03.001).\n\n- Droughts: Recurrent droughts have exacerbated forest degradation, weakening tree resilience and delaying post-fire recovery [(Gouveia et al., 2012)](https://nhess.copernicus.org/articles/12/3123/2012/). The severe droughts of **2005 and 2017** significantly reduced vegetation cover, affecting Mediterranean pine and oak forests [(Vidal-Macua et al., 2017)](https://doi.org/10.1016/j.foreco.2017.10.011).\n\n**Human Activities**\n\n- Farmland Abandonment: Rural depopulation since the 1990s has driven widespread farmland abandonment, leading to natural forest regeneration in regions like **northern Portugal and Galicia** [(Palmero-Iniesta et al., 2021)](https://api.semanticscholar.org/CorpusID:238829360).\n\n- Logging and Eucalyptus Plantations: Logging activities and the expansion of eucalyptus plantations in **Galicia and central Portugal** have caused periodic forest loss and regrowth [(Silva et al., 2011)](https://doi.org/10.1016/j.landurbplan.2011.03.001).\n\n**Policy and Economic Influences**\n\n- Afforestation and Reforestation: EU-funded afforestation programs under the **Common Agricultural Policy (CAP)** promoted the planting of trees on abandoned farmland, particularly in **Castilla y León, Extremadura, and parts of Portugal**, leading to abrupt increases in forest cover [(Sevillano et al., 2018)](https://doi.org/10.1016/j.landusepol.2018.06.054).\n\n- Natura 2000 Network and Conservation Policies: The **Natura 2000 network** has contributed to forest conservation efforts, particularly in regions like **Sierra de Guadarrama (Madrid)** and **Peneda-Gerês National Park (Portugal)** [(Regos et al., 2015)](https://doi.org/10.1016/j.jag.2014.11.010).\n\n- Economic Downturn (2008–2013): The economic crisis reduced logging temporarily alleviating deforestation pressures (Mateus & Fernandes, 2014).\n\n**Methodology**\n\n- Certain shifts in forest land cover may be related to data collection and processing rather than actual forest changes. For instance, transitions between different satellite sensors can introduce artificial breaks in the time series [(Chelali et al., 2019)](https://doi.org/10.1109/JURSE.2019.8808967).\n\n- **The LC dataset is generated from a multi-sensor surface reflectance time series derived from several satellite missions. As sensor technology evolved, different instruments were used to produce consistent global composites.**\n\n
", "text_with_prefix": "EQC Quality Assessment: \"Satellite Land Cover trend assessment for Spatial Planning and Land Management\"\nDataset: satellite-land-cover [CDS]\nAspect: completeness_q03 | Category: Satellite_ECVs\nSection: Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 2. Inspect and view data for the defined AoI (Iberian Peninsula) > Potential Drivers and Methodological Considerations in Afforestation/Deforestation Trends in the Iberian Peninsula (1992–2022)\n---\nForest dynamics in the Iberian Peninsula over the last three decades have been shaped by a combination of **natural events**, **human activities**, and **policy interventions**. Before analysing trends in a land use and land cover (LULC) time series, it is crucial to account for abrupt shifts that may distort trend estimates. These shifts, often referred to as breakpoints, can result from:\n\n**Natural Events**\n\n- Wildfires: The Iberian Peninsula experiences frequent and intense wildfires, particularly in **Galicia, Catalonia, and central Portugal**. Major fire seasons occurred in **2003, 2005, and 2017**, causing the destruction of vast areas of forest [(Mato et al., 2014)](https://api.semanticscholar.org/CorpusID:265869582). Wildfires have become a key driver of land cover change, especially in fire-prone regions of Portugal and Spain [(Silva et al., 2011)](https://doi.org/10.1016/j.landurbplan.2011.03.001).\n\n- Droughts: Recurrent droughts have exacerbated forest degradation, weakening tree resilience and delaying post-fire recovery [(Gouveia et al., 2012)](https://nhess.copernicus.org/articles/12/3123/2012/). The severe droughts of **2005 and 2017** significantly reduced vegetation cover, affecting Mediterranean pine and oak forests [(Vidal-Macua et al., 2017)](https://doi.org/10.1016/j.foreco.2017.10.011).\n\n**Human Activities**\n\n- Farmland Abandonment: Rural depopulation since the 1990s has driven widespread farmland abandonment, leading to natural forest regeneration in regions like **northern Portugal and Galicia** [(Palmero-Iniesta et al., 2021)](https://api.semanticscholar.org/CorpusID:238829360).\n\n- Logging and Eucalyptus Plantations: Logging activities and the expansion of eucalyptus plantations in **Galicia and central Portugal** have caused periodic forest loss and regrowth [(Silva et al., 2011)](https://doi.org/10.1016/j.landurbplan.2011.03.001).\n\n**Policy and Economic Influences**\n\n- Afforestation and Reforestation: EU-funded afforestation programs under the **Common Agricultural Policy (CAP)** promoted the planting of trees on abandoned farmland, particularly in **Castilla y León, Extremadura, and parts of Portugal**, leading to abrupt increases in forest cover [(Sevillano et al., 2018)](https://doi.org/10.1016/j.landusepol.2018.06.054).\n\n- Natura 2000 Network and Conservation Policies: The **Natura 2000 network** has contributed to forest conservation efforts, particularly in regions like **Sierra de Guadarrama (Madrid)** and **Peneda-Gerês National Park (Portugal)** [(Regos et al., 2015)](https://doi.org/10.1016/j.jag.2014.11.010).\n\n- Economic Downturn (2008–2013): The economic crisis reduced logging temporarily alleviating deforestation pressures (Mateus & Fernandes, 2014).\n\n**Methodology**\n\n- Certain shifts in forest land cover may be related to data collection and processing rather than actual forest changes. For instance, transitions between different satellite sensors can introduce artificial breaks in the time series [(Chelali et al., 2019)](https://doi.org/10.1109/JURSE.2019.8808967).\n\n- **The LC dataset is generated from a multi-sensor surface reflectance time series derived from several satellite missions. As sensor technology evolved, different instruments were used to produce consistent global composites.**\n\n
"} {"chunk_id": "satellite_satellite-land-cover_completeness_q03__d8b1421693af", "report_id": "satellite_satellite-land-cover_completeness_q03", "dataset_id": "satellite-land-cover", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q03", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 2. Inspect and view data for the defined AoI (Iberian Peninsula) > Potential Drivers and Methodological Considerations in Afforestation/Deforestation Trends in the Iberian Peninsula (1992–2022)", "title": "Satellite Land Cover trend assessment for Spatial Planning and Land Management", "chunk_index": 15, "token_count": 1150, "text_raw": "ensor surface reflectance time series derived from several satellite missions. As sensor technology evolved, different instruments were used to produce consistent global composites.**\n\n
\n\n| Surface Reflectance (SR) Input | Reference Period | Satellite Sensor / Mission | Main Characteristics |\n|---|---|---|---|\n| AVHRR SR composites | 1992–1999 | AVHRR-2 (NOAA-11, NOAA-14) | ~1 km spatial resolution, visible and near-infrared observations |\n| SPOT-VGT SR composites | 1999–2013 | SPOT-4 / SPOT-5 VEGETATION | 1 km spatial resolution, 4 spectral bands (blue, red, NIR, SWIR) |\n| MERIS SR composites | 2003–2012 | Envisat MERIS | ~300 m spatial resolution, 15 spectral bands in visible and near-infrared |\n| PROBA-V SR composites | 2013–2019 | PROBA-V | ~300 m spatial resolution, 4 spectral bands (blue, red, NIR, SWIR) |\n| Sentinel-3 SR composites | 2020 | Sentinel-3 OLCI | ~300 m spatial resolution, multispectral observations |\n| Sentinel-3 SR composites | 2021–2022 | Sentinel-3 OLCI + SLSTR | Optical and thermal observations supporting land-cover mapping |\n
\n\n\n- In addition to optical surface reflectance observations, auxiliary datasets are used to improve land-cover mapping. In particular, **Envisat ASAR Wide Swath Mode (WSM)** radar observations (2005–2012) are incorporated as ancillary information [(C3S - LC Target Requirements and Gap Analysis Document, 2024)](https://dast.copernicus-climate.eu/documents/satellite-land-cover/WP3-TR-GAD-2023_LC_v1.1_final.pdf).\n\n- While the dataset relies on multiple satellite sensors, it is not generated through simple sensor-to-sensor transitions. Instead, the global land-cover time series is built around a baseline land-cover map derived from MERIS observations (2003–2012). Earlier periods (1992–2003) are reconstructed through back-dating using AVHRR and SPOT-VGT data, while later years are produced through incremental updates using SPOT-VGT, PROBA-V, and Sentinel-3 observations. During several periods, multiple sensors are used simultaneously, with different roles (e.g., temporal consistency versus spatial refinement).\n\n
\n\n| Global LC database | Reference period | Satellite data source |\n|---|---|---|\n| Baseline 10-year global LC map | 2003–2012 | • MERIS FR/RR global SR composites between 2003 and 2012 |\n| Global annual LC maps | 1992–1999 | • Baseline 10-year global LC map
• AVHRR global SR composites between 1992 and 1999 for back-dating the baseline |\n| | 1999–2013 | • Baseline 10-year global LC map
• SPOT-VGT global SR composites between 1999 and 2013 for up- and back-dating the baseline
• MERIS FR global SR composites between 2003 and 2012 to delineate the identified changes at 300 m spatial resolution
• PROBA-V global SR composites at 300 m for year 2013 to delineate the identified changes at 300 m spatial resolution |\n| | 2014–2019 | • PROBA-V global SR composites at 1 km for years 2014 to 2019 for updating the baseline
• PROBA-V time series at 300 m for 2014 to 2019 to delineate the identified changes at the LC map spatial resolution |\n| | 2020–2022 | • S3 global SR composites at 1 km for years 2020 to 2022 for updating the baseline
• S3 time series at 300 m for 2020 to 2022 to delineate the identified changes at the LC map spatial resolution |\n
\n\n\n(code-section-3)=", "text_with_prefix": "EQC Quality Assessment: \"Satellite Land Cover trend assessment for Spatial Planning and Land Management\"\nDataset: satellite-land-cover [CDS]\nAspect: completeness_q03 | Category: Satellite_ECVs\nSection: Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 2. Inspect and view data for the defined AoI (Iberian Peninsula) > Potential Drivers and Methodological Considerations in Afforestation/Deforestation Trends in the Iberian Peninsula (1992–2022)\n---\nensor surface reflectance time series derived from several satellite missions. As sensor technology evolved, different instruments were used to produce consistent global composites.**\n\n
\n\n| Surface Reflectance (SR) Input | Reference Period | Satellite Sensor / Mission | Main Characteristics |\n|---|---|---|---|\n| AVHRR SR composites | 1992–1999 | AVHRR-2 (NOAA-11, NOAA-14) | ~1 km spatial resolution, visible and near-infrared observations |\n| SPOT-VGT SR composites | 1999–2013 | SPOT-4 / SPOT-5 VEGETATION | 1 km spatial resolution, 4 spectral bands (blue, red, NIR, SWIR) |\n| MERIS SR composites | 2003–2012 | Envisat MERIS | ~300 m spatial resolution, 15 spectral bands in visible and near-infrared |\n| PROBA-V SR composites | 2013–2019 | PROBA-V | ~300 m spatial resolution, 4 spectral bands (blue, red, NIR, SWIR) |\n| Sentinel-3 SR composites | 2020 | Sentinel-3 OLCI | ~300 m spatial resolution, multispectral observations |\n| Sentinel-3 SR composites | 2021–2022 | Sentinel-3 OLCI + SLSTR | Optical and thermal observations supporting land-cover mapping |\n
\n\n\n- In addition to optical surface reflectance observations, auxiliary datasets are used to improve land-cover mapping. In particular, **Envisat ASAR Wide Swath Mode (WSM)** radar observations (2005–2012) are incorporated as ancillary information [(C3S - LC Target Requirements and Gap Analysis Document, 2024)](https://dast.copernicus-climate.eu/documents/satellite-land-cover/WP3-TR-GAD-2023_LC_v1.1_final.pdf).\n\n- While the dataset relies on multiple satellite sensors, it is not generated through simple sensor-to-sensor transitions. Instead, the global land-cover time series is built around a baseline land-cover map derived from MERIS observations (2003–2012). Earlier periods (1992–2003) are reconstructed through back-dating using AVHRR and SPOT-VGT data, while later years are produced through incremental updates using SPOT-VGT, PROBA-V, and Sentinel-3 observations. During several periods, multiple sensors are used simultaneously, with different roles (e.g., temporal consistency versus spatial refinement).\n\n
\n\n| Global LC database | Reference period | Satellite data source |\n|---|---|---|\n| Baseline 10-year global LC map | 2003–2012 | • MERIS FR/RR global SR composites between 2003 and 2012 |\n| Global annual LC maps | 1992–1999 | • Baseline 10-year global LC map
• AVHRR global SR composites between 1992 and 1999 for back-dating the baseline |\n| | 1999–2013 | • Baseline 10-year global LC map
• SPOT-VGT global SR composites between 1999 and 2013 for up- and back-dating the baseline
• MERIS FR global SR composites between 2003 and 2012 to delineate the identified changes at 300 m spatial resolution
• PROBA-V global SR composites at 300 m for year 2013 to delineate the identified changes at 300 m spatial resolution |\n| | 2014–2019 | • PROBA-V global SR composites at 1 km for years 2014 to 2019 for updating the baseline
• PROBA-V time series at 300 m for 2014 to 2019 to delineate the identified changes at the LC map spatial resolution |\n| | 2020–2022 | • S3 global SR composites at 1 km for years 2020 to 2022 for updating the baseline
• S3 time series at 300 m for 2020 to 2022 to delineate the identified changes at the LC map spatial resolution |\n
\n\n\n(code-section-3)="} {"chunk_id": "satellite_satellite-land-cover_completeness_q03__2bdf54f2e524", "report_id": "satellite_satellite-land-cover_completeness_q03", "dataset_id": "satellite-land-cover", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q03", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 3. Breakpoint Detection", "title": "Satellite Land Cover trend assessment for Spatial Planning and Land Management", "chunk_index": 16, "token_count": 952, "text_raw": "A first step in assessing temporal behaviour is to analyse the **rate of change** of the forest-area time series. This provides an intuitive view of how the pace of forest expansion evolves over time and helps highlight periods where the trajectory deviates from its typical pattern.\n\nHowever, while such exploratory diagnostics identify points of interest, they do not by themselves confirm whether the observed changes are structurally meaningful. Year-to-year variations may reflect noise, smoothing effects, or gradual adjustments in the underlying product. To formally assess structural changes, a **statistical breakpoint detection method** is applied.\n\nThe analysis therefore combines **rate-of-change diagnostics** with **statistical segmentation**.\n\n---\n\n**Rate of Change Calculation**\n\n- The **first derivative** of the forest-area time series is computed to quantify the **year-to-year change in forest extent**.\n- This highlights **periods where the pace of forest expansion deviates from its typical behaviour**, providing an initial indication of potential anomalies or shifts in the series.\n\n---\n\n**Statistical Thresholding of Derivative Spikes**\n\n- To identify unusually strong deviations, a dynamic threshold is defined as:\n\n**Threshold = mean(|rate of change|) + 1.5 × standard deviation**\n\n- Years where the **absolute rate of change exceeds this threshold** are flagged as **derivative spikes**.\n- These spikes represent periods where the magnitude of change is unusually large relative to the typical variability of the series.\n- This step serves as a **diagnostic indicator of anomalous growth behaviour**, but does not in itself define structural breakpoints.\n\n---\n\n**Breakpoint Detection Using the PELT Algorithm**\n\n- The **Pruned Exact Linear Time (PELT)** algorithm is used to detect **changes in the statistical behaviour of the growth rate**, rather than in the cumulative forest-area series itself.\n\n- The algorithm is therefore applied to the **first derivative (growth rate)**, allowing detection of:\n - transitions between periods with different **rates of forest expansion**, rather than absolute levels.\n\n- The implementation uses:\n - the **L2 cost function** (`model=\"l2\"`) to minimise variance within segments,\n - a **fixed penalty parameter** (`pen = 2.5`) applied to a standardised growth-rate series,\n - and a **minimum segment length (`min_size = 3`)** to avoid short, unstable segments.\n\n- Prior to segmentation, the growth-rate series may be **standardised** to ensure comparability across regions with different magnitudes of forest change.\n\n---\n\n**Interpreting Breakpoint Timing**\n\nBreakpoint positions should be interpreted as **approximate indicators of when a change in growth dynamics begins**, rather than exact points of maximum change.\n\nBecause segmentation identifies **changes in statistical behaviour**, the detected breakpoint may:\n\n- precede the most visible increase in growth rate, or \n- occur within a broader transition period.\n\nAs a result, breakpoint timing should be considered **indicative rather than exact**, and interpreted in conjunction with derivative-based diagnostics.\n\nCalculate breakpoints and derivative spikes for each region\nNumber of subplot columns in the regional figure layout\nThreshold multiplier used to flag unusually large derivative spikes:\nspike if |growth rate| > mean(|growth rate|) + spike_std_factor * std(growth rate)\nMinimum number of derivative observations allowed in each PELT segment.\nPenalty term for PELT breakpoint detection.\nStandardize the growth-rate series before applying PELT so that breakpoint\ndetection is comparable across regions with different magnitudes of forest change.\nremove invalid and terminal interval\n==================================================\n==================================================\n\nPlot Breakpoints Series\nStore spikes\nStore breakpoints\nBackground phases\nRemove empty axes\nProcessing phases legend\nLine/signal legend\nCompact spacing between legends and plots\n\nPlot the aligned breakpoints and derivative spikes with known methodological changes present in more than half of the regions\nRename Interval column\nAdd relation to processing transition\nDrop helper column\nSort by number of regions, then by interval\nBuild tables\nDisplay nicely\n\n```text\n### Breakpoint Table ###\n```\n\n```text\nintervals number of regions \\\n0 2001-2002 12 \n1 1996-1997 10 \n2 2006-2007 1 \n3 2011-2012 1 \n4 2016-2017 1", "text_with_prefix": "EQC Quality Assessment: \"Satellite Land Cover trend assessment for Spatial Planning and Land Management\"\nDataset: satellite-land-cover [CDS]\nAspect: completeness_q03 | Category: Satellite_ECVs\nSection: Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 3. Breakpoint Detection\n---\nA first step in assessing temporal behaviour is to analyse the **rate of change** of the forest-area time series. This provides an intuitive view of how the pace of forest expansion evolves over time and helps highlight periods where the trajectory deviates from its typical pattern.\n\nHowever, while such exploratory diagnostics identify points of interest, they do not by themselves confirm whether the observed changes are structurally meaningful. Year-to-year variations may reflect noise, smoothing effects, or gradual adjustments in the underlying product. To formally assess structural changes, a **statistical breakpoint detection method** is applied.\n\nThe analysis therefore combines **rate-of-change diagnostics** with **statistical segmentation**.\n\n---\n\n**Rate of Change Calculation**\n\n- The **first derivative** of the forest-area time series is computed to quantify the **year-to-year change in forest extent**.\n- This highlights **periods where the pace of forest expansion deviates from its typical behaviour**, providing an initial indication of potential anomalies or shifts in the series.\n\n---\n\n**Statistical Thresholding of Derivative Spikes**\n\n- To identify unusually strong deviations, a dynamic threshold is defined as:\n\n**Threshold = mean(|rate of change|) + 1.5 × standard deviation**\n\n- Years where the **absolute rate of change exceeds this threshold** are flagged as **derivative spikes**.\n- These spikes represent periods where the magnitude of change is unusually large relative to the typical variability of the series.\n- This step serves as a **diagnostic indicator of anomalous growth behaviour**, but does not in itself define structural breakpoints.\n\n---\n\n**Breakpoint Detection Using the PELT Algorithm**\n\n- The **Pruned Exact Linear Time (PELT)** algorithm is used to detect **changes in the statistical behaviour of the growth rate**, rather than in the cumulative forest-area series itself.\n\n- The algorithm is therefore applied to the **first derivative (growth rate)**, allowing detection of:\n - transitions between periods with different **rates of forest expansion**, rather than absolute levels.\n\n- The implementation uses:\n - the **L2 cost function** (`model=\"l2\"`) to minimise variance within segments,\n - a **fixed penalty parameter** (`pen = 2.5`) applied to a standardised growth-rate series,\n - and a **minimum segment length (`min_size = 3`)** to avoid short, unstable segments.\n\n- Prior to segmentation, the growth-rate series may be **standardised** to ensure comparability across regions with different magnitudes of forest change.\n\n---\n\n**Interpreting Breakpoint Timing**\n\nBreakpoint positions should be interpreted as **approximate indicators of when a change in growth dynamics begins**, rather than exact points of maximum change.\n\nBecause segmentation identifies **changes in statistical behaviour**, the detected breakpoint may:\n\n- precede the most visible increase in growth rate, or \n- occur within a broader transition period.\n\nAs a result, breakpoint timing should be considered **indicative rather than exact**, and interpreted in conjunction with derivative-based diagnostics.\n\nCalculate breakpoints and derivative spikes for each region\nNumber of subplot columns in the regional figure layout\nThreshold multiplier used to flag unusually large derivative spikes:\nspike if |growth rate| > mean(|growth rate|) + spike_std_factor * std(growth rate)\nMinimum number of derivative observations allowed in each PELT segment.\nPenalty term for PELT breakpoint detection.\nStandardize the growth-rate series before applying PELT so that breakpoint\ndetection is comparable across regions with different magnitudes of forest change.\nremove invalid and terminal interval\n==================================================\n==================================================\n\nPlot Breakpoints Series\nStore spikes\nStore breakpoints\nBackground phases\nRemove empty axes\nProcessing phases legend\nLine/signal legend\nCompact spacing between legends and plots\n\nPlot the aligned breakpoints and derivative spikes with known methodological changes present in more than half of the regions\nRename Interval column\nAdd relation to processing transition\nDrop helper column\nSort by number of regions, then by interval\nBuild tables\nDisplay nicely\n\n```text\n### Breakpoint Table ###\n```\n\n```text\nintervals number of regions \\\n0 2001-2002 12 \n1 1996-1997 10 \n2 2006-2007 1 \n3 2011-2012 1 \n4 2016-2017 1"} {"chunk_id": "satellite_satellite-land-cover_completeness_q03__baba85c43acf", "report_id": "satellite_satellite-land-cover_completeness_q03", "dataset_id": "satellite-land-cover", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q03", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 3. Breakpoint Detection", "title": "Satellite Land Cover trend assessment for Spatial Planning and Land Management", "chunk_index": 17, "token_count": 935, "text_raw": "1996-1997 10 \n2 2006-2007 1 \n3 2011-2012 1 \n4 2016-2017 1\n\nregions name \\\n0 Andalucía, Aragón, Castilla y León, Castilla-La Mancha, Cataluña, Comunidad Foral de Navarra, Comunitat Valenciana, Galicia, La Rioja, Norte, País Vasco, Área Metropolitana de Lisboa \n1 Alentejo, Andalucía, Cantabria, Comunidad Foral de Navarra, Extremadura, Galicia, La Rioja, Norte, País Vasco, Área Metropolitana de Lisboa \n2 Cantabria \n3 Comunidad de Madrid \n4 Área Metropolitana de Lisboa\n\nnearest processing changes \n0 None within 2 years \n1 None within 2 years \n2 None within 2 years \n3 None within 2 years \n4 None within 2 years\n```\n\n```text\n### Derivative Spike Table ###\n```\n\n```text\nintervals number of regions \\\n0 1998-1999 15 \n1 2003-2004 5 \n2 2015-2016 4 \n3 1994-1995 2 \n4 2001-2002 2 \n5 1997-1998 1 \n6 1999-2000 1 \n7 2000-2001 1 \n8 2002-2003 1 \n9 2004-2005 1 \n10 2011-2012 1 \n11 2016-2017 1 \n12 2017-2018 1 \n13 2020-2021 1 \n14 2021-2022 1\n\nregions name \\\n0 Alentejo, Algarve, Andalucía, Aragón, Cantabria, Castilla-La Mancha, Cataluña, Comunidad Foral de Navarra, Extremadura, Galicia, Illes Balears, La Rioja, País Vasco, Principado de Asturias, Área Metropolitana de Lisboa \n1 Centro (PT), Galicia, Norte, Principado de Asturias, Región de Murcia \n2 Alentejo, Castilla y León, Comunidad de Madrid, Área Metropolitana de Lisboa \n3 Aragón, Comunitat Valenciana \n4 Galicia, Norte \n5 La Rioja \n6 Región de Murcia \n7 Principado de Asturias \n8 Cantabria \n9 Algarve \n10 Comunitat Valenciana \n11 Comunidad de Madrid \n12 Castilla y León \n13 Extremadura \n14 Región de Murcia\n\nnearest processing changes \n0 None within 2 years \n1 MERIS_baseline_start (2003) \n2 None within 2 years \n3 None within 2 years \n4 None within 2 years \n5 None within 2 years \n6 AVHRR_finish (1999); SPOT-VGT_start (1999) \n7 AVHRR_finish (1999); SPOT-VGT_start (1999) \n8 None within 2 years \n9 MERIS_baseline_start (2003) \n10 None within 2 years \n11 None within 2 years \n12 None within 2 years \n13 S3_OLCI_start (2020) \n14 S3_OLCI_start (2020); S3_OLCI_SLSTR_start (2021)\n```", "text_with_prefix": "EQC Quality Assessment: \"Satellite Land Cover trend assessment for Spatial Planning and Land Management\"\nDataset: satellite-land-cover [CDS]\nAspect: completeness_q03 | Category: Satellite_ECVs\nSection: Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 3. Breakpoint Detection\n---\n1996-1997 10 \n2 2006-2007 1 \n3 2011-2012 1 \n4 2016-2017 1\n\nregions name \\\n0 Andalucía, Aragón, Castilla y León, Castilla-La Mancha, Cataluña, Comunidad Foral de Navarra, Comunitat Valenciana, Galicia, La Rioja, Norte, País Vasco, Área Metropolitana de Lisboa \n1 Alentejo, Andalucía, Cantabria, Comunidad Foral de Navarra, Extremadura, Galicia, La Rioja, Norte, País Vasco, Área Metropolitana de Lisboa \n2 Cantabria \n3 Comunidad de Madrid \n4 Área Metropolitana de Lisboa\n\nnearest processing changes \n0 None within 2 years \n1 None within 2 years \n2 None within 2 years \n3 None within 2 years \n4 None within 2 years\n```\n\n```text\n### Derivative Spike Table ###\n```\n\n```text\nintervals number of regions \\\n0 1998-1999 15 \n1 2003-2004 5 \n2 2015-2016 4 \n3 1994-1995 2 \n4 2001-2002 2 \n5 1997-1998 1 \n6 1999-2000 1 \n7 2000-2001 1 \n8 2002-2003 1 \n9 2004-2005 1 \n10 2011-2012 1 \n11 2016-2017 1 \n12 2017-2018 1 \n13 2020-2021 1 \n14 2021-2022 1\n\nregions name \\\n0 Alentejo, Algarve, Andalucía, Aragón, Cantabria, Castilla-La Mancha, Cataluña, Comunidad Foral de Navarra, Extremadura, Galicia, Illes Balears, La Rioja, País Vasco, Principado de Asturias, Área Metropolitana de Lisboa \n1 Centro (PT), Galicia, Norte, Principado de Asturias, Región de Murcia \n2 Alentejo, Castilla y León, Comunidad de Madrid, Área Metropolitana de Lisboa \n3 Aragón, Comunitat Valenciana \n4 Galicia, Norte \n5 La Rioja \n6 Región de Murcia \n7 Principado de Asturias \n8 Cantabria \n9 Algarve \n10 Comunitat Valenciana \n11 Comunidad de Madrid \n12 Castilla y León \n13 Extremadura \n14 Región de Murcia\n\nnearest processing changes \n0 None within 2 years \n1 MERIS_baseline_start (2003) \n2 None within 2 years \n3 None within 2 years \n4 None within 2 years \n5 None within 2 years \n6 AVHRR_finish (1999); SPOT-VGT_start (1999) \n7 AVHRR_finish (1999); SPOT-VGT_start (1999) \n8 None within 2 years \n9 MERIS_baseline_start (2003) \n10 None within 2 years \n11 None within 2 years \n12 None within 2 years \n13 S3_OLCI_start (2020) \n14 S3_OLCI_start (2020); S3_OLCI_SLSTR_start (2021)\n```"} {"chunk_id": "satellite_satellite-land-cover_completeness_q03__a111f03c7c12", "report_id": "satellite_satellite-land-cover_completeness_q03", "dataset_id": "satellite-land-cover", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q03", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 3. Breakpoint Detection > Breakpoint and Derivative-Spike Analysis", "title": "Satellite Land Cover trend assessment for Spatial Planning and Land Management", "chunk_index": 18, "token_count": 1008, "text_raw": "Breakpoint detection and derivative analysis describe related but not identical aspects of temporal change in land-cover time series. Breakpoints indicate potential structural shifts, while derivative spikes identify years with strong year-to-year variation ([Chelali et al., 2019](https://doi.org/10.1109/JURSE.2019.8808967); [Chang et al., 2018](https://doi.org/10.1088/1755-1315/113/1/012087)). However, the repeated concentration of breakpoints and derivative spikes around similar years suggests that some of the observed variability may reflect methodological sensitivity, rather than clearly separable ecological or land-use processes.\n\nSignals are considered more robust when they show broad spatial coherence across regions, but temporal clustering alone is not treated as evidence of multiple independent change events. In satellite-derived land-cover products, changes in sensors, classification schemes, reference layers, or processing chains can introduce apparent discontinuities or amplified year-to-year variation. Therefore, breakpoint and derivative-spike results are assessed together, with attention to both spatial consistency and possible methodological artefacts ([Chang et al., 2018](https://doi.org/10.1088/1755-1315/113/1/012087)).\n\nThe **1998–1999 derivative signal**, affecting 15 regions, is the most spatially widespread short-term variability signal. Its limited correspondence with a clear breakpoint suggests that it may represent transient variability, classification sensitivity in mixed or transitional land-cover classes, or early inconsistencies in the time series.\n\nA major breakpoint cluster is also observed around **2001–2002**, affecting 12 regions. Although this is the most spatially coherent breakpoint signal, no clearly documented methodological transition is identified within a close temporal window. This weakens any direct attribution to a known processing-chain change. The signal may partly reflect real land-cover dynamics, including wildfire activity, forest management changes, or broader land-use transitions in the Iberian Peninsula [(Silva et al., 2011)](https://doi.org/10.1016/j.landurbplan.2011.03.001). Nevertheless, because breakpoint and derivative signals cluster around nearby years, this period is best interpreted as a strong but not fully attributable signal, potentially combining environmental change with methodological effects.\n\nA secondary derivative signal occurs around **2003–2004**, affecting 5 regions. This timing coincides with major wildfire activity and forest loss in parts of the Iberian Peninsula [(Mato et al., 2014)](https://api.semanticscholar.org/CorpusID:265869582), as well as improvements in detection capacity.\n\nAn earlier breakpoint cluster around **1996–1997**, affecting 10 regions, suggests a possible structural shift in western and northern regions. Its timing is consistent with broader land-use transitions, including farmland abandonment and natural forest regeneration [(Palmero-Iniesta et al., 2021)](https://api.semanticscholar.org/CorpusID:238829360), as well as changes in forest management. However, given its proximity to the stronger 1998–1999 derivative signal, it should not be treated as a fully independent event without further validation.\n\nOverall, the recurrence of breakpoint and derivative-spike detections around similar years indicates that the time series contains periods of heightened instability, but these should not be over-interpreted as discrete ecological events. The most defensible interpretation is that the late 1990s to early 2000s represent an interval of elevated variability possibly shaped by a combination of land-cover change, disturbance processes, and methodological sensitivity in the data product.\n\nAdditional breakpoint and derivative-spike detections affect only a small number of regions. These are best interpreted as localised or context-specific dynamics, potentially linked to wildfires, drought impacts, regional land-use change, plantation dynamics, or natural regeneration following abandonment [(Gouveia et al., 2012](https://nhess.copernicus.org/articles/12/3123/2012/)[; Vidal-Macua et al., 2017)](https://doi.org/10.1016/j.foreco.2017.10.011); [(Silva et al., 2011)](https://doi.org/10.1016/j.landurbplan.2011.03.001); [(Palmero-Iniesta et al., 2021)](https://api.semanticscholar.org/CorpusID:238829360).\n\n(code-section-4)=", "text_with_prefix": "EQC Quality Assessment: \"Satellite Land Cover trend assessment for Spatial Planning and Land Management\"\nDataset: satellite-land-cover [CDS]\nAspect: completeness_q03 | Category: Satellite_ECVs\nSection: Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 3. Breakpoint Detection > Breakpoint and Derivative-Spike Analysis\n---\nBreakpoint detection and derivative analysis describe related but not identical aspects of temporal change in land-cover time series. Breakpoints indicate potential structural shifts, while derivative spikes identify years with strong year-to-year variation ([Chelali et al., 2019](https://doi.org/10.1109/JURSE.2019.8808967); [Chang et al., 2018](https://doi.org/10.1088/1755-1315/113/1/012087)). However, the repeated concentration of breakpoints and derivative spikes around similar years suggests that some of the observed variability may reflect methodological sensitivity, rather than clearly separable ecological or land-use processes.\n\nSignals are considered more robust when they show broad spatial coherence across regions, but temporal clustering alone is not treated as evidence of multiple independent change events. In satellite-derived land-cover products, changes in sensors, classification schemes, reference layers, or processing chains can introduce apparent discontinuities or amplified year-to-year variation. Therefore, breakpoint and derivative-spike results are assessed together, with attention to both spatial consistency and possible methodological artefacts ([Chang et al., 2018](https://doi.org/10.1088/1755-1315/113/1/012087)).\n\nThe **1998–1999 derivative signal**, affecting 15 regions, is the most spatially widespread short-term variability signal. Its limited correspondence with a clear breakpoint suggests that it may represent transient variability, classification sensitivity in mixed or transitional land-cover classes, or early inconsistencies in the time series.\n\nA major breakpoint cluster is also observed around **2001–2002**, affecting 12 regions. Although this is the most spatially coherent breakpoint signal, no clearly documented methodological transition is identified within a close temporal window. This weakens any direct attribution to a known processing-chain change. The signal may partly reflect real land-cover dynamics, including wildfire activity, forest management changes, or broader land-use transitions in the Iberian Peninsula [(Silva et al., 2011)](https://doi.org/10.1016/j.landurbplan.2011.03.001). Nevertheless, because breakpoint and derivative signals cluster around nearby years, this period is best interpreted as a strong but not fully attributable signal, potentially combining environmental change with methodological effects.\n\nA secondary derivative signal occurs around **2003–2004**, affecting 5 regions. This timing coincides with major wildfire activity and forest loss in parts of the Iberian Peninsula [(Mato et al., 2014)](https://api.semanticscholar.org/CorpusID:265869582), as well as improvements in detection capacity.\n\nAn earlier breakpoint cluster around **1996–1997**, affecting 10 regions, suggests a possible structural shift in western and northern regions. Its timing is consistent with broader land-use transitions, including farmland abandonment and natural forest regeneration [(Palmero-Iniesta et al., 2021)](https://api.semanticscholar.org/CorpusID:238829360), as well as changes in forest management. However, given its proximity to the stronger 1998–1999 derivative signal, it should not be treated as a fully independent event without further validation.\n\nOverall, the recurrence of breakpoint and derivative-spike detections around similar years indicates that the time series contains periods of heightened instability, but these should not be over-interpreted as discrete ecological events. The most defensible interpretation is that the late 1990s to early 2000s represent an interval of elevated variability possibly shaped by a combination of land-cover change, disturbance processes, and methodological sensitivity in the data product.\n\nAdditional breakpoint and derivative-spike detections affect only a small number of regions. These are best interpreted as localised or context-specific dynamics, potentially linked to wildfires, drought impacts, regional land-use change, plantation dynamics, or natural regeneration following abandonment [(Gouveia et al., 2012](https://nhess.copernicus.org/articles/12/3123/2012/)[; Vidal-Macua et al., 2017)](https://doi.org/10.1016/j.foreco.2017.10.011); [(Silva et al., 2011)](https://doi.org/10.1016/j.landurbplan.2011.03.001); [(Palmero-Iniesta et al., 2021)](https://api.semanticscholar.org/CorpusID:238829360).\n\n(code-section-4)="} {"chunk_id": "satellite_satellite-land-cover_completeness_q03__b6976dca8a9a", "report_id": "satellite_satellite-land-cover_completeness_q03", "dataset_id": "satellite-land-cover", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q03", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 4. Trend Assessment", "title": "Satellite Land Cover trend assessment for Spatial Planning and Land Management", "chunk_index": 19, "token_count": 612, "text_raw": "The next and final step is to determine whether the trend should be computed for the entire period (total trend) or divided into segments based on detected breakpoints.\n\nThe decision between using the total trend or segmented trends is based on Sen’s Slope (Trend Magnitude) and Mann-Kendall p-values (Trend Significance). Sen’s Slope is a non-parametric estimator that calculates the median rate of change over time, making it robust to outliers. First, Sen’s Slope is computed for the total trend, capturing the overall rate of forest change. Then, Sen’s Slope is calculated for each segmented trend to assess whether breakpoints introduce significant shifts in trend magnitude. The Mann-Kendall test is a non-parametric statistical test used to assess the presence of upward or downward trend in a time series without requiring the data to follow any particular distribution. In parallel, the Mann-Kendall p-value is evaluated for both the total and segmented trends to measure trend significance—lower p-values indicate stronger evidence of a significant trend.\n\nThe final decision is based on the following approach:\n\n- If segmented trends show substantially different Sen’s Slopes compared to the total trend and exhibit stronger statistical significance (lower p-values), segmentation is preferred.\n- If segmented trends closely resemble the total trend or do not provide a clear statistical advantage, the total trend is used to maintain a simpler and more robust interpretation.\n\nPrepare data for trend calculation\nKeep only valid rows\n\nCalculate trends\nCompute total + segmented trends\nTotal Sen slope\nConvert to % area per decade, as requested by reviewer\nMann-Kendall\nBreakpoints and segmented trends\n\nThe following summary table reports:\n\n- **Growth rate (%/decade)** → long-term rate of forest expansion \n- **Significance** → statistical robustness of the full trend \n - `***` → highly significant (p < 0.001) \n - `**` → significant (p < 0.01) \n - `*` → moderately significant (p < 0.05) \n - no symbol → not statistically significant \n- **Breakpoints** → number of structural changes detected \n- **Segmented trend needed?** → whether segmented trends change the interpretation \n- **Confidence** → degree of agreement between full and segmented trend significance\n\nBuild summary table\nforest area for each region in the latest year\nSort by growth rate\nRound for display\nNormalize text values\nHTML table display\n\n```text\n21 regions were assessed. Segmented trends are not required in 18 of 21 regions. Confidence is High in 10 regions, Medium in 7, and Low in 4.\n```", "text_with_prefix": "EQC Quality Assessment: \"Satellite Land Cover trend assessment for Spatial Planning and Land Management\"\nDataset: satellite-land-cover [CDS]\nAspect: completeness_q03 | Category: Satellite_ECVs\nSection: Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 4. Trend Assessment\n---\nThe next and final step is to determine whether the trend should be computed for the entire period (total trend) or divided into segments based on detected breakpoints.\n\nThe decision between using the total trend or segmented trends is based on Sen’s Slope (Trend Magnitude) and Mann-Kendall p-values (Trend Significance). Sen’s Slope is a non-parametric estimator that calculates the median rate of change over time, making it robust to outliers. First, Sen’s Slope is computed for the total trend, capturing the overall rate of forest change. Then, Sen’s Slope is calculated for each segmented trend to assess whether breakpoints introduce significant shifts in trend magnitude. The Mann-Kendall test is a non-parametric statistical test used to assess the presence of upward or downward trend in a time series without requiring the data to follow any particular distribution. In parallel, the Mann-Kendall p-value is evaluated for both the total and segmented trends to measure trend significance—lower p-values indicate stronger evidence of a significant trend.\n\nThe final decision is based on the following approach:\n\n- If segmented trends show substantially different Sen’s Slopes compared to the total trend and exhibit stronger statistical significance (lower p-values), segmentation is preferred.\n- If segmented trends closely resemble the total trend or do not provide a clear statistical advantage, the total trend is used to maintain a simpler and more robust interpretation.\n\nPrepare data for trend calculation\nKeep only valid rows\n\nCalculate trends\nCompute total + segmented trends\nTotal Sen slope\nConvert to % area per decade, as requested by reviewer\nMann-Kendall\nBreakpoints and segmented trends\n\nThe following summary table reports:\n\n- **Growth rate (%/decade)** → long-term rate of forest expansion \n- **Significance** → statistical robustness of the full trend \n - `***` → highly significant (p < 0.001) \n - `**` → significant (p < 0.01) \n - `*` → moderately significant (p < 0.05) \n - no symbol → not statistically significant \n- **Breakpoints** → number of structural changes detected \n- **Segmented trend needed?** → whether segmented trends change the interpretation \n- **Confidence** → degree of agreement between full and segmented trend significance\n\nBuild summary table\nforest area for each region in the latest year\nSort by growth rate\nRound for display\nNormalize text values\nHTML table display\n\n```text\n21 regions were assessed. Segmented trends are not required in 18 of 21 regions. Confidence is High in 10 regions, Medium in 7, and Low in 4.\n```"} {"chunk_id": "satellite_satellite-land-cover_completeness_q03__a761882a8f7f", "report_id": "satellite_satellite-land-cover_completeness_q03", "dataset_id": "satellite-land-cover", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q03", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 4. Trend Assessment > Trend Analysis", "title": "Satellite Land Cover trend assessment for Spatial Planning and Land Management", "chunk_index": 20, "token_count": 311, "text_raw": "The results demonstrate that while breakpoints exist, whether driven by real environmental changes or influenced by sensor/methodological shifts, they do not necessarily require separate trend calculations. In most cases, the total trend provides a stable and reliable representation of forest evolution across the full dataset, reinforcing the idea that forest cover follows a largely continuous trajectory over time.\n\nHowever, certain regions exhibit segmented trends that deviate significantly from the total trend. These cases merit special attention, as they may reflect actual shifts in forest dynamics due to land management, climate events, or localised disturbances.\n\nA key consideration is the geographical and ecological diversity of the Iberian Peninsula. Larger regions with varied landscapes may show more gradual changes in forest cover, making the total trend more representative. In contrast, smaller or ecologically distinct regions may experience more abrupt shifts that justify the use of segmented trends in specific cases.\n\nUltimately, while breakpoints provide useful insights into short-term variations, they do not invalidate the overall trend. This analysis confirms that long-term forest trends in the Iberian Peninsula can be effectively assessed using the full dataset, with segmentation being relevant only when trends deviate sharply and are supported by contextual or statistical evidence.\n\n(code-section-5)=", "text_with_prefix": "EQC Quality Assessment: \"Satellite Land Cover trend assessment for Spatial Planning and Land Management\"\nDataset: satellite-land-cover [CDS]\nAspect: completeness_q03 | Category: Satellite_ECVs\nSection: Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 4. Trend Assessment > Trend Analysis\n---\nThe results demonstrate that while breakpoints exist, whether driven by real environmental changes or influenced by sensor/methodological shifts, they do not necessarily require separate trend calculations. In most cases, the total trend provides a stable and reliable representation of forest evolution across the full dataset, reinforcing the idea that forest cover follows a largely continuous trajectory over time.\n\nHowever, certain regions exhibit segmented trends that deviate significantly from the total trend. These cases merit special attention, as they may reflect actual shifts in forest dynamics due to land management, climate events, or localised disturbances.\n\nA key consideration is the geographical and ecological diversity of the Iberian Peninsula. Larger regions with varied landscapes may show more gradual changes in forest cover, making the total trend more representative. In contrast, smaller or ecologically distinct regions may experience more abrupt shifts that justify the use of segmented trends in specific cases.\n\nUltimately, while breakpoints provide useful insights into short-term variations, they do not invalidate the overall trend. This analysis confirms that long-term forest trends in the Iberian Peninsula can be effectively assessed using the full dataset, with segmentation being relevant only when trends deviate sharply and are supported by contextual or statistical evidence.\n\n(code-section-5)="} {"chunk_id": "satellite_satellite-land-cover_completeness_q03__0ed8fdce42eb", "report_id": "satellite_satellite-land-cover_completeness_q03", "dataset_id": "satellite-land-cover", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q03", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 5. Discussion", "title": "Satellite Land Cover trend assessment for Spatial Planning and Land Management", "chunk_index": 21, "token_count": 437, "text_raw": "The results of this analysis highlight the value of the dataset for monitoring forest dynamics in the Iberian Peninsula. The dataset proves robust for long-term trend detection, with methodological transitions and sensor shifts,having none to limited impact on overarching patterns of forest change.\n\nFor applications in spatial planning and land management, this consistency is crucial. The ability to track forest evolution across decades allows regional planners and policymakers to evaluate the outcomes of afforestation policies, conservation efforts, and land abandonment. Even in the presence of minor reclassifications or local anomalies, the dataset supports confident assessments of broader land cover trends.\n\nNonetheless, the analysis also underscores the importance of methodological awareness. Breakpoints, particularly those associated with known sensor transitions, can introduce artificial changes in the time series. These do not necessarily reflect ecological shifts but rather improvements in resolution, classification algorithms, or spectral capabilities. Users must remain cautious when interpreting abrupt changes without considering the underlying data lineage.\n\nGeographic context further shapes interpretability. Larger, heterogeneous regions tend to smooth out classification noise, while smaller or fragmented areas may reflect sharper, localized breaks. This has direct implications for regional land management strategies, where scale and land use diversity influence the reliability of detected trends.\n\nIn practical terms, the dataset is well-suited for identifying general trends in forest cover, but users should:\n\n- Complement statistical analysis with contextual information (e.g., known policy events, fire records),\n- Use segmentation carefully—reserving it for regions with strong evidence of structural change,\n- Remain cautious when interpreting data near known sensor transition years.\n\nIn conclusion, the dataset provides a strong foundation for forest monitoring and planning. However, optimal use requires not just quantitative analysis, but also critical interpretation of when and why changes appear in the data. This fusion of methods and context is essential for informed decision-making in land and forest management.", "text_with_prefix": "EQC Quality Assessment: \"Satellite Land Cover trend assessment for Spatial Planning and Land Management\"\nDataset: satellite-land-cover [CDS]\nAspect: completeness_q03 | Category: Satellite_ECVs\nSection: Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 5. Discussion\n---\nThe results of this analysis highlight the value of the dataset for monitoring forest dynamics in the Iberian Peninsula. The dataset proves robust for long-term trend detection, with methodological transitions and sensor shifts,having none to limited impact on overarching patterns of forest change.\n\nFor applications in spatial planning and land management, this consistency is crucial. The ability to track forest evolution across decades allows regional planners and policymakers to evaluate the outcomes of afforestation policies, conservation efforts, and land abandonment. Even in the presence of minor reclassifications or local anomalies, the dataset supports confident assessments of broader land cover trends.\n\nNonetheless, the analysis also underscores the importance of methodological awareness. Breakpoints, particularly those associated with known sensor transitions, can introduce artificial changes in the time series. These do not necessarily reflect ecological shifts but rather improvements in resolution, classification algorithms, or spectral capabilities. Users must remain cautious when interpreting abrupt changes without considering the underlying data lineage.\n\nGeographic context further shapes interpretability. Larger, heterogeneous regions tend to smooth out classification noise, while smaller or fragmented areas may reflect sharper, localized breaks. This has direct implications for regional land management strategies, where scale and land use diversity influence the reliability of detected trends.\n\nIn practical terms, the dataset is well-suited for identifying general trends in forest cover, but users should:\n\n- Complement statistical analysis with contextual information (e.g., known policy events, fire records),\n- Use segmentation carefully—reserving it for regions with strong evidence of structural change,\n- Remain cautious when interpreting data near known sensor transition years.\n\nIn conclusion, the dataset provides a strong foundation for forest monitoring and planning. However, optimal use requires not just quantitative analysis, but also critical interpretation of when and why changes appear in the data. This fusion of methods and context is essential for informed decision-making in land and forest management."} {"chunk_id": "satellite_satellite-land-cover_completeness_q03__748eeae3b912", "report_id": "satellite_satellite-land-cover_completeness_q03", "dataset_id": "satellite-land-cover", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q03", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Land Cover trend assessment for Spatial Planning and Land Management > ℹ️ If you want to know more > Key Resources", "title": "Satellite Land Cover trend assessment for Spatial Planning and Land Management", "chunk_index": 22, "token_count": 344, "text_raw": "* The CDS catalogue entry for the data used was [Land cover classification gridded maps from 1992 to present derived from satellite observations](https://cds.climate.copernicus.eu/datasets/satellite-land-cover?tab=overview)\n\n* Product User Guide and Specification of the dataset [version 2.1](https://dast.copernicus-climate.eu/documents/satellite-land-cover/D5.3.1_PUGS_ICDR_LC_v2.1.x_PRODUCTS_v1.1.pdf) and [version 2.0](https://dast.copernicus-climate.eu/documents/satellite-land-cover/D3.3.11-v1.0_PUGS_CDR_LC-CCI_v2.0.7cds_Products_v1.0.1_APPROVED_Ver1.pdf)\n\n* [Eurostat NUTS](https://ec.europa.eu/eurostat/web/nuts) (Nomenclature of territorial units for statistics)\n\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), c3s_eqc_automatic_quality_control, prepared by [B-Open](https://www.bopen.eu/)", "text_with_prefix": "EQC Quality Assessment: \"Satellite Land Cover trend assessment for Spatial Planning and Land Management\"\nDataset: satellite-land-cover [CDS]\nAspect: completeness_q03 | Category: Satellite_ECVs\nSection: Satellite Land Cover trend assessment for Spatial Planning and Land Management > ℹ️ If you want to know more > Key Resources\n---\n* The CDS catalogue entry for the data used was [Land cover classification gridded maps from 1992 to present derived from satellite observations](https://cds.climate.copernicus.eu/datasets/satellite-land-cover?tab=overview)\n\n* Product User Guide and Specification of the dataset [version 2.1](https://dast.copernicus-climate.eu/documents/satellite-land-cover/D5.3.1_PUGS_ICDR_LC_v2.1.x_PRODUCTS_v1.1.pdf) and [version 2.0](https://dast.copernicus-climate.eu/documents/satellite-land-cover/D3.3.11-v1.0_PUGS_CDR_LC-CCI_v2.0.7cds_Products_v1.0.1_APPROVED_Ver1.pdf)\n\n* [Eurostat NUTS](https://ec.europa.eu/eurostat/web/nuts) (Nomenclature of territorial units for statistics)\n\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), c3s_eqc_automatic_quality_control, prepared by [B-Open](https://www.bopen.eu/)"} {"chunk_id": "satellite_satellite-land-cover_completeness_q03__6aae0fe163a4", "report_id": "satellite_satellite-land-cover_completeness_q03", "dataset_id": "satellite-land-cover", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q03", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Land Cover trend assessment for Spatial Planning and Land Management > ℹ️ If you want to know more > References", "title": "Satellite Land Cover trend assessment for Spatial Planning and Land Management", "chunk_index": 23, "token_count": 1007, "text_raw": "- [EUROSTAT: Forests, forestry and logging](https://ec.europa.eu/eurostat/statistics-explained/index.php?title=Forests,_forestry_and_logging)\n\n- [New EU Forest Strategy for 2030](https://commission.europa.eu/document/cf3294e1-8358-4c93-8de4-3e1503b95201_en)\n\n- [Mato, M.M., Legido, J.L., Miguez, E., Carballas, T., Jiménez, E., Caselles, V., & Andrade, M.I. (2014). Analysis of burnt areas and number of forest fires in the Iberian Peninsula.](https://api.semanticscholar.org/CorpusID:265869582)\n\n- [Silva, J., Vaz, P.G., Moreira, F., Catry, F.X., & Rego, F.C. (2011). Wildfires as a major driver of landscape dynamics in three fire-prone areas of Portugal. *Landscape and forest Planning, 101*, 349-358.](https://api.semanticscholar.org/CorpusID:83540326)\n\n- [Gouveia, C.M., Bastos, A., Trigo, R.M., & DaCamara, C.C. (2012). Drought impacts on vegetation in the pre- and post-fire events over Iberian Peninsula. *Natural Hazards and Earth System Sciences, 12*, 3123-3137.](https://nhess.copernicus.org/articles/12/3123/2012/)\n\n- [Vidal-Macua, J.J., Ninyerola, M., Zabala, A., Domingo‐Marimon, C., & Pons, X. (2017). Factors affecting forest dynamics in the Iberian Peninsula from 1987 to 2012: the role of topography and drought. *Forest Ecology and Management, 406*, 290-306.](https://doi.org/10.1016/j.foreco.2017.10.011)\n\n- [Sevillano, E.H., Contador, J.F., Schnabel, S., Pulido, M., & Ibáñez, J.Q. (2018). Using spatial models of temporal tree dynamics to evaluate the implementation of EU afforestation policies in rangelands of SW Spain. *Land Use Policy.*](https://doi.org/10.1016/j.landusepol.2018.06.054)\n\n- [Palmero-Iniesta, M., Espelta, J.M., Padial-Iglesias, M., González-Guerrero, Ó., Pesquer, L., Domingo‐Marimon, C., Ninyerola, M., Pons, X., & Pino, J. (2021). The Role of Recent (1985–2014) Patterns of Land Abandonment and Environmental Factors in the Establishment and Growth of Secondary Forests in the Iberian Peninsula. *Land.*](https://api.semanticscholar.org/CorpusID:238829360)\n\n- [Silva, J., Vaz, P.G., Moreira, F., Catry, F.X., & Rego, F.C. (2011). Wildfires as a major driver of landscape dynamics in three fire-prone areas of Portugal. *Landscape and forest Planning, 101*, 349-358.](https://doi.org/10.1016/j.landurbplan.2011.03.001)\n\n- [Regos, A., Ninyerola, M., Moré, G., & Pons, X. (2015). Linking land cover dynamics with driving forces in mountain landscapes of the Northwestern Iberian Peninsula. *Int. J. Appl. Earth Obs. Geoinformation, 38*, 1-14.](https://doi.org/10.1016/j.jag.2014.11.010)\n\n- [Toté, C., Swinnen, E., & Henocq, C. (2024). An Evaluation of Sentinel-3 SYN VGT Products in Comparison to the SPOT/VEGETATION and PROBA-V Archives. Remote Sensing, 16(20), 3822. https://doi.org/10.3390/rs16203822 ](https://doi.org/10.3390/rs16203822)", "text_with_prefix": "EQC Quality Assessment: \"Satellite Land Cover trend assessment for Spatial Planning and Land Management\"\nDataset: satellite-land-cover [CDS]\nAspect: completeness_q03 | Category: Satellite_ECVs\nSection: Satellite Land Cover trend assessment for Spatial Planning and Land Management > ℹ️ If you want to know more > References\n---\n- [EUROSTAT: Forests, forestry and logging](https://ec.europa.eu/eurostat/statistics-explained/index.php?title=Forests,_forestry_and_logging)\n\n- [New EU Forest Strategy for 2030](https://commission.europa.eu/document/cf3294e1-8358-4c93-8de4-3e1503b95201_en)\n\n- [Mato, M.M., Legido, J.L., Miguez, E., Carballas, T., Jiménez, E., Caselles, V., & Andrade, M.I. (2014). Analysis of burnt areas and number of forest fires in the Iberian Peninsula.](https://api.semanticscholar.org/CorpusID:265869582)\n\n- [Silva, J., Vaz, P.G., Moreira, F., Catry, F.X., & Rego, F.C. (2011). Wildfires as a major driver of landscape dynamics in three fire-prone areas of Portugal. *Landscape and forest Planning, 101*, 349-358.](https://api.semanticscholar.org/CorpusID:83540326)\n\n- [Gouveia, C.M., Bastos, A., Trigo, R.M., & DaCamara, C.C. (2012). Drought impacts on vegetation in the pre- and post-fire events over Iberian Peninsula. *Natural Hazards and Earth System Sciences, 12*, 3123-3137.](https://nhess.copernicus.org/articles/12/3123/2012/)\n\n- [Vidal-Macua, J.J., Ninyerola, M., Zabala, A., Domingo‐Marimon, C., & Pons, X. (2017). Factors affecting forest dynamics in the Iberian Peninsula from 1987 to 2012: the role of topography and drought. *Forest Ecology and Management, 406*, 290-306.](https://doi.org/10.1016/j.foreco.2017.10.011)\n\n- [Sevillano, E.H., Contador, J.F., Schnabel, S., Pulido, M., & Ibáñez, J.Q. (2018). Using spatial models of temporal tree dynamics to evaluate the implementation of EU afforestation policies in rangelands of SW Spain. *Land Use Policy.*](https://doi.org/10.1016/j.landusepol.2018.06.054)\n\n- [Palmero-Iniesta, M., Espelta, J.M., Padial-Iglesias, M., González-Guerrero, Ó., Pesquer, L., Domingo‐Marimon, C., Ninyerola, M., Pons, X., & Pino, J. (2021). The Role of Recent (1985–2014) Patterns of Land Abandonment and Environmental Factors in the Establishment and Growth of Secondary Forests in the Iberian Peninsula. *Land.*](https://api.semanticscholar.org/CorpusID:238829360)\n\n- [Silva, J., Vaz, P.G., Moreira, F., Catry, F.X., & Rego, F.C. (2011). Wildfires as a major driver of landscape dynamics in three fire-prone areas of Portugal. *Landscape and forest Planning, 101*, 349-358.](https://doi.org/10.1016/j.landurbplan.2011.03.001)\n\n- [Regos, A., Ninyerola, M., Moré, G., & Pons, X. (2015). Linking land cover dynamics with driving forces in mountain landscapes of the Northwestern Iberian Peninsula. *Int. J. Appl. Earth Obs. Geoinformation, 38*, 1-14.](https://doi.org/10.1016/j.jag.2014.11.010)\n\n- [Toté, C., Swinnen, E., & Henocq, C. (2024). An Evaluation of Sentinel-3 SYN VGT Products in Comparison to the SPOT/VEGETATION and PROBA-V Archives. Remote Sensing, 16(20), 3822. https://doi.org/10.3390/rs16203822 ](https://doi.org/10.3390/rs16203822)"} {"chunk_id": "satellite_satellite-land-cover_completeness_q03__c68f98945542", "report_id": "satellite_satellite-land-cover_completeness_q03", "dataset_id": "satellite-land-cover", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q03", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Land Cover trend assessment for Spatial Planning and Land Management > ℹ️ If you want to know more > References", "title": "Satellite Land Cover trend assessment for Spatial Planning and Land Management", "chunk_index": 24, "token_count": 208, "text_raw": "BA-V Archives. Remote Sensing, 16(20), 3822. https://doi.org/10.3390/rs16203822 ](https://doi.org/10.3390/rs16203822)\n\n- [Zhao, T., Zhang, X., Gao, Y., Mi, J., Liu, W., Wang, J., Jiang, M., & Liu, L. (2023). Assessing the Accuracy and Consistency of Six Fine-Resolution Global Land Cover Products Using a Novel Stratified Random Sampling Validation Dataset. Remote Sensing, 15(9)](https://doi.org/10.3390/rs15092285)", "text_with_prefix": "EQC Quality Assessment: \"Satellite Land Cover trend assessment for Spatial Planning and Land Management\"\nDataset: satellite-land-cover [CDS]\nAspect: completeness_q03 | Category: Satellite_ECVs\nSection: Satellite Land Cover trend assessment for Spatial Planning and Land Management > ℹ️ If you want to know more > References\n---\nBA-V Archives. Remote Sensing, 16(20), 3822. https://doi.org/10.3390/rs16203822 ](https://doi.org/10.3390/rs16203822)\n\n- [Zhao, T., Zhang, X., Gao, Y., Mi, J., Liu, W., Wang, J., Jiang, M., & Liu, L. (2023). Assessing the Accuracy and Consistency of Six Fine-Resolution Global Land Cover Products Using a Novel Stratified Random Sampling Validation Dataset. Remote Sensing, 15(9)](https://doi.org/10.3390/rs15092285)"} {"chunk_id": "satellite_satellite-land-cover_consistency-assessment_q01__576466c2fb72", "report_id": "satellite_satellite-land-cover_consistency-assessment_q01", "dataset_id": "satellite-land-cover", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency-assessment_q01", "aspect_base": "consistency-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Land Cover completeness for Spatial Planning and Land Management", "title": "Satellite Land Cover completeness for Spatial Planning and Land Management", "chunk_index": 0, "token_count": 91, "text_raw": "Production date: 05-06-2026\n\nProduced by: Inês Girão e Luís Figueiredo(+ATLANTIC)", "text_with_prefix": "EQC Quality Assessment: \"Satellite Land Cover completeness for Spatial Planning and Land Management\"\nDataset: satellite-land-cover [CDS]\nAspect: consistency-assessment_q01 | Category: Satellite_ECVs\nSection: Satellite Land Cover completeness for Spatial Planning and Land Management\n---\nProduction date: 05-06-2026\n\nProduced by: Inês Girão e Luís Figueiredo(+ATLANTIC)"} {"chunk_id": "satellite_satellite-land-cover_consistency-assessment_q01__5f9f67f22ad4", "report_id": "satellite_satellite-land-cover_consistency-assessment_q01", "dataset_id": "satellite-land-cover", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency-assessment_q01", "aspect_base": "consistency-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Land Cover completeness for Spatial Planning and Land Management > Quality assessment question", "title": "Satellite Land Cover completeness for Spatial Planning and Land Management", "chunk_index": 1, "token_count": 260, "text_raw": "* **How consistent with other datasets are satellite observations in capturing land cover changes, such as urbanisation?**\n\nLand cover data is a vital resource across a wide range of fields, from climate change research to urban and regional planning. Products with long historical timelines allow scientists, policymakers, and planners to assess how land use and land cover have evolved over time, supporting evidence-based decision-making.\n\nIn this notebook, we use the ***Land Cover Classification Gridded Maps from 1992 to present derived from satellite observations*** (hereafter referred to as **LC**) provided by the Climate Data Store (CDS) of the [Copernicus Climate Change Service (C3S)](https://climate.copernicus.eu/esotc/2023). The analysis focuses on evaluating the consistency of C3S-derived settlement (Artificial Land) classes with independent statistical records from **EUROSTAT**, across selected years and NUTS2 regions in the Iberian Peninsula.", "text_with_prefix": "EQC Quality Assessment: \"Satellite Land Cover completeness for Spatial Planning and Land Management\"\nDataset: satellite-land-cover [CDS]\nAspect: consistency-assessment_q01 | Category: Satellite_ECVs\nSection: Satellite Land Cover completeness for Spatial Planning and Land Management > Quality assessment question\n---\n* **How consistent with other datasets are satellite observations in capturing land cover changes, such as urbanisation?**\n\nLand cover data is a vital resource across a wide range of fields, from climate change research to urban and regional planning. Products with long historical timelines allow scientists, policymakers, and planners to assess how land use and land cover have evolved over time, supporting evidence-based decision-making.\n\nIn this notebook, we use the ***Land Cover Classification Gridded Maps from 1992 to present derived from satellite observations*** (hereafter referred to as **LC**) provided by the Climate Data Store (CDS) of the [Copernicus Climate Change Service (C3S)](https://climate.copernicus.eu/esotc/2023). The analysis focuses on evaluating the consistency of C3S-derived settlement (Artificial Land) classes with independent statistical records from **EUROSTAT**, across selected years and NUTS2 regions in the Iberian Peninsula."} {"chunk_id": "satellite_satellite-land-cover_consistency-assessment_q01__a39131057aab", "report_id": "satellite_satellite-land-cover_consistency-assessment_q01", "dataset_id": "satellite-land-cover", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency-assessment_q01", "aspect_base": "consistency-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Land Cover completeness for Spatial Planning and Land Management > Quality assessment statement", "title": "Satellite Land Cover completeness for Spatial Planning and Land Management", "chunk_index": 2, "token_count": 399, "text_raw": "These are the key outcomes of this assessment\n\n* The dataset demonstrates physical consistency and precision among the different datasets as it aligns closely with equivalent statistics from official sources [(EUROSTAT,2025)](https://ec.europa.eu/eurostat/databrowser/bookmark/63d3fd90-2f5b-4bec-99f3-4f88814e2624?lang=en). Specifically, the analysis shows that the Urban/Settlements category closely matches EUROSTAT's statistics for the Artificial Land category in the most populated NUTs 2 regions. Differences are typically small, within 0.3–5% absolute difference for the majority of regions and years assessed.\n\n* Beyond the scope of this analysis, additional work could examine whether the dataset aligns with findings reported in peer-reviewed studies, such as [Fernández-Nogueira and Corbelle-Rico (2018)](https://doi.org/10.1016/j.apgeog.2009.07.003). This would help validate the robustness and reliability of the dataset through comparison with multiple datasets.\n\n* Another aspect of the analysis supporting the dataset's precision is the identification of trade-offs between urban expansion and the decline of agricultural areas, a well-documented phenomenon. Numerous studies, including observations across multiple European regions, note that urban growth frequently replaces agricultural lands [Bagan and Yamagata (2014)](https://doi.org/10.1088/1748-9326/9/6/064015).\n\n```\n\n![Urbanization_Map_Series_v2.png](attachment:Urbanization_Map_Series_v2.png)", "text_with_prefix": "EQC Quality Assessment: \"Satellite Land Cover completeness for Spatial Planning and Land Management\"\nDataset: satellite-land-cover [CDS]\nAspect: consistency-assessment_q01 | Category: Satellite_ECVs\nSection: Satellite Land Cover completeness for Spatial Planning and Land Management > Quality assessment statement\n---\nThese are the key outcomes of this assessment\n\n* The dataset demonstrates physical consistency and precision among the different datasets as it aligns closely with equivalent statistics from official sources [(EUROSTAT,2025)](https://ec.europa.eu/eurostat/databrowser/bookmark/63d3fd90-2f5b-4bec-99f3-4f88814e2624?lang=en). Specifically, the analysis shows that the Urban/Settlements category closely matches EUROSTAT's statistics for the Artificial Land category in the most populated NUTs 2 regions. Differences are typically small, within 0.3–5% absolute difference for the majority of regions and years assessed.\n\n* Beyond the scope of this analysis, additional work could examine whether the dataset aligns with findings reported in peer-reviewed studies, such as [Fernández-Nogueira and Corbelle-Rico (2018)](https://doi.org/10.1016/j.apgeog.2009.07.003). This would help validate the robustness and reliability of the dataset through comparison with multiple datasets.\n\n* Another aspect of the analysis supporting the dataset's precision is the identification of trade-offs between urban expansion and the decline of agricultural areas, a well-documented phenomenon. Numerous studies, including observations across multiple European regions, note that urban growth frequently replaces agricultural lands [Bagan and Yamagata (2014)](https://doi.org/10.1088/1748-9326/9/6/064015).\n\n```\n\n![Urbanization_Map_Series_v2.png](attachment:Urbanization_Map_Series_v2.png)"} {"chunk_id": "satellite_satellite-land-cover_consistency-assessment_q01__7d6a0c82a7d1", "report_id": "satellite_satellite-land-cover_consistency-assessment_q01", "dataset_id": "satellite-land-cover", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency-assessment_q01", "aspect_base": "consistency-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Land Cover completeness for Spatial Planning and Land Management > Methodology", "title": "Satellite Land Cover completeness for Spatial Planning and Land Management", "chunk_index": 3, "token_count": 109, "text_raw": "**[](code-section-1)**\n\n**[](code-section-2)**\n\n**[](code-section-3)**\n\n**[](code-section-4)**\n\n**[](code-section-5)**\n\n**[](code-section-6)**", "text_with_prefix": "EQC Quality Assessment: \"Satellite Land Cover completeness for Spatial Planning and Land Management\"\nDataset: satellite-land-cover [CDS]\nAspect: consistency-assessment_q01 | Category: Satellite_ECVs\nSection: Satellite Land Cover completeness for Spatial Planning and Land Management > Methodology\n---\n**[](code-section-1)**\n\n**[](code-section-2)**\n\n**[](code-section-3)**\n\n**[](code-section-4)**\n\n**[](code-section-5)**\n\n**[](code-section-6)**"} {"chunk_id": "satellite_satellite-land-cover_consistency-assessment_q01__48c1d910b09f", "report_id": "satellite_satellite-land-cover_consistency-assessment_q01", "dataset_id": "satellite-land-cover", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency-assessment_q01", "aspect_base": "consistency-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Land Cover completeness for Spatial Planning and Land Management > Analysis and results > 1. Define the AoI, search and download LC data.", "title": "Satellite Land Cover completeness for Spatial Planning and Land Management", "chunk_index": 4, "token_count": 112, "text_raw": "Before we begin we must prepare our environment. This includes installing the Application Programming Interface (API) of the CDS, and importing the various python libraries that we will need.", "text_with_prefix": "EQC Quality Assessment: \"Satellite Land Cover completeness for Spatial Planning and Land Management\"\nDataset: satellite-land-cover [CDS]\nAspect: consistency-assessment_q01 | Category: Satellite_ECVs\nSection: Satellite Land Cover completeness for Spatial Planning and Land Management > Analysis and results > 1. Define the AoI, search and download LC data.\n---\nBefore we begin we must prepare our environment. This includes installing the Application Programming Interface (API) of the CDS, and importing the various python libraries that we will need."} {"chunk_id": "satellite_satellite-land-cover_consistency-assessment_q01__ed6e1add15cf", "report_id": "satellite_satellite-land-cover_consistency-assessment_q01", "dataset_id": "satellite-land-cover", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency-assessment_q01", "aspect_base": "consistency-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Land Cover completeness for Spatial Planning and Land Management > Analysis and results > 1. Define the AoI, search and download LC data. > Import all the libraries/packages", "title": "Satellite Land Cover completeness for Spatial Planning and Land Management", "chunk_index": 5, "token_count": 130, "text_raw": "We will be working with data in NetCDF format. To best handle this type of data we will use libraries for working with multidimensional arrays, in particular Xarray. \nWe will also need libraries for plotting and viewing data.", "text_with_prefix": "EQC Quality Assessment: \"Satellite Land Cover completeness for Spatial Planning and Land Management\"\nDataset: satellite-land-cover [CDS]\nAspect: consistency-assessment_q01 | Category: Satellite_ECVs\nSection: Satellite Land Cover completeness for Spatial Planning and Land Management > Analysis and results > 1. Define the AoI, search and download LC data. > Import all the libraries/packages\n---\nWe will be working with data in NetCDF format. To best handle this type of data we will use libraries for working with multidimensional arrays, in particular Xarray. \nWe will also need libraries for plotting and viewing data."} {"chunk_id": "satellite_satellite-land-cover_consistency-assessment_q01__82751f6f201f", "report_id": "satellite_satellite-land-cover_consistency-assessment_q01", "dataset_id": "satellite-land-cover", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency-assessment_q01", "aspect_base": "consistency-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Land Cover completeness for Spatial Planning and Land Management > Analysis and results > 1. Define the AoI, search and download LC data. > Data Overview", "title": "Satellite Land Cover completeness for Spatial Planning and Land Management", "chunk_index": 6, "token_count": 824, "text_raw": "To search for data, visit the CDS website: http://cds.climate.copernicus.eu Here you can search for 'Satellite observations land' using the search bar. The data we need for this tutorial is the ***Land cover classification gridded maps from 1992 to present derived from satellite observations***. This catalogue entry provides global Land Cover Classification (LCC) maps with a very high spatial resolution, with a L4 processing level, on an annual basis with a one-year delay, following the [Global Climate Observing System (GCOS) convention requirements](https://library.wmo.int/records/item/58111-the-2022-gcos-ecvs-requirements-gcos-245), under which land cover is recognised as an Essential Climate Variable (ECV) due to its importance for monitoring climate change and its impacts on terrestrial ecosystems. These Land Cover (LC) maps correspond to a global classification scheme encompassing 22 land cover classes.\n\n**Note:** Although the dataset is officially described as having 22 global land cover classes, many of these classes are internally subdivided into more detailed subcategories (e.g., types of cropland, forest canopy structure, types of shrubland). These subdivisions allow for greater ecological detail but can be aggregated back into the 22 main classes for standard analysis and intercomparison purposes.\n\nData specifications for this use case:\n- **Years:** 2009, 2012, 2015, 2018 and 2022\n- **Versions:** v2.0.7 for years up to 2015; v2.1.1 for 2018\n- **Format:** Downloaded as .zip files\n\nAt the end of the data request form on CDS, select `Show API request` to generate Python code, which can be pasted directly into a Jupyter Notebook cell. Running this cell will retrieve the requested files, provided that you have accepted the dataset's `terms and conditions` on the CDS platform. It is advisable to define the desired time period and Area of Interest (AoI) explicitly when preparing the API request, as shown in the example cells below.\n\nUrban areas are represented by **class 190: Urban areas** on the LC dataset. In the IPCC aggregation used by the product, this class belongs to the **Settlement** category. In the underlying UN/FAO Land Cover Classification System coding, class 190 is mapped to B15 Artificial Surfaces, labelled as “Artificial surfaces and associated areas.” The product guide does not provide a more detailed breakdown of what this class includes, such as specific built-up elements or infrastructure types. It only identifies the class as Urban areas / Artificial Surfaces and notes that it relies on external urban reference datasets, specifically the Global Human Settlement Layer and the Global Urban Footprint.\n\nYears to download\n\nList of requests to retrieve data\n\nDownload and regionalise by AoI\n\n```text\n100%|██████████| 5/5 [00:00<00:00, 14.65it/s]\n/data/common/miniforge3/envs/wp5/lib/python3.12/site-packages/earthkit/data/readers/netcdf/fieldlist.py:202: FutureWarning:\n\nIn a future version of xarray the default value for data_vars will change from data_vars='all' to data_vars=None. This is likely to lead to different results when multiple datasets have matching variables with overlapping values. To opt in to new defaults and get rid of these warnings now use `set_options(use_new_combine_kwarg_defaults=True) or set data_vars explicitly.\n```\n\n(code-section-2)=", "text_with_prefix": "EQC Quality Assessment: \"Satellite Land Cover completeness for Spatial Planning and Land Management\"\nDataset: satellite-land-cover [CDS]\nAspect: consistency-assessment_q01 | Category: Satellite_ECVs\nSection: Satellite Land Cover completeness for Spatial Planning and Land Management > Analysis and results > 1. Define the AoI, search and download LC data. > Data Overview\n---\nTo search for data, visit the CDS website: http://cds.climate.copernicus.eu Here you can search for 'Satellite observations land' using the search bar. The data we need for this tutorial is the ***Land cover classification gridded maps from 1992 to present derived from satellite observations***. This catalogue entry provides global Land Cover Classification (LCC) maps with a very high spatial resolution, with a L4 processing level, on an annual basis with a one-year delay, following the [Global Climate Observing System (GCOS) convention requirements](https://library.wmo.int/records/item/58111-the-2022-gcos-ecvs-requirements-gcos-245), under which land cover is recognised as an Essential Climate Variable (ECV) due to its importance for monitoring climate change and its impacts on terrestrial ecosystems. These Land Cover (LC) maps correspond to a global classification scheme encompassing 22 land cover classes.\n\n**Note:** Although the dataset is officially described as having 22 global land cover classes, many of these classes are internally subdivided into more detailed subcategories (e.g., types of cropland, forest canopy structure, types of shrubland). These subdivisions allow for greater ecological detail but can be aggregated back into the 22 main classes for standard analysis and intercomparison purposes.\n\nData specifications for this use case:\n- **Years:** 2009, 2012, 2015, 2018 and 2022\n- **Versions:** v2.0.7 for years up to 2015; v2.1.1 for 2018\n- **Format:** Downloaded as .zip files\n\nAt the end of the data request form on CDS, select `Show API request` to generate Python code, which can be pasted directly into a Jupyter Notebook cell. Running this cell will retrieve the requested files, provided that you have accepted the dataset's `terms and conditions` on the CDS platform. It is advisable to define the desired time period and Area of Interest (AoI) explicitly when preparing the API request, as shown in the example cells below.\n\nUrban areas are represented by **class 190: Urban areas** on the LC dataset. In the IPCC aggregation used by the product, this class belongs to the **Settlement** category. In the underlying UN/FAO Land Cover Classification System coding, class 190 is mapped to B15 Artificial Surfaces, labelled as “Artificial surfaces and associated areas.” The product guide does not provide a more detailed breakdown of what this class includes, such as specific built-up elements or infrastructure types. It only identifies the class as Urban areas / Artificial Surfaces and notes that it relies on external urban reference datasets, specifically the Global Human Settlement Layer and the Global Urban Footprint.\n\nYears to download\n\nList of requests to retrieve data\n\nDownload and regionalise by AoI\n\n```text\n100%|██████████| 5/5 [00:00<00:00, 14.65it/s]\n/data/common/miniforge3/envs/wp5/lib/python3.12/site-packages/earthkit/data/readers/netcdf/fieldlist.py:202: FutureWarning:\n\nIn a future version of xarray the default value for data_vars will change from data_vars='all' to data_vars=None. This is likely to lead to different results when multiple datasets have matching variables with overlapping values. To opt in to new defaults and get rid of these warnings now use `set_options(use_new_combine_kwarg_defaults=True) or set data_vars explicitly.\n```\n\n(code-section-2)="} {"chunk_id": "satellite_satellite-land-cover_consistency-assessment_q01__41d4e2ff3685", "report_id": "satellite_satellite-land-cover_consistency-assessment_q01", "dataset_id": "satellite-land-cover", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency-assessment_q01", "aspect_base": "consistency-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Land Cover completeness for Spatial Planning and Land Management > Analysis and results > 2. Inspect and view data", "title": "Satellite Land Cover completeness for Spatial Planning and Land Management", "chunk_index": 7, "token_count": 113, "text_raw": "Now that we have downloaded the data, we can inspect it. In the previous step, the data were downloaded and automatically loaded into an Xarray dataset using the `download.download_and_transform()` helper function.", "text_with_prefix": "EQC Quality Assessment: \"Satellite Land Cover completeness for Spatial Planning and Land Management\"\nDataset: satellite-land-cover [CDS]\nAspect: consistency-assessment_q01 | Category: Satellite_ECVs\nSection: Satellite Land Cover completeness for Spatial Planning and Land Management > Analysis and results > 2. Inspect and view data\n---\nNow that we have downloaded the data, we can inspect it. In the previous step, the data were downloaded and automatically loaded into an Xarray dataset using the `download.download_and_transform()` helper function."} {"chunk_id": "satellite_satellite-land-cover_consistency-assessment_q01__c0dae4fc5cb0", "report_id": "satellite_satellite-land-cover_consistency-assessment_q01", "dataset_id": "satellite-land-cover", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency-assessment_q01", "aspect_base": "consistency-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Land Cover completeness for Spatial Planning and Land Management > Analysis and results > 2. Inspect and view data > Label Color Definition and Class Correspondence", "title": "Satellite Land Cover completeness for Spatial Planning and Land Management", "chunk_index": 8, "token_count": 919, "text_raw": "To facilitate visual inspection of the Land Cover (LC) classes, we define a dictionary containing each class label (the \"keys\"), the corresponding color code (the \"colors\"), and the associated numeric identifier (the \"values\"). In addition, we create a second dictionary to establish the correspondence between the original land cover classes provided in the metadata and the aggregated IPCC classes, as described in the Product User Guide (see resources).\n\nDefine LC labels dictionary\nHelper function to format the labels text\n\n```text\n{'No Data': ('#000000', np.uint8(0)),\n 'Cropland Rainfed': ('#ffff64', np.uint8(10)),\n 'Cropland Rainfed Herbaceous Cover': ('#ffff64', np.uint8(11)),\n 'Cropland Rainfed Tree Or Shrub Cover': ('#ffff00', np.uint8(12)),\n 'Cropland Irrigated': ('#aaf0f0', np.uint8(20)),\n 'Mosaic Cropland': ('#dcf064', np.uint8(30)),\n 'Mosaic Natural Vegetation': ('#c8c864', np.uint8(40)),\n 'Tree Broadleaved Evergreen Closed To Open': ('#006400', np.uint8(50)),\n 'Tree Broadleaved Deciduous Closed To Open': ('#00a000', np.uint8(60)),\n 'Tree Broadleaved Deciduous Closed': ('#00a000', np.uint8(61)),\n 'Tree Broadleaved Deciduous Open': ('#aac800', np.uint8(62)),\n 'Tree Needleleaved Evergreen Closed To Open': ('#003c00', np.uint8(70)),\n 'Tree Needleleaved Evergreen Closed': ('#003c00', np.uint8(71)),\n 'Tree Needleleaved Evergreen Open': ('#005000', np.uint8(72)),\n 'Tree Needleleaved Deciduous Closed To Open': ('#285000', np.uint8(80)),\n 'Tree Needleleaved Deciduous Closed': ('#285000', np.uint8(81)),\n 'Tree Needleleaved Deciduous Open': ('#286400', np.uint8(82)),\n 'Tree Mixed': ('#788200', np.uint8(90)),\n 'Mosaic Tree And Shrub': ('#8ca000', np.uint8(100)),\n 'Mosaic Herbaceous': ('#be9600', np.uint8(110)),\n 'Shrubland': ('#966400', np.uint8(120)),\n 'Shrubland Evergreen': ('#966400', np.uint8(121)),\n 'Shrubland Deciduous': ('#966400', np.uint8(122)),\n 'Grassland': ('#ffb432', np.uint8(130)),\n 'Lichens And Mosses': ('#ffdcd2', np.uint8(140)),\n 'Sparse Vegetation': ('#ffebaf', np.uint8(150)),\n 'Sparse Tree': ('#ffc864', np.uint8(151)),\n 'Sparse Shrub': ('#ffd278', np.uint8(152)),\n 'Sparse Herbaceous': ('#ffebaf', np.uint8(153)),\n 'Tree Cover Flooded Fresh Or Brakish Water': ('#00785a', np.uint8(160)),\n 'Tree Cover Flooded Saline Water': ('#009678', np.uint8(170)),\n 'Shrub Or Herbaceous Cover Flooded': ('#00dc82', np.uint8(180)),\n 'Urban': ('#c31400', np.uint8(190)),\n 'Bare Areas': ('#fff5d7', np.uint8(200)),\n 'Bare Areas Consolidated': ('#dcdcdc', np.uint8(201)),\n 'Bare Areas Unconsolidated': ('#fff5d7', np.uint8(202)),\n 'Water': ('#0046c8', np.uint8(210)),\n 'Snow And Ice': ('#ffffff', np.uint8(220))}\n```\n\nDefine IPCC labels dictionary", "text_with_prefix": "EQC Quality Assessment: \"Satellite Land Cover completeness for Spatial Planning and Land Management\"\nDataset: satellite-land-cover [CDS]\nAspect: consistency-assessment_q01 | Category: Satellite_ECVs\nSection: Satellite Land Cover completeness for Spatial Planning and Land Management > Analysis and results > 2. Inspect and view data > Label Color Definition and Class Correspondence\n---\nTo facilitate visual inspection of the Land Cover (LC) classes, we define a dictionary containing each class label (the \"keys\"), the corresponding color code (the \"colors\"), and the associated numeric identifier (the \"values\"). In addition, we create a second dictionary to establish the correspondence between the original land cover classes provided in the metadata and the aggregated IPCC classes, as described in the Product User Guide (see resources).\n\nDefine LC labels dictionary\nHelper function to format the labels text\n\n```text\n{'No Data': ('#000000', np.uint8(0)),\n 'Cropland Rainfed': ('#ffff64', np.uint8(10)),\n 'Cropland Rainfed Herbaceous Cover': ('#ffff64', np.uint8(11)),\n 'Cropland Rainfed Tree Or Shrub Cover': ('#ffff00', np.uint8(12)),\n 'Cropland Irrigated': ('#aaf0f0', np.uint8(20)),\n 'Mosaic Cropland': ('#dcf064', np.uint8(30)),\n 'Mosaic Natural Vegetation': ('#c8c864', np.uint8(40)),\n 'Tree Broadleaved Evergreen Closed To Open': ('#006400', np.uint8(50)),\n 'Tree Broadleaved Deciduous Closed To Open': ('#00a000', np.uint8(60)),\n 'Tree Broadleaved Deciduous Closed': ('#00a000', np.uint8(61)),\n 'Tree Broadleaved Deciduous Open': ('#aac800', np.uint8(62)),\n 'Tree Needleleaved Evergreen Closed To Open': ('#003c00', np.uint8(70)),\n 'Tree Needleleaved Evergreen Closed': ('#003c00', np.uint8(71)),\n 'Tree Needleleaved Evergreen Open': ('#005000', np.uint8(72)),\n 'Tree Needleleaved Deciduous Closed To Open': ('#285000', np.uint8(80)),\n 'Tree Needleleaved Deciduous Closed': ('#285000', np.uint8(81)),\n 'Tree Needleleaved Deciduous Open': ('#286400', np.uint8(82)),\n 'Tree Mixed': ('#788200', np.uint8(90)),\n 'Mosaic Tree And Shrub': ('#8ca000', np.uint8(100)),\n 'Mosaic Herbaceous': ('#be9600', np.uint8(110)),\n 'Shrubland': ('#966400', np.uint8(120)),\n 'Shrubland Evergreen': ('#966400', np.uint8(121)),\n 'Shrubland Deciduous': ('#966400', np.uint8(122)),\n 'Grassland': ('#ffb432', np.uint8(130)),\n 'Lichens And Mosses': ('#ffdcd2', np.uint8(140)),\n 'Sparse Vegetation': ('#ffebaf', np.uint8(150)),\n 'Sparse Tree': ('#ffc864', np.uint8(151)),\n 'Sparse Shrub': ('#ffd278', np.uint8(152)),\n 'Sparse Herbaceous': ('#ffebaf', np.uint8(153)),\n 'Tree Cover Flooded Fresh Or Brakish Water': ('#00785a', np.uint8(160)),\n 'Tree Cover Flooded Saline Water': ('#009678', np.uint8(170)),\n 'Shrub Or Herbaceous Cover Flooded': ('#00dc82', np.uint8(180)),\n 'Urban': ('#c31400', np.uint8(190)),\n 'Bare Areas': ('#fff5d7', np.uint8(200)),\n 'Bare Areas Consolidated': ('#dcdcdc', np.uint8(201)),\n 'Bare Areas Unconsolidated': ('#fff5d7', np.uint8(202)),\n 'Water': ('#0046c8', np.uint8(210)),\n 'Snow And Ice': ('#ffffff', np.uint8(220))}\n```\n\nDefine IPCC labels dictionary"} {"chunk_id": "satellite_satellite-land-cover_consistency-assessment_q01__bd1df001623b", "report_id": "satellite_satellite-land-cover_consistency-assessment_q01", "dataset_id": "satellite-land-cover", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency-assessment_q01", "aspect_base": "consistency-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Land Cover completeness for Spatial Planning and Land Management > Analysis and results > 2. Inspect and view data > Label Color Definition and Class Correspondence", "title": "Satellite Land Cover completeness for Spatial Planning and Land Management", "chunk_index": 9, "token_count": 309, "text_raw": "8(202)),\n 'Water': ('#0046c8', np.uint8(210)),\n 'Snow And Ice': ('#ffffff', np.uint8(220))}\n```\n\nDefine IPCC labels dictionary\n\n```text\n{'No Data': ('#000000', [0]),\n 'Agriculture': ('#ffffcc', [10, 11, 12, 20, 30, 40]),\n 'Forest': ('#4c9900',\n [50, 60, 61, 62, 70, 71, 72, 80, 81, 82, 90, 100, 160, 170]),\n 'Grassland': ('#ccff99', [110, 130]),\n 'Settlement': ('#ff0000', [190]),\n 'Wetland': ('#99ffff', [180]),\n 'Other': ('#0000ff',\n [120, 121, 122, 140, 150, 151, 152, 153, 200, 201, 202, 210])}\n```", "text_with_prefix": "EQC Quality Assessment: \"Satellite Land Cover completeness for Spatial Planning and Land Management\"\nDataset: satellite-land-cover [CDS]\nAspect: consistency-assessment_q01 | Category: Satellite_ECVs\nSection: Satellite Land Cover completeness for Spatial Planning and Land Management > Analysis and results > 2. Inspect and view data > Label Color Definition and Class Correspondence\n---\n8(202)),\n 'Water': ('#0046c8', np.uint8(210)),\n 'Snow And Ice': ('#ffffff', np.uint8(220))}\n```\n\nDefine IPCC labels dictionary\n\n```text\n{'No Data': ('#000000', [0]),\n 'Agriculture': ('#ffffcc', [10, 11, 12, 20, 30, 40]),\n 'Forest': ('#4c9900',\n [50, 60, 61, 62, 70, 71, 72, 80, 81, 82, 90, 100, 160, 170]),\n 'Grassland': ('#ccff99', [110, 130]),\n 'Settlement': ('#ff0000', [190]),\n 'Wetland': ('#99ffff', [180]),\n 'Other': ('#0000ff',\n [120, 121, 122, 140, 150, 151, 152, 153, 200, 201, 202, 210])}\n```"} {"chunk_id": "satellite_satellite-land-cover_consistency-assessment_q01__6c32b9e0ee65", "report_id": "satellite_satellite-land-cover_consistency-assessment_q01", "dataset_id": "satellite-land-cover", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency-assessment_q01", "aspect_base": "consistency-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Land Cover completeness for Spatial Planning and Land Management > Analysis and results > 2. Inspect and view data > Plot maps", "title": "Satellite Land Cover completeness for Spatial Planning and Land Management", "chunk_index": 10, "token_count": 194, "text_raw": "Having defined the color and legends for the IPCC classes and using the metadata of the dataset to get the colors and legends for each Land Cover class it is now possible to plot our data either with the original colors or with the IPCC previously defined colors.\n\nThe function below plots the LC maps for the year of your choice , using both land cover schemes. From the output, we can already distinguish the differences in classification schemes.\n\nunique codes\nFilter both legends\n1) LCCS map (only present classes in legend)\n2) IPCC map (only present groups in legend)\n\n(code-section-3)=", "text_with_prefix": "EQC Quality Assessment: \"Satellite Land Cover completeness for Spatial Planning and Land Management\"\nDataset: satellite-land-cover [CDS]\nAspect: consistency-assessment_q01 | Category: Satellite_ECVs\nSection: Satellite Land Cover completeness for Spatial Planning and Land Management > Analysis and results > 2. Inspect and view data > Plot maps\n---\nHaving defined the color and legends for the IPCC classes and using the metadata of the dataset to get the colors and legends for each Land Cover class it is now possible to plot our data either with the original colors or with the IPCC previously defined colors.\n\nThe function below plots the LC maps for the year of your choice , using both land cover schemes. From the output, we can already distinguish the differences in classification schemes.\n\nunique codes\nFilter both legends\n1) LCCS map (only present classes in legend)\n2) IPCC map (only present groups in legend)\n\n(code-section-3)="} {"chunk_id": "satellite_satellite-land-cover_consistency-assessment_q01__7b69714921f1", "report_id": "satellite_satellite-land-cover_consistency-assessment_q01", "dataset_id": "satellite-land-cover", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency-assessment_q01", "aspect_base": "consistency-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Land Cover completeness for Spatial Planning and Land Management > Analysis and results > 3. Data preparation and area-based calculations", "title": "Satellite Land Cover completeness for Spatial Planning and Land Management", "chunk_index": 11, "token_count": 293, "text_raw": "To further identify changes in LC patterns, in this user question, Nomenclature of Territorial Units for Statistics (NUTS) 2 will be used, providing the information regarding the main regions/parcels of the Iberian Peninsula.\n\nThe [NUTS](https://ec.europa.eu/eurostat/web/nuts) are a hierarchical system divided into 3 levels. NUTS 1 correspond to major socio-economic regions, NUTS 2 correspond to basic regions for the application of regional policies, and NUTS 3 correspond to small regions for specific diagnoses. Additionally a NUTS 0 level, usually co-incident with national boundaries is also available. The NUTS legislation is periodically amended; therefore multiple years are available for download.\n\nThe step below masks the Land Cover data according to the NUTS 2 boundaries and calculates the area of each pixel (weighted by Latitude). For each NUTS 2 region, we proceed with the analysis and visual inspection of Land Cover areas per class and corresponding percentages during the selected period.", "text_with_prefix": "EQC Quality Assessment: \"Satellite Land Cover completeness for Spatial Planning and Land Management\"\nDataset: satellite-land-cover [CDS]\nAspect: consistency-assessment_q01 | Category: Satellite_ECVs\nSection: Satellite Land Cover completeness for Spatial Planning and Land Management > Analysis and results > 3. Data preparation and area-based calculations\n---\nTo further identify changes in LC patterns, in this user question, Nomenclature of Territorial Units for Statistics (NUTS) 2 will be used, providing the information regarding the main regions/parcels of the Iberian Peninsula.\n\nThe [NUTS](https://ec.europa.eu/eurostat/web/nuts) are a hierarchical system divided into 3 levels. NUTS 1 correspond to major socio-economic regions, NUTS 2 correspond to basic regions for the application of regional policies, and NUTS 3 correspond to small regions for specific diagnoses. Additionally a NUTS 0 level, usually co-incident with national boundaries is also available. The NUTS legislation is periodically amended; therefore multiple years are available for download.\n\nThe step below masks the Land Cover data according to the NUTS 2 boundaries and calculates the area of each pixel (weighted by Latitude). For each NUTS 2 region, we proceed with the analysis and visual inspection of Land Cover areas per class and corresponding percentages during the selected period."} {"chunk_id": "satellite_satellite-land-cover_consistency-assessment_q01__2f083005d2f4", "report_id": "satellite_satellite-land-cover_consistency-assessment_q01", "dataset_id": "satellite-land-cover", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency-assessment_q01", "aspect_base": "consistency-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Land Cover completeness for Spatial Planning and Land Management > Analysis and results > 3. Data preparation and area-based calculations > Mask regions", "title": "Satellite Land Cover completeness for Spatial Planning and Land Management", "chunk_index": 12, "token_count": 198, "text_raw": "First, we need to establish the geometry of the NUTS region (level 2) in order to make the corresponding statistics.\n\nShapefile with regions (NUTS2)\n\nDefine the CRS for the Iberian Peninsula (WGS84)\nData filter: load the shapefile and filter for NUTS2 regions in Spain and Portugal\nCreate a mask for the NUTS2 regions that intersect with the Iberian Peninsula\nApply the mask to the dataset\nOnly keep land areas (not ocean) by applying the mask\nThis ensures that only land pixels are used in the area calculations", "text_with_prefix": "EQC Quality Assessment: \"Satellite Land Cover completeness for Spatial Planning and Land Management\"\nDataset: satellite-land-cover [CDS]\nAspect: consistency-assessment_q01 | Category: Satellite_ECVs\nSection: Satellite Land Cover completeness for Spatial Planning and Land Management > Analysis and results > 3. Data preparation and area-based calculations > Mask regions\n---\nFirst, we need to establish the geometry of the NUTS region (level 2) in order to make the corresponding statistics.\n\nShapefile with regions (NUTS2)\n\nDefine the CRS for the Iberian Peninsula (WGS84)\nData filter: load the shapefile and filter for NUTS2 regions in Spain and Portugal\nCreate a mask for the NUTS2 regions that intersect with the Iberian Peninsula\nApply the mask to the dataset\nOnly keep land areas (not ocean) by applying the mask\nThis ensures that only land pixels are used in the area calculations"} {"chunk_id": "satellite_satellite-land-cover_consistency-assessment_q01__b03cb96fed96", "report_id": "satellite_satellite-land-cover_consistency-assessment_q01", "dataset_id": "satellite-land-cover", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency-assessment_q01", "aspect_base": "consistency-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Land Cover completeness for Spatial Planning and Land Management > Analysis and results > 3. Data preparation and area-based calculations > Compute cell area", "title": "Satellite Land Cover completeness for Spatial Planning and Land Management", "chunk_index": 13, "token_count": 138, "text_raw": "Then, we can calculate the area of each pixel taking into consideration the curvature of the earth (i.e., weighted by Latitude).\n\n[If you want to know more](https://rdrr.io/cran/raster/man/area.html)\n\nCalculate Pixel Area after applying the mask\n\n(code-section-4)=", "text_with_prefix": "EQC Quality Assessment: \"Satellite Land Cover completeness for Spatial Planning and Land Management\"\nDataset: satellite-land-cover [CDS]\nAspect: consistency-assessment_q01 | Category: Satellite_ECVs\nSection: Satellite Land Cover completeness for Spatial Planning and Land Management > Analysis and results > 3. Data preparation and area-based calculations > Compute cell area\n---\nThen, we can calculate the area of each pixel taking into consideration the curvature of the earth (i.e., weighted by Latitude).\n\n[If you want to know more](https://rdrr.io/cran/raster/man/area.html)\n\nCalculate Pixel Area after applying the mask\n\n(code-section-4)="} {"chunk_id": "satellite_satellite-land-cover_consistency-assessment_q01__9040c0ea573d", "report_id": "satellite_satellite-land-cover_consistency-assessment_q01", "dataset_id": "satellite-land-cover", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency-assessment_q01", "aspect_base": "consistency-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Land Cover completeness for Spatial Planning and Land Management > Analysis and results > 4. Analysis and visualisation of Iberian Peninsula land cover changes > Bar Charts", "title": "Satellite Land Cover completeness for Spatial Planning and Land Management", "chunk_index": 14, "token_count": 276, "text_raw": "Having the area calculated and the NUTS 2 regions assigned to each pixel, we can now proceed to create the plots of the LC areas per class, by year. First, let's inspect the total area of each LC class in this AoI. We will use the original LC classes to highlight which ones show more changes.\n\nFuction to calculate class area with original classes\nSelect the data for the given year\nInitialize a dictionary to store the area for each land cover class\nLoop over each unique land cover class (lccs_class)\nCreate a mask for the current LC class\nCalculate the total area for the current LC class\nCalculate areas for 1992 and 2022\n\nCombine area dictionaries into a single dictionary\nEnsure lc_classes excludes NaN\nPrepare area lists for plotting\nGet class names and colors from the mapping dictionaries\nBar plot settings\nUnique hatching patterns for visual separation by year\nPlot bars for each year\nFinal plot customization", "text_with_prefix": "EQC Quality Assessment: \"Satellite Land Cover completeness for Spatial Planning and Land Management\"\nDataset: satellite-land-cover [CDS]\nAspect: consistency-assessment_q01 | Category: Satellite_ECVs\nSection: Satellite Land Cover completeness for Spatial Planning and Land Management > Analysis and results > 4. Analysis and visualisation of Iberian Peninsula land cover changes > Bar Charts\n---\nHaving the area calculated and the NUTS 2 regions assigned to each pixel, we can now proceed to create the plots of the LC areas per class, by year. First, let's inspect the total area of each LC class in this AoI. We will use the original LC classes to highlight which ones show more changes.\n\nFuction to calculate class area with original classes\nSelect the data for the given year\nInitialize a dictionary to store the area for each land cover class\nLoop over each unique land cover class (lccs_class)\nCreate a mask for the current LC class\nCalculate the total area for the current LC class\nCalculate areas for 1992 and 2022\n\nCombine area dictionaries into a single dictionary\nEnsure lc_classes excludes NaN\nPrepare area lists for plotting\nGet class names and colors from the mapping dictionaries\nBar plot settings\nUnique hatching patterns for visual separation by year\nPlot bars for each year\nFinal plot customization"} {"chunk_id": "satellite_satellite-land-cover_consistency-assessment_q01__9b94b5661a4c", "report_id": "satellite_satellite-land-cover_consistency-assessment_q01", "dataset_id": "satellite-land-cover", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency-assessment_q01", "aspect_base": "consistency-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Land Cover completeness for Spatial Planning and Land Management > Analysis and results > 4. Analysis and visualisation of Iberian Peninsula land cover changes > Sankey Diagrams", "title": "Satellite Land Cover completeness for Spatial Planning and Land Management", "chunk_index": 15, "token_count": 487, "text_raw": "In this section we take the land-cover maps for consecutive years and quantify how much area changes from one category to another. For each time step (e.g., 2009→2012, 2012→2015, 2015→2018), every grid cell is compared between the two years. If the class changes, we add that cell’s area (km²) to the corresponding “from → to” transition. This produces a table of transitions in km² for each period. The transitions are then aggregated to broader IPCC land-cover groups (e.g., Forest, Cropland, Grassland, Settlements, etc.) so the results are easier to interpret. Finally, these aggregated transitions are visualised as Sankey diagrams.\n\nEach Sankey diagram corresponds to one period (for example 2009→2012). The left side lists the IPCC groups in the first year and the right side lists the IPCC groups in the second year. The ribbons represent land that changed from one group to another; the thickness of a ribbon is the area that changed (km²). Hovering a ribbon shows the same area in km² and also the percentage of the source group’s total change that went to that destination. Only changes between different groups are shown.\n\nWeighted class transitions (km²) for consecutive years\nBroadcast 1D or 2D area -> (lat, lon) like a single year slice\nClass codes from CF-style flags; exclude nodata_code\nLUT: raw code -> contiguous [0..nC-1], else -1\nAggregate class→class transitions to IPCC groups\nMembership matrix M: (nC x nG)\nSankey: km² + ONLY % of source outflow;\nDenominator for outflow %: sum of non-self transitions per source row (if ignore_self)\nLabels: optionally show totals for sources (outflow only)\nNode hover suppression (no tooltips on nodes)\n\nColoring graphs\n\nPlot Sankey Diagrams", "text_with_prefix": "EQC Quality Assessment: \"Satellite Land Cover completeness for Spatial Planning and Land Management\"\nDataset: satellite-land-cover [CDS]\nAspect: consistency-assessment_q01 | Category: Satellite_ECVs\nSection: Satellite Land Cover completeness for Spatial Planning and Land Management > Analysis and results > 4. Analysis and visualisation of Iberian Peninsula land cover changes > Sankey Diagrams\n---\nIn this section we take the land-cover maps for consecutive years and quantify how much area changes from one category to another. For each time step (e.g., 2009→2012, 2012→2015, 2015→2018), every grid cell is compared between the two years. If the class changes, we add that cell’s area (km²) to the corresponding “from → to” transition. This produces a table of transitions in km² for each period. The transitions are then aggregated to broader IPCC land-cover groups (e.g., Forest, Cropland, Grassland, Settlements, etc.) so the results are easier to interpret. Finally, these aggregated transitions are visualised as Sankey diagrams.\n\nEach Sankey diagram corresponds to one period (for example 2009→2012). The left side lists the IPCC groups in the first year and the right side lists the IPCC groups in the second year. The ribbons represent land that changed from one group to another; the thickness of a ribbon is the area that changed (km²). Hovering a ribbon shows the same area in km² and also the percentage of the source group’s total change that went to that destination. Only changes between different groups are shown.\n\nWeighted class transitions (km²) for consecutive years\nBroadcast 1D or 2D area -> (lat, lon) like a single year slice\nClass codes from CF-style flags; exclude nodata_code\nLUT: raw code -> contiguous [0..nC-1], else -1\nAggregate class→class transitions to IPCC groups\nMembership matrix M: (nC x nG)\nSankey: km² + ONLY % of source outflow;\nDenominator for outflow %: sum of non-self transitions per source row (if ignore_self)\nLabels: optionally show totals for sources (outflow only)\nNode hover suppression (no tooltips on nodes)\n\nColoring graphs\n\nPlot Sankey Diagrams"} {"chunk_id": "satellite_satellite-land-cover_consistency-assessment_q01__70de8b20467f", "report_id": "satellite_satellite-land-cover_consistency-assessment_q01", "dataset_id": "satellite-land-cover", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency-assessment_q01", "aspect_base": "consistency-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Land Cover completeness for Spatial Planning and Land Management > Analysis and results > 4. Analysis and visualisation of Iberian Peninsula land cover changes > Bar chart and Sankey Diagrams Analysis", "title": "Satellite Land Cover completeness for Spatial Planning and Land Management", "chunk_index": 16, "token_count": 639, "text_raw": "The land cover comparison across the selected years (2009, 2012, 2015, 2018 and 2022) shows a clear and persistent increase in urban area. The bar chart provides the net view of these changes, showing a steady rise in the Urban class across all years, while other classes change more gradually without abrupt or implausible shifts. This supports the interpretation that the observed urban growth is systematic over time rather than being driven by isolated anomalies.\n\nThe Sankey transition diagrams provide further evidence that the observed urban growth follows realistic land-use change pathways. When analysed using the percentage share of each source class outflow, Agriculture→Settlement transitions account for 64.46% (2009–2012) and 65.91% (2012–2015) of total agricultural change in the respective periods. In the later intervals, this proportion decreases to 23.26% (2015–2018) and 13.49% (2018–2022), indicating that although agriculture remains a major contributor to new Settlement area in absolute terms, a smaller share of overall agricultural transitions is directed toward urban expansion in the later years.\n\nForest→Settlement transitions represent a consistently smaller proportion of total forest outflow, accounting for 3.58%, 6.81%, 15.50%, and 0.87% across the four respective periods. These values indicate that only a limited fraction of forest change results in direct urban conversion. Grassland→Settlement transitions account for 64.72%, 74.35%, 13.46%, and 3.22% of total grassland outflow across the same periods, indicating that a larger share of grassland change was directed toward urban expansion in the earlier periods than in the later years.\n\nThese outflow-based transition patterns confirm that urban growth primarily interacts with agricultural and grassland dynamics, while forest areas play a comparatively minor direct role in settlement expansion. This behaviour aligns with Feranec et al. (2010), who show that urbanisation flows predominantly originate from agricultural classes at the European scale. Similarly, Bagan and Yamagata (2014) identify cropland as the principal donor class to urban growth in global city analyses. In the Iberian Peninsula, Fernández-Nogueira and Corbelle-Rico (2018) report that increases in artificial surfaces occurred alongside stable or increasing forest area, further supporting the interpretation that urban growth does not primarily proceed through direct forest conversion.\n\nOverall, the combined bar chart and Sankey analysis indicates that the dataset captures urbanisation patterns that are coherent across time, internally consistent, and aligned with established land-use change processes documented in the literature.\n\n(code-section-5)=", "text_with_prefix": "EQC Quality Assessment: \"Satellite Land Cover completeness for Spatial Planning and Land Management\"\nDataset: satellite-land-cover [CDS]\nAspect: consistency-assessment_q01 | Category: Satellite_ECVs\nSection: Satellite Land Cover completeness for Spatial Planning and Land Management > Analysis and results > 4. Analysis and visualisation of Iberian Peninsula land cover changes > Bar chart and Sankey Diagrams Analysis\n---\nThe land cover comparison across the selected years (2009, 2012, 2015, 2018 and 2022) shows a clear and persistent increase in urban area. The bar chart provides the net view of these changes, showing a steady rise in the Urban class across all years, while other classes change more gradually without abrupt or implausible shifts. This supports the interpretation that the observed urban growth is systematic over time rather than being driven by isolated anomalies.\n\nThe Sankey transition diagrams provide further evidence that the observed urban growth follows realistic land-use change pathways. When analysed using the percentage share of each source class outflow, Agriculture→Settlement transitions account for 64.46% (2009–2012) and 65.91% (2012–2015) of total agricultural change in the respective periods. In the later intervals, this proportion decreases to 23.26% (2015–2018) and 13.49% (2018–2022), indicating that although agriculture remains a major contributor to new Settlement area in absolute terms, a smaller share of overall agricultural transitions is directed toward urban expansion in the later years.\n\nForest→Settlement transitions represent a consistently smaller proportion of total forest outflow, accounting for 3.58%, 6.81%, 15.50%, and 0.87% across the four respective periods. These values indicate that only a limited fraction of forest change results in direct urban conversion. Grassland→Settlement transitions account for 64.72%, 74.35%, 13.46%, and 3.22% of total grassland outflow across the same periods, indicating that a larger share of grassland change was directed toward urban expansion in the earlier periods than in the later years.\n\nThese outflow-based transition patterns confirm that urban growth primarily interacts with agricultural and grassland dynamics, while forest areas play a comparatively minor direct role in settlement expansion. This behaviour aligns with Feranec et al. (2010), who show that urbanisation flows predominantly originate from agricultural classes at the European scale. Similarly, Bagan and Yamagata (2014) identify cropland as the principal donor class to urban growth in global city analyses. In the Iberian Peninsula, Fernández-Nogueira and Corbelle-Rico (2018) report that increases in artificial surfaces occurred alongside stable or increasing forest area, further supporting the interpretation that urban growth does not primarily proceed through direct forest conversion.\n\nOverall, the combined bar chart and Sankey analysis indicates that the dataset captures urbanisation patterns that are coherent across time, internally consistent, and aligned with established land-use change processes documented in the literature.\n\n(code-section-5)="} {"chunk_id": "satellite_satellite-land-cover_consistency-assessment_q01__7d3c50cbeeee", "report_id": "satellite_satellite-land-cover_consistency-assessment_q01", "dataset_id": "satellite-land-cover", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency-assessment_q01", "aspect_base": "consistency-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Land Cover completeness for Spatial Planning and Land Management > Analysis and results > 5. Calculate the area percentage and the area percentage change for the most populated regions", "title": "Satellite Land Cover completeness for Spatial Planning and Land Management", "chunk_index": 17, "token_count": 326, "text_raw": "Having identified general urbanisation trends in the Iberian Peninsula, we now focus on specific NUTS2 regions in greater detail. These regions were selected based on their inclusion of the largest urban areas in terms of population, according to the Urban Audit Indicators dataset from the European Commission (Eurostat).\n\nThe selected NUTS2 regions are:\n\n- **Comunidad de Madrid**, which includes the city of Madrid (5,098,717 inhabitants in 2022), the capital of Spain;\n- **Cataluña**, which includes Barcelona (3,755,512 inhabitants in 2022);\n- **Área Metropolitana de Lisboa**, which includes Lisbon (1,872,036 inhabitants in 2022), the capital of Portugal;\n- **Comunitat Valenciana**, which includes Valencia (1,417,464 inhabitants in 2022);\n- **Norte**, which includes Porto (955,864 inhabitants in 2022).\n\nFor each of these regions, we analyse land cover dynamics in terms of **area percentages by IPCC classes**, in order to highlight broader, aggregated land cover changes. Additionally, we compare dataset results with EUROSTAT statistics to assess the consistency between datasets.", "text_with_prefix": "EQC Quality Assessment: \"Satellite Land Cover completeness for Spatial Planning and Land Management\"\nDataset: satellite-land-cover [CDS]\nAspect: consistency-assessment_q01 | Category: Satellite_ECVs\nSection: Satellite Land Cover completeness for Spatial Planning and Land Management > Analysis and results > 5. Calculate the area percentage and the area percentage change for the most populated regions\n---\nHaving identified general urbanisation trends in the Iberian Peninsula, we now focus on specific NUTS2 regions in greater detail. These regions were selected based on their inclusion of the largest urban areas in terms of population, according to the Urban Audit Indicators dataset from the European Commission (Eurostat).\n\nThe selected NUTS2 regions are:\n\n- **Comunidad de Madrid**, which includes the city of Madrid (5,098,717 inhabitants in 2022), the capital of Spain;\n- **Cataluña**, which includes Barcelona (3,755,512 inhabitants in 2022);\n- **Área Metropolitana de Lisboa**, which includes Lisbon (1,872,036 inhabitants in 2022), the capital of Portugal;\n- **Comunitat Valenciana**, which includes Valencia (1,417,464 inhabitants in 2022);\n- **Norte**, which includes Porto (955,864 inhabitants in 2022).\n\nFor each of these regions, we analyse land cover dynamics in terms of **area percentages by IPCC classes**, in order to highlight broader, aggregated land cover changes. Additionally, we compare dataset results with EUROSTAT statistics to assess the consistency between datasets."} {"chunk_id": "satellite_satellite-land-cover_consistency-assessment_q01__bce1c7530458", "report_id": "satellite_satellite-land-cover_consistency-assessment_q01", "dataset_id": "satellite-land-cover", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency-assessment_q01", "aspect_base": "consistency-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Land Cover completeness for Spatial Planning and Land Management > Analysis and results > 5. Calculate the area percentage and the area percentage change for the most populated regions > Calculation of the percentage area, absolute change, and relative percentage change for each IPCC class category", "title": "Satellite Land Cover completeness for Spatial Planning and Land Management", "chunk_index": 18, "token_count": 462, "text_raw": "1. **Area Percentage Coverage**: \n *Example*: In 1992, forest land covered 35% of the total area, while urban areas occupied 10%. By 2022, forest coverage decreased to 30%, and urban areas expanded to 15%. This metric gives the proportion of the total land occupied by each land cover class.\n\n2. **Absolute Percentage Difference**: \n *Example*: In 1992, 10% of the region was classified as agricultural land. By 2022, this had decreased to 8%. The absolute percentage difference in agricultural land coverage is −2% (from 10% in 1992 to 8% in 2022, representing a 2% decrease in total land area occupied by agriculture).\n\n3. **Relative Percentage Difference**: \n *Example*: In 1992, 10% of the area was classified as wetlands. By 2022, wetlands accounted for 12% of the total area. The relative percentage difference is calculated as ((12−10)/10)∗100 = +20%. This means there was a 20% increase in wetland area relative to its size in 1992.\n\nFilter the GeoDataFrame to include only the regions of interest\nExpand cell_area ONCE outside the loops\nPrepare latitude and longitude grids ONCE\n\nCreate region mask using shapely\nCalculate IPCC class % coverage for a given year and region\nCompute absolute and relative differences\nPadded y-axis limits\n\nPloting fuctions\nPlot absolute and relative differences (side-by-side bars)\nAbsolute difference\nRelative difference\n\nYear to check\nStorage dictionary for later reuse\nCalculate everything\nFor plotting (only cov_2009 vs cov_2022)\nCalculate global plot limits\nPlot results", "text_with_prefix": "EQC Quality Assessment: \"Satellite Land Cover completeness for Spatial Planning and Land Management\"\nDataset: satellite-land-cover [CDS]\nAspect: consistency-assessment_q01 | Category: Satellite_ECVs\nSection: Satellite Land Cover completeness for Spatial Planning and Land Management > Analysis and results > 5. Calculate the area percentage and the area percentage change for the most populated regions > Calculation of the percentage area, absolute change, and relative percentage change for each IPCC class category\n---\n1. **Area Percentage Coverage**: \n *Example*: In 1992, forest land covered 35% of the total area, while urban areas occupied 10%. By 2022, forest coverage decreased to 30%, and urban areas expanded to 15%. This metric gives the proportion of the total land occupied by each land cover class.\n\n2. **Absolute Percentage Difference**: \n *Example*: In 1992, 10% of the region was classified as agricultural land. By 2022, this had decreased to 8%. The absolute percentage difference in agricultural land coverage is −2% (from 10% in 1992 to 8% in 2022, representing a 2% decrease in total land area occupied by agriculture).\n\n3. **Relative Percentage Difference**: \n *Example*: In 1992, 10% of the area was classified as wetlands. By 2022, wetlands accounted for 12% of the total area. The relative percentage difference is calculated as ((12−10)/10)∗100 = +20%. This means there was a 20% increase in wetland area relative to its size in 1992.\n\nFilter the GeoDataFrame to include only the regions of interest\nExpand cell_area ONCE outside the loops\nPrepare latitude and longitude grids ONCE\n\nCreate region mask using shapely\nCalculate IPCC class % coverage for a given year and region\nCompute absolute and relative differences\nPadded y-axis limits\n\nPloting fuctions\nPlot absolute and relative differences (side-by-side bars)\nAbsolute difference\nRelative difference\n\nYear to check\nStorage dictionary for later reuse\nCalculate everything\nFor plotting (only cov_2009 vs cov_2022)\nCalculate global plot limits\nPlot results"} {"chunk_id": "satellite_satellite-land-cover_consistency-assessment_q01__8e989ecb46e5", "report_id": "satellite_satellite-land-cover_consistency-assessment_q01", "dataset_id": "satellite-land-cover", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency-assessment_q01", "aspect_base": "consistency-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Land Cover completeness for Spatial Planning and Land Management > Analysis and results > 5. Calculate the area percentage and the area percentage change for the most populated regions > Bar Chart Analysis", "title": "Satellite Land Cover completeness for Spatial Planning and Land Management", "chunk_index": 19, "token_count": 347, "text_raw": "Across the five most populated NUTS2 regions clear land cover change patterns are observed between 2009 and 2022, especially in relation to urban expansion and agricultural decline.\n\n- **Settlement/Urban areas** show a consistent increase in all regions, with the most pronounced growth occurring in Área Metropolitana de Lisboa and Comunidad de Madrid.\n\n- **Agricultural land** has decreased across all regions, with the sharpest reduction seen in Área Metropolitana de Lisboa. The decline appears moderate in Comunidad de Madrid, Cataluña, and Comunitat Valenciana, suggesting a gradual conversion of agricultural land to urban or other uses.\n\n- **Forest and grassland areas** remained mostly stable in several regions, with small increases in Comunitat Valenciana and Cataluña, indicating preservation or limited natural regeneration in non‑urbanised areas. However, in Norte and Valencia, forest cover declined despite only a minimal change of agricultural land. This pattern suggests that the reduction in forest area is linked to the expansion of settlements.\n\n- **Wetlands and other land cover categories** continue to represent a negligible share of total area. Their relative and absolute changes are minimal, with little to no impact on the overall land use structure of the regions.", "text_with_prefix": "EQC Quality Assessment: \"Satellite Land Cover completeness for Spatial Planning and Land Management\"\nDataset: satellite-land-cover [CDS]\nAspect: consistency-assessment_q01 | Category: Satellite_ECVs\nSection: Satellite Land Cover completeness for Spatial Planning and Land Management > Analysis and results > 5. Calculate the area percentage and the area percentage change for the most populated regions > Bar Chart Analysis\n---\nAcross the five most populated NUTS2 regions clear land cover change patterns are observed between 2009 and 2022, especially in relation to urban expansion and agricultural decline.\n\n- **Settlement/Urban areas** show a consistent increase in all regions, with the most pronounced growth occurring in Área Metropolitana de Lisboa and Comunidad de Madrid.\n\n- **Agricultural land** has decreased across all regions, with the sharpest reduction seen in Área Metropolitana de Lisboa. The decline appears moderate in Comunidad de Madrid, Cataluña, and Comunitat Valenciana, suggesting a gradual conversion of agricultural land to urban or other uses.\n\n- **Forest and grassland areas** remained mostly stable in several regions, with small increases in Comunitat Valenciana and Cataluña, indicating preservation or limited natural regeneration in non‑urbanised areas. However, in Norte and Valencia, forest cover declined despite only a minimal change of agricultural land. This pattern suggests that the reduction in forest area is linked to the expansion of settlements.\n\n- **Wetlands and other land cover categories** continue to represent a negligible share of total area. Their relative and absolute changes are minimal, with little to no impact on the overall land use structure of the regions."} {"chunk_id": "satellite_satellite-land-cover_consistency-assessment_q01__1339f84b4710", "report_id": "satellite_satellite-land-cover_consistency-assessment_q01", "dataset_id": "satellite-land-cover", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency-assessment_q01", "aspect_base": "consistency-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Land Cover completeness for Spatial Planning and Land Management > Analysis and results > 5. Calculate the area percentage and the area percentage change for the most populated regions > Comparison with EUROSTAT Data", "title": "Satellite Land Cover completeness for Spatial Planning and Land Management", "chunk_index": 20, "token_count": 505, "text_raw": "To validate the consistency of the C3S Land Cover dataset in the context of urbanisation, we compared the percentage of land classified as Settlement with [EUROSTAT](https://ec.europa.eu/eurostat/databrowser/view/lan_lcv_ovw/default/table?lang=en&category=lan) artificial land statistics for the five most populated NUTS2 regions in the Iberian Peninsula. In addition, key non-urban land cover classes—Forest, Grassland, and Agriculture—are included in the comparison to provide context for the land cover changes associated with urban expansion.\n\nEUROSTAT land-cover and land-use statistics are based on the [LUCAS survey](https://ec.europa.eu/eurostat/web/lucas/methodology), an in-situ survey in which field surveyors classify the observed land cover and visible land use at sampled points using harmonised LUCAS classifications. LUCAS explicitly separates **land cover**, meaning the physical cover of the Earth’s surface, from **land use**, meaning the socio-economic function of the land. The LUCAS land-cover classification is hierarchical and includes **A00 Artificial land** as one of the main land-cover categories.\n\nIn the LUCAS classification, the settlement/artificial-surface category is constructed from the following land-cover classes:\n\n- **A10 Roofed built-up areas**, including buildings and greenhouses;\n- **A20 Artificial non built-up areas**, including sealed area features such as yards, farmyards, cemeteries and car parking areas, as well as linear features such as streets, roads, railways and runways;\n- **A30 Other artificial areas**, including bridges and viaducts, mobile homes, solar panels, power plants, electrical substations, pipelines, sewage plants and open dump sites.\n\nThe table below presents values for the five years for which EUROSTAT land cover data are available (2009, 2012, 2015, 2018 and 2022).\n\nEUROSTAT reference values (percent)\nCombined table for all classes and years", "text_with_prefix": "EQC Quality Assessment: \"Satellite Land Cover completeness for Spatial Planning and Land Management\"\nDataset: satellite-land-cover [CDS]\nAspect: consistency-assessment_q01 | Category: Satellite_ECVs\nSection: Satellite Land Cover completeness for Spatial Planning and Land Management > Analysis and results > 5. Calculate the area percentage and the area percentage change for the most populated regions > Comparison with EUROSTAT Data\n---\nTo validate the consistency of the C3S Land Cover dataset in the context of urbanisation, we compared the percentage of land classified as Settlement with [EUROSTAT](https://ec.europa.eu/eurostat/databrowser/view/lan_lcv_ovw/default/table?lang=en&category=lan) artificial land statistics for the five most populated NUTS2 regions in the Iberian Peninsula. In addition, key non-urban land cover classes—Forest, Grassland, and Agriculture—are included in the comparison to provide context for the land cover changes associated with urban expansion.\n\nEUROSTAT land-cover and land-use statistics are based on the [LUCAS survey](https://ec.europa.eu/eurostat/web/lucas/methodology), an in-situ survey in which field surveyors classify the observed land cover and visible land use at sampled points using harmonised LUCAS classifications. LUCAS explicitly separates **land cover**, meaning the physical cover of the Earth’s surface, from **land use**, meaning the socio-economic function of the land. The LUCAS land-cover classification is hierarchical and includes **A00 Artificial land** as one of the main land-cover categories.\n\nIn the LUCAS classification, the settlement/artificial-surface category is constructed from the following land-cover classes:\n\n- **A10 Roofed built-up areas**, including buildings and greenhouses;\n- **A20 Artificial non built-up areas**, including sealed area features such as yards, farmyards, cemeteries and car parking areas, as well as linear features such as streets, roads, railways and runways;\n- **A30 Other artificial areas**, including bridges and viaducts, mobile homes, solar panels, power plants, electrical substations, pipelines, sewage plants and open dump sites.\n\nThe table below presents values for the five years for which EUROSTAT land cover data are available (2009, 2012, 2015, 2018 and 2022).\n\nEUROSTAT reference values (percent)\nCombined table for all classes and years"} {"chunk_id": "satellite_satellite-land-cover_consistency-assessment_q01__d0f9930cc6de", "report_id": "satellite_satellite-land-cover_consistency-assessment_q01", "dataset_id": "satellite-land-cover", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency-assessment_q01", "aspect_base": "consistency-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Land Cover completeness for Spatial Planning and Land Management > Analysis and results > 5. Calculate the area percentage and the area percentage change for the most populated regions > Comparison with EUROSTAT Data", "title": "Satellite Land Cover completeness for Spatial Planning and Land Management", "chunk_index": 21, "token_count": 1108, "text_raw": "and 2022).\n\nEUROSTAT reference values (percent)\nCombined table for all classes and years\n\n```text\nRegion Year C3S Settlement (%) EUROSTAT Artificial (%) Δ Settlement (%) C3S Forest (%) EUROSTAT Woodland (%) Δ Forest (%) C3S Grassland (%) EUROSTAT Grassland (%) Δ Grassland (%) C3S Agriculture (%) EUROSTAT Cropland (%) Δ Agriculture (%)\n Cataluña 2009 3.29 6.3 -3.01 45.66 41.5 4.16 4.98 13.1 -8.12 40.30 21.7 18.60\n Cataluña 2012 3.49 6.4 -2.91 45.52 44.8 0.72 4.99 12.8 -7.81 40.21 19.5 20.71\n Cataluña 2015 3.71 6.5 -2.79 45.42 44.8 0.62 5.01 13.4 -8.39 40.08 18.4 21.68\n Cataluña 2018 3.93 5.3 -1.37 45.51 49.6 -4.09 5.04 8.9 -3.86 39.79 24.4 15.39\n Cataluña 2022 3.98 5.9 -1.92 45.23 48.2 -2.97 5.16 9.2 -4.04 39.84 19.9 19.94\n Comunidad de Madrid 2009 9.09 9.7 -0.61 27.87 15.3 12.57 14.63 25.3 -10.67 43.16 15.5 27.66\n Comunidad de Madrid 2012 9.62 10.1 -0.48 27.71 20.1 7.61 14.55 24.4 -9.85 42.83 14.0 28.83\n Comunidad de Madrid 2015 10.30 10.6 -0.30 27.44 24.1 3.34 14.43 24.9 -10.47 42.51 15.6 26.91\n Comunidad de Madrid 2018 12.23 11.7 0.53 28.62 30.6 -1.98 13.65 20.4 -6.75 40.58 17.7 22.88\n Comunidad de Madrid 2022 12.47 11.5 0.97 28.80 32.1 -3.30 13.50 17.9 -4.40 40.39 15.1 25.29\n Comunitat Valenciana 2009 3.16 6.1 -2.94 36.00 23.4 12.60 5.96 13.4 -7.44 38.93 11.4 27.53\n Comunitat Valenciana 2012 3.46 6.2 -2.74 35.29 29.2 6.09 5.94 14.1 -8.16 38.62 14.0 24.62\n Comunitat Valenciana 2015 3.86 6.5 -2.64 35.45 29.8 5.65 5.90 14.0 -8.10 38.19 12.2 25.99\n Comunitat Valenciana 2018 4.58", "text_with_prefix": "EQC Quality Assessment: \"Satellite Land Cover completeness for Spatial Planning and Land Management\"\nDataset: satellite-land-cover [CDS]\nAspect: consistency-assessment_q01 | Category: Satellite_ECVs\nSection: Satellite Land Cover completeness for Spatial Planning and Land Management > Analysis and results > 5. Calculate the area percentage and the area percentage change for the most populated regions > Comparison with EUROSTAT Data\n---\nand 2022).\n\nEUROSTAT reference values (percent)\nCombined table for all classes and years\n\n```text\nRegion Year C3S Settlement (%) EUROSTAT Artificial (%) Δ Settlement (%) C3S Forest (%) EUROSTAT Woodland (%) Δ Forest (%) C3S Grassland (%) EUROSTAT Grassland (%) Δ Grassland (%) C3S Agriculture (%) EUROSTAT Cropland (%) Δ Agriculture (%)\n Cataluña 2009 3.29 6.3 -3.01 45.66 41.5 4.16 4.98 13.1 -8.12 40.30 21.7 18.60\n Cataluña 2012 3.49 6.4 -2.91 45.52 44.8 0.72 4.99 12.8 -7.81 40.21 19.5 20.71\n Cataluña 2015 3.71 6.5 -2.79 45.42 44.8 0.62 5.01 13.4 -8.39 40.08 18.4 21.68\n Cataluña 2018 3.93 5.3 -1.37 45.51 49.6 -4.09 5.04 8.9 -3.86 39.79 24.4 15.39\n Cataluña 2022 3.98 5.9 -1.92 45.23 48.2 -2.97 5.16 9.2 -4.04 39.84 19.9 19.94\n Comunidad de Madrid 2009 9.09 9.7 -0.61 27.87 15.3 12.57 14.63 25.3 -10.67 43.16 15.5 27.66\n Comunidad de Madrid 2012 9.62 10.1 -0.48 27.71 20.1 7.61 14.55 24.4 -9.85 42.83 14.0 28.83\n Comunidad de Madrid 2015 10.30 10.6 -0.30 27.44 24.1 3.34 14.43 24.9 -10.47 42.51 15.6 26.91\n Comunidad de Madrid 2018 12.23 11.7 0.53 28.62 30.6 -1.98 13.65 20.4 -6.75 40.58 17.7 22.88\n Comunidad de Madrid 2022 12.47 11.5 0.97 28.80 32.1 -3.30 13.50 17.9 -4.40 40.39 15.1 25.29\n Comunitat Valenciana 2009 3.16 6.1 -2.94 36.00 23.4 12.60 5.96 13.4 -7.44 38.93 11.4 27.53\n Comunitat Valenciana 2012 3.46 6.2 -2.74 35.29 29.2 6.09 5.94 14.1 -8.16 38.62 14.0 24.62\n Comunitat Valenciana 2015 3.86 6.5 -2.64 35.45 29.8 5.65 5.90 14.0 -8.10 38.19 12.2 25.99\n Comunitat Valenciana 2018 4.58"} {"chunk_id": "satellite_satellite-land-cover_consistency-assessment_q01__1ff3f59f4b3a", "report_id": "satellite_satellite-land-cover_consistency-assessment_q01", "dataset_id": "satellite-land-cover", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency-assessment_q01", "aspect_base": "consistency-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Land Cover completeness for Spatial Planning and Land Management > Analysis and results > 5. Calculate the area percentage and the area percentage change for the most populated regions > Comparison with EUROSTAT Data", "title": "Satellite Land Cover completeness for Spatial Planning and Land Management", "chunk_index": 22, "token_count": 972, "text_raw": "5.65 5.90 14.0 -8.10 38.19 12.2 25.99\n Comunitat Valenciana 2018 4.58\n\n 7.2 -2.62 35.66 35.5 0.16 5.77 5.6 0.17 37.54 25.8 11.74\n Comunitat Valenciana 2022 4.67 6.9 -2.23 36.03 34.0 2.03 5.67 6.6 -0.93 37.43 25.6 11.83\n Norte 2009 4.33 6.8 -2.47 50.91 22.6 28.31 1.63 17.7 -16.07 40.17 14.0 26.17\n Norte 2012 4.68 6.9 -2.22 50.32 24.1 26.22 1.70 18.2 -16.50 40.36 12.2 28.16\n Norte 2015 5.08 7.0 -1.92 49.55 26.1 23.45 1.79 16.9 -15.11 40.62 11.5 29.12\n Norte 2018 5.14 9.4 -4.26 50.07 28.9 21.17 1.77 14.3 -12.53 39.99 18.0 21.99\n Norte 2022 5.25 8.0 -2.75 50.32 26.7 23.62 1.73 13.3 -11.57 39.70 19.1 20.60\nÁrea Metropolitana de Lisboa 2009 15.46 14.6 0.86 18.62 11.4 7.22 0.12 21.4 -21.28 50.80 18.5 32.30\nÁrea Metropolitana de Lisboa 2012 16.58 15.4 1.18 18.58 22.1 -3.52 0.12 25.7 -25.58 49.77 14.9 34.87\nÁrea Metropolitana de Lisboa 2015 17.66 15.8 1.86 18.62 23.2 -4.58 0.12 22.7 -22.58 48.64 13.6 35.04\nÁrea Metropolitana de Lisboa 2018 18.12 18.9 -0.78 19.02 28.1 -9.08 0.11 16.5 -16.39 47.82 17.4 30.42\nÁrea Metropolitana de Lisboa 2022 18.25 17.7 0.55 19.35 24.5 -5.15 0.11 22.2 -22.09 47.34 15.9 31.44\n```", "text_with_prefix": "EQC Quality Assessment: \"Satellite Land Cover completeness for Spatial Planning and Land Management\"\nDataset: satellite-land-cover [CDS]\nAspect: consistency-assessment_q01 | Category: Satellite_ECVs\nSection: Satellite Land Cover completeness for Spatial Planning and Land Management > Analysis and results > 5. Calculate the area percentage and the area percentage change for the most populated regions > Comparison with EUROSTAT Data\n---\n5.65 5.90 14.0 -8.10 38.19 12.2 25.99\n Comunitat Valenciana 2018 4.58\n\n 7.2 -2.62 35.66 35.5 0.16 5.77 5.6 0.17 37.54 25.8 11.74\n Comunitat Valenciana 2022 4.67 6.9 -2.23 36.03 34.0 2.03 5.67 6.6 -0.93 37.43 25.6 11.83\n Norte 2009 4.33 6.8 -2.47 50.91 22.6 28.31 1.63 17.7 -16.07 40.17 14.0 26.17\n Norte 2012 4.68 6.9 -2.22 50.32 24.1 26.22 1.70 18.2 -16.50 40.36 12.2 28.16\n Norte 2015 5.08 7.0 -1.92 49.55 26.1 23.45 1.79 16.9 -15.11 40.62 11.5 29.12\n Norte 2018 5.14 9.4 -4.26 50.07 28.9 21.17 1.77 14.3 -12.53 39.99 18.0 21.99\n Norte 2022 5.25 8.0 -2.75 50.32 26.7 23.62 1.73 13.3 -11.57 39.70 19.1 20.60\nÁrea Metropolitana de Lisboa 2009 15.46 14.6 0.86 18.62 11.4 7.22 0.12 21.4 -21.28 50.80 18.5 32.30\nÁrea Metropolitana de Lisboa 2012 16.58 15.4 1.18 18.58 22.1 -3.52 0.12 25.7 -25.58 49.77 14.9 34.87\nÁrea Metropolitana de Lisboa 2015 17.66 15.8 1.86 18.62 23.2 -4.58 0.12 22.7 -22.58 48.64 13.6 35.04\nÁrea Metropolitana de Lisboa 2018 18.12 18.9 -0.78 19.02 28.1 -9.08 0.11 16.5 -16.39 47.82 17.4 30.42\nÁrea Metropolitana de Lisboa 2022 18.25 17.7 0.55 19.35 24.5 -5.15 0.11 22.2 -22.09 47.34 15.9 31.44\n```"} {"chunk_id": "satellite_satellite-land-cover_consistency-assessment_q01__b0f101da6834", "report_id": "satellite_satellite-land-cover_consistency-assessment_q01", "dataset_id": "satellite-land-cover", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency-assessment_q01", "aspect_base": "consistency-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Land Cover completeness for Spatial Planning and Land Management > Analysis and results > 5. Calculate the area percentage and the area percentage change for the most populated regions > Comparison with EUROSTAT Data", "title": "Satellite Land Cover completeness for Spatial Planning and Land Management", "chunk_index": 23, "token_count": 810, "text_raw": "24.5 -5.15 0.11 22.2 -22.09 47.34 15.9 31.44\n```\n\nAcross all regions, both datasets indicate a systematic increase in urban land over time, with comparable magnitudes and trajectories. Absolute differences in settlement/artificial land percentages are generally small, particularly in highly urbanised regions. For Comunidad de Madrid, differences remain below ±1% throughout the period, while Área Metropolitana de Lisboa shows differences mostly within ±2%, despite its higher overall urban fraction.\n\nIn regions with more dispersed or heterogeneous urban structures—such as Cataluña, Comunitat Valenciana, and Norte—C3S settlement values are consistently lower than EUROSTAT artificial land by approximately 1–3%. This offset is stable across years and likely reflects both definitional differences and differences in dataset construction, including spatial resolution. Because C3S is a satellite-derived 300 m gridded product, small or fragmented artificial surfaces may be represented differently from EUROSTAT/LUCAS artificial land, which is based on in-situ point observations aggregated statistically. Crucially, both datasets capture similar rates and directions of urban expansion, indicating that urbanisation dynamics are robustly identified despite these differences.\n\nBetween 2009 and 2022, urban land increased in all regions in both datasets, with comparable absolute changes. For example, in Comunidad de Madrid, C3S settlement increased by +3.38 percentage points (from 9.09% to 12.47%), while EUROSTAT artificial land increased by +1.8 percentage points (from 9.7% to 11.5%). Similarly, in Área Metropolitana de Lisboa, C3S reports an increase of +2.79 percentage points, compared to +3.1 percentage points in EUROSTAT. These comparable magnitudes indicate that both datasets capture a similar volume of land converted into urban use.\n\nThe land cover losses associated with this urban expansion are also consistent in direction. Across all regions, urban growth coincides primarily with reductions in agricultural and open land classes. For instance, in Cataluña, the cumulative increase in settlement between 2009 and 2022 (approximately +0.7 percentage points in C3S) corresponds to a decrease in agricultural land of around –0.5 percentage points, while EUROSTAT reports a comparable decline in cropland over the same period. Similar patterns are observed in Comunitat Valenciana and Norte, where modest but persistent urban gains are accompanied by declining agricultural shares in both datasets.\n\nAlthough C3S and EUROSTAT differ in how agricultural, grassland, and mixed-use areas are classified, these differences do not alter the identification of urban-driven land take. Both datasets indicate that urban expansion primarily occurs at the expense of agricultural and open land, while forest areas remain comparatively stable. This agreement confirms that, in terms of urbanisation impact, the datasets provide equivalent signals of land cover loss.\n\nOverall, the comparison demonstrates that the C3S Land Cover dataset is consistent with EUROSTAT in capturing the magnitude and evolution of urban expansion at the NUTS2 level. Minor systematic offsets in absolute values do not compromise the dataset’s suitability for analysing urban growth, land take, and associated land cover loss. For regional-scale monitoring of urbanisation processes, the C3S dataset therefore provides results that are comparable to official statistics, while offering the added benefit of spatially explicit, satellite-based observations.\n\n(code-section-6)=", "text_with_prefix": "EQC Quality Assessment: \"Satellite Land Cover completeness for Spatial Planning and Land Management\"\nDataset: satellite-land-cover [CDS]\nAspect: consistency-assessment_q01 | Category: Satellite_ECVs\nSection: Satellite Land Cover completeness for Spatial Planning and Land Management > Analysis and results > 5. Calculate the area percentage and the area percentage change for the most populated regions > Comparison with EUROSTAT Data\n---\n24.5 -5.15 0.11 22.2 -22.09 47.34 15.9 31.44\n```\n\nAcross all regions, both datasets indicate a systematic increase in urban land over time, with comparable magnitudes and trajectories. Absolute differences in settlement/artificial land percentages are generally small, particularly in highly urbanised regions. For Comunidad de Madrid, differences remain below ±1% throughout the period, while Área Metropolitana de Lisboa shows differences mostly within ±2%, despite its higher overall urban fraction.\n\nIn regions with more dispersed or heterogeneous urban structures—such as Cataluña, Comunitat Valenciana, and Norte—C3S settlement values are consistently lower than EUROSTAT artificial land by approximately 1–3%. This offset is stable across years and likely reflects both definitional differences and differences in dataset construction, including spatial resolution. Because C3S is a satellite-derived 300 m gridded product, small or fragmented artificial surfaces may be represented differently from EUROSTAT/LUCAS artificial land, which is based on in-situ point observations aggregated statistically. Crucially, both datasets capture similar rates and directions of urban expansion, indicating that urbanisation dynamics are robustly identified despite these differences.\n\nBetween 2009 and 2022, urban land increased in all regions in both datasets, with comparable absolute changes. For example, in Comunidad de Madrid, C3S settlement increased by +3.38 percentage points (from 9.09% to 12.47%), while EUROSTAT artificial land increased by +1.8 percentage points (from 9.7% to 11.5%). Similarly, in Área Metropolitana de Lisboa, C3S reports an increase of +2.79 percentage points, compared to +3.1 percentage points in EUROSTAT. These comparable magnitudes indicate that both datasets capture a similar volume of land converted into urban use.\n\nThe land cover losses associated with this urban expansion are also consistent in direction. Across all regions, urban growth coincides primarily with reductions in agricultural and open land classes. For instance, in Cataluña, the cumulative increase in settlement between 2009 and 2022 (approximately +0.7 percentage points in C3S) corresponds to a decrease in agricultural land of around –0.5 percentage points, while EUROSTAT reports a comparable decline in cropland over the same period. Similar patterns are observed in Comunitat Valenciana and Norte, where modest but persistent urban gains are accompanied by declining agricultural shares in both datasets.\n\nAlthough C3S and EUROSTAT differ in how agricultural, grassland, and mixed-use areas are classified, these differences do not alter the identification of urban-driven land take. Both datasets indicate that urban expansion primarily occurs at the expense of agricultural and open land, while forest areas remain comparatively stable. This agreement confirms that, in terms of urbanisation impact, the datasets provide equivalent signals of land cover loss.\n\nOverall, the comparison demonstrates that the C3S Land Cover dataset is consistent with EUROSTAT in capturing the magnitude and evolution of urban expansion at the NUTS2 level. Minor systematic offsets in absolute values do not compromise the dataset’s suitability for analysing urban growth, land take, and associated land cover loss. For regional-scale monitoring of urbanisation processes, the C3S dataset therefore provides results that are comparable to official statistics, while offering the added benefit of spatially explicit, satellite-based observations.\n\n(code-section-6)="} {"chunk_id": "satellite_satellite-land-cover_consistency-assessment_q01__86b299302a69", "report_id": "satellite_satellite-land-cover_consistency-assessment_q01", "dataset_id": "satellite-land-cover", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency-assessment_q01", "aspect_base": "consistency-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Land Cover completeness for Spatial Planning and Land Management > Analysis and results > 6. Main Takeaways", "title": "Satellite Land Cover completeness for Spatial Planning and Land Management", "chunk_index": 24, "token_count": 586, "text_raw": "- The dataset demonstrates physical consistency with official sources, aligning well with [EUROSTAT statistics](https://ec.europa.eu/eurostat/databrowser/bookmark/63d3fd90-2f5b-4bec-99f3-4f88814e2624?lang=en) and previously reported European land-cover analyses (e.g. [Feranec et al., 2010](https://doi.org/10.1016/j.apgeog.2009.07.003); [Fernández-Nogueira & Corbelle-Rico, 2018](https://doi.org/10.3390/land7030099)). Between 2009 and 2022, urbanisation occurred across major NUTS2 regions, with settlement areas doubling in some cases. This expansion largely came at the expense of Agriculture and Grassland classes. The observed magnitude and direction of change are consistent with broader European urbanisation trends identified using the GHSL dataset ([Alberti et al., 2019](https://publications.jrc.ec.europa.eu/repository/handle/JRC116711)). Small discrepancies (<5%) with EUROSTAT artificial land statistics do not compromise the dataset’s reliability for identifying overall land-use change trends, though users focusing on very localised patterns should interpret results with caution.\n\n- Urban expansion mainly replaced agricultural and open land, while Forest areas remained largely stable over the study period. Settlements increased their share of total land area across regions and remain a minority land cover class. These patterns are consistent with spatial trade-off dynamics reported in global urban land-cover change analyses, where urban growth frequently draws from agricultural land rather than direct forest conversion ([Bagan & Yamagata, 2014](https://doi.org/10.1088/1748-9326/9/6/064015)).\n\n- The use of both detailed and aggregated IPCC classes provided a comprehensive view of land cover dynamics, balancing precision and generalisation. Although class aggregation introduces some uncertainty (e.g. between Agriculture and Grassland), it enables clearer identification of dominant transitions such as urban expansion, consistent with structured land-cover change assessment approaches reported in the literature.\n\n- Minor classification differences between datasets (e.g. coding of mixed-use or greenhouse areas) can introduce uncertainty. Nonetheless, the dataset maintains strong performance in identifying urban expansion trends, in line with consistency and accuracy assessments of global land-cover products ([Zhao et al., 2023](https://doi.org/10.3390/rs15092285)).", "text_with_prefix": "EQC Quality Assessment: \"Satellite Land Cover completeness for Spatial Planning and Land Management\"\nDataset: satellite-land-cover [CDS]\nAspect: consistency-assessment_q01 | Category: Satellite_ECVs\nSection: Satellite Land Cover completeness for Spatial Planning and Land Management > Analysis and results > 6. Main Takeaways\n---\n- The dataset demonstrates physical consistency with official sources, aligning well with [EUROSTAT statistics](https://ec.europa.eu/eurostat/databrowser/bookmark/63d3fd90-2f5b-4bec-99f3-4f88814e2624?lang=en) and previously reported European land-cover analyses (e.g. [Feranec et al., 2010](https://doi.org/10.1016/j.apgeog.2009.07.003); [Fernández-Nogueira & Corbelle-Rico, 2018](https://doi.org/10.3390/land7030099)). Between 2009 and 2022, urbanisation occurred across major NUTS2 regions, with settlement areas doubling in some cases. This expansion largely came at the expense of Agriculture and Grassland classes. The observed magnitude and direction of change are consistent with broader European urbanisation trends identified using the GHSL dataset ([Alberti et al., 2019](https://publications.jrc.ec.europa.eu/repository/handle/JRC116711)). Small discrepancies (<5%) with EUROSTAT artificial land statistics do not compromise the dataset’s reliability for identifying overall land-use change trends, though users focusing on very localised patterns should interpret results with caution.\n\n- Urban expansion mainly replaced agricultural and open land, while Forest areas remained largely stable over the study period. Settlements increased their share of total land area across regions and remain a minority land cover class. These patterns are consistent with spatial trade-off dynamics reported in global urban land-cover change analyses, where urban growth frequently draws from agricultural land rather than direct forest conversion ([Bagan & Yamagata, 2014](https://doi.org/10.1088/1748-9326/9/6/064015)).\n\n- The use of both detailed and aggregated IPCC classes provided a comprehensive view of land cover dynamics, balancing precision and generalisation. Although class aggregation introduces some uncertainty (e.g. between Agriculture and Grassland), it enables clearer identification of dominant transitions such as urban expansion, consistent with structured land-cover change assessment approaches reported in the literature.\n\n- Minor classification differences between datasets (e.g. coding of mixed-use or greenhouse areas) can introduce uncertainty. Nonetheless, the dataset maintains strong performance in identifying urban expansion trends, in line with consistency and accuracy assessments of global land-cover products ([Zhao et al., 2023](https://doi.org/10.3390/rs15092285))."} {"chunk_id": "satellite_satellite-land-cover_consistency-assessment_q01__cfe770c545ef", "report_id": "satellite_satellite-land-cover_consistency-assessment_q01", "dataset_id": "satellite-land-cover", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency-assessment_q01", "aspect_base": "consistency-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Land Cover completeness for Spatial Planning and Land Management > ℹ️ If you want to know more > Key resources", "title": "Satellite Land Cover completeness for Spatial Planning and Land Management", "chunk_index": 25, "token_count": 367, "text_raw": "Some key resources and further reading were linked throughout this assessment.\n\nThe CDS catalogue entry for the data used were:\n* Land cover classification gridded maps from 1992 to present derived from satellite observations:\n (https://cds.climate.copernicus.eu/datasets/satellite-land-cover?tab=overview)\n\n* Product User Guide and Specification of the dataset [version 2.1](https://dast.copernicus-climate.eu/documents/satellite-land-cover/D5.3.1_PUGS_ICDR_LC_v2.1.x_PRODUCTS_v1.1.pdf) and [version 2.0](https://dast.copernicus-climate.eu/documents/satellite-land-cover/D3.3.11-v1.0_PUGS_CDR_LC-CCI_v2.0.7cds_Products_v1.0.1_APPROVED_Ver1.pdf)\n\nAdditional resources:\n* [Eurostat NUTS](https://ec.europa.eu/eurostat/web/nuts) (Nomenclature of territorial units for statistics)\n\nCode libraries used:\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)", "text_with_prefix": "EQC Quality Assessment: \"Satellite Land Cover completeness for Spatial Planning and Land Management\"\nDataset: satellite-land-cover [CDS]\nAspect: consistency-assessment_q01 | Category: Satellite_ECVs\nSection: Satellite Land Cover completeness for Spatial Planning and Land Management > ℹ️ If you want to know more > Key resources\n---\nSome key resources and further reading were linked throughout this assessment.\n\nThe CDS catalogue entry for the data used were:\n* Land cover classification gridded maps from 1992 to present derived from satellite observations:\n (https://cds.climate.copernicus.eu/datasets/satellite-land-cover?tab=overview)\n\n* Product User Guide and Specification of the dataset [version 2.1](https://dast.copernicus-climate.eu/documents/satellite-land-cover/D5.3.1_PUGS_ICDR_LC_v2.1.x_PRODUCTS_v1.1.pdf) and [version 2.0](https://dast.copernicus-climate.eu/documents/satellite-land-cover/D3.3.11-v1.0_PUGS_CDR_LC-CCI_v2.0.7cds_Products_v1.0.1_APPROVED_Ver1.pdf)\n\nAdditional resources:\n* [Eurostat NUTS](https://ec.europa.eu/eurostat/web/nuts) (Nomenclature of territorial units for statistics)\n\nCode libraries used:\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)"} {"chunk_id": "satellite_satellite-land-cover_consistency-assessment_q01__2ede1b89a87e", "report_id": "satellite_satellite-land-cover_consistency-assessment_q01", "dataset_id": "satellite-land-cover", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency-assessment_q01", "aspect_base": "consistency-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Land Cover completeness for Spatial Planning and Land Management > ℹ️ If you want to know more > References", "title": "Satellite Land Cover completeness for Spatial Planning and Land Management", "chunk_index": 26, "token_count": 951, "text_raw": "[Feranec, J., Jaffrain, G., Soukup, T., & Hazeu, G. (2010). Determining changes and flows in European landscapes 1990–2000 using CORINE land cover data. Applied geography, 30(1), 19-35.](https://doi.org/10.1016/j.apgeog.2009.07.003)\n\n[Fernández-Nogueira, D., & Corbelle-Rico, E. (2018). Land use changes in Iberian Peninsula 1990–2012. Land, 7(3), 99.](https://doi.org/10.3390/land7030099)\n\n[Eurostat, Statistical Atlas](https://ec.europa.eu/statistical-atlas/viewer/?ch=gridvizChapter&mids=BKGCNT,totalPop21,CNTOVL&o=1,1,0.7¢er=40.94812,-2.2287,5&lcis=totalPop21&)\n\n[Bagan, H., & Yamagata, Y. (2014). Land-cover change analysis in 50 global cities by using a combination of Landsat data and analysis of grid cells. Environmental Research Letters, 9.](https://doi.org/10.1088/1748-9326/9/6/064015)\n\n[Alberti, V., Alonso Raposo, M., Attardo, C., Auteri, D., Ribeiro Barranco, R., Batista E Silva, F., Benczur, P., Bertoldi, P., Bono, F., Bussolari, I., Louro Caldeira, S., Carlsson, J., Christidis, P., Christodoulou, A., Ciuffo, B., Corrado, S., Fioretti, C., Galassi, M., Galbusera, L., Gawlik, B., Giusti, F., Gomez Prieto, J., Grosso, M., Martinho Guimaraes Pires Pereira, A., Jacobs, C., Kavalov, B., Kompil, M., Kucas, A., Kona, A., Lavalle, C., Leip, A., Lyons, L., Manca, A., Melchiorri, M., Monforti-Ferrario, F., Montalto, V., Mortara, B., Natale, F., Panella, F., Pasi, G., Perpiña Castillo, C., Pertoldi, M., Pisoni, E., Roque Mendes Polvora, A., Rainoldi, A., Rembges, D., Rissola, G., Sala, S., Schade, S., Serra, N., Spirito, L., Tsakalidis, A., Schiavina, M., Tintori, G., Vaccari, L., Vandyck, T., Vanham, D., Van Heerden, S., Van Noordt, C., Vespe, M., Vetters, N., Vilahur Chiaraviglio, N., Vizcaino, M., Von Estorff, U. and Zulian, G., The Future of Cities, Vandecasteele, I., Baranzelli, C., Siragusa, A. and Aurambout, J. editor(s), EUR 29752 EN, Publications Office of the European Union, Luxembourg, 2019, ISBN 978-92-76-03847-4, doi:10.2760/375209, JRC116711.](https://publications.jrc.ec.europa.eu/repository/handle/JRC116711)\n\n[Zhao, T., Zhang, X., Gao, Y., Mi, J., Liu, W., Wang, J., Jiang, M., & Liu, L. (2023). Assessing the Accuracy and Consistency of Six Fine-Resolution Global Land Cover Products Using a Novel Stratified Random Sampling Validation Dataset. Remote. Sens., 15, 2285.](https://doi.org/10.3390/rs15092285)", "text_with_prefix": "EQC Quality Assessment: \"Satellite Land Cover completeness for Spatial Planning and Land Management\"\nDataset: satellite-land-cover [CDS]\nAspect: consistency-assessment_q01 | Category: Satellite_ECVs\nSection: Satellite Land Cover completeness for Spatial Planning and Land Management > ℹ️ If you want to know more > References\n---\n[Feranec, J., Jaffrain, G., Soukup, T., & Hazeu, G. (2010). Determining changes and flows in European landscapes 1990–2000 using CORINE land cover data. Applied geography, 30(1), 19-35.](https://doi.org/10.1016/j.apgeog.2009.07.003)\n\n[Fernández-Nogueira, D., & Corbelle-Rico, E. (2018). Land use changes in Iberian Peninsula 1990–2012. Land, 7(3), 99.](https://doi.org/10.3390/land7030099)\n\n[Eurostat, Statistical Atlas](https://ec.europa.eu/statistical-atlas/viewer/?ch=gridvizChapter&mids=BKGCNT,totalPop21,CNTOVL&o=1,1,0.7¢er=40.94812,-2.2287,5&lcis=totalPop21&)\n\n[Bagan, H., & Yamagata, Y. (2014). Land-cover change analysis in 50 global cities by using a combination of Landsat data and analysis of grid cells. Environmental Research Letters, 9.](https://doi.org/10.1088/1748-9326/9/6/064015)\n\n[Alberti, V., Alonso Raposo, M., Attardo, C., Auteri, D., Ribeiro Barranco, R., Batista E Silva, F., Benczur, P., Bertoldi, P., Bono, F., Bussolari, I., Louro Caldeira, S., Carlsson, J., Christidis, P., Christodoulou, A., Ciuffo, B., Corrado, S., Fioretti, C., Galassi, M., Galbusera, L., Gawlik, B., Giusti, F., Gomez Prieto, J., Grosso, M., Martinho Guimaraes Pires Pereira, A., Jacobs, C., Kavalov, B., Kompil, M., Kucas, A., Kona, A., Lavalle, C., Leip, A., Lyons, L., Manca, A., Melchiorri, M., Monforti-Ferrario, F., Montalto, V., Mortara, B., Natale, F., Panella, F., Pasi, G., Perpiña Castillo, C., Pertoldi, M., Pisoni, E., Roque Mendes Polvora, A., Rainoldi, A., Rembges, D., Rissola, G., Sala, S., Schade, S., Serra, N., Spirito, L., Tsakalidis, A., Schiavina, M., Tintori, G., Vaccari, L., Vandyck, T., Vanham, D., Van Heerden, S., Van Noordt, C., Vespe, M., Vetters, N., Vilahur Chiaraviglio, N., Vizcaino, M., Von Estorff, U. and Zulian, G., The Future of Cities, Vandecasteele, I., Baranzelli, C., Siragusa, A. and Aurambout, J. editor(s), EUR 29752 EN, Publications Office of the European Union, Luxembourg, 2019, ISBN 978-92-76-03847-4, doi:10.2760/375209, JRC116711.](https://publications.jrc.ec.europa.eu/repository/handle/JRC116711)\n\n[Zhao, T., Zhang, X., Gao, Y., Mi, J., Liu, W., Wang, J., Jiang, M., & Liu, L. (2023). Assessing the Accuracy and Consistency of Six Fine-Resolution Global Land Cover Products Using a Novel Stratified Random Sampling Validation Dataset. Remote. Sens., 15, 2285.](https://doi.org/10.3390/rs15092285)"} {"chunk_id": "satellite_satellite-land-cover_trend-assessment_q02__f2d6d51bfc35", "report_id": "satellite_satellite-land-cover_trend-assessment_q02", "dataset_id": "satellite-land-cover", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q02", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Land Cover trend assessment for Spatial Planning and Land Management > Quality assessment question", "title": "Satellite Land Cover trend assessment for Spatial Planning and Land Management", "chunk_index": 0, "token_count": 276, "text_raw": "* **Is the dataset accurate and consistent for the analysis of urbanisation trends in the Iberian Peninsula?**\n\nLand Cover data is an invaluable resource for a wide range of fields, from climate change research to urban planning. Land Cover products that provide historical timelines enable scientists, policymakers, and planners to understand and analyse the transformation of land cover over recent decades ([Vargo et al, 2013](https://doi.org/10.1016/j.jenvman.2012.10.007); [Chang et al., 2018](https://doi.org/10.1088/1755-1315/113/1/012087)).\n\nThis notebook will access the ***Land cover classification gridded maps from 1992 to present derived from satellite observations*** (henceforth, LC) data from the Climate Data Store (CDS) of the Copernicus Climate Change Service (C3S), and analyse the spatial patterns of the LC over a given Area of Interest (AoI) and time.", "text_with_prefix": "EQC Quality Assessment: \"Satellite Land Cover trend assessment for Spatial Planning and Land Management\"\nDataset: satellite-land-cover [CDS]\nAspect: trend-assessment_q02 | Category: Satellite_ECVs\nSection: Satellite Land Cover trend assessment for Spatial Planning and Land Management > Quality assessment question\n---\n* **Is the dataset accurate and consistent for the analysis of urbanisation trends in the Iberian Peninsula?**\n\nLand Cover data is an invaluable resource for a wide range of fields, from climate change research to urban planning. Land Cover products that provide historical timelines enable scientists, policymakers, and planners to understand and analyse the transformation of land cover over recent decades ([Vargo et al, 2013](https://doi.org/10.1016/j.jenvman.2012.10.007); [Chang et al., 2018](https://doi.org/10.1088/1755-1315/113/1/012087)).\n\nThis notebook will access the ***Land cover classification gridded maps from 1992 to present derived from satellite observations*** (henceforth, LC) data from the Climate Data Store (CDS) of the Copernicus Climate Change Service (C3S), and analyse the spatial patterns of the LC over a given Area of Interest (AoI) and time."} {"chunk_id": "satellite_satellite-land-cover_trend-assessment_q02__e54104b0962e", "report_id": "satellite_satellite-land-cover_trend-assessment_q02", "dataset_id": "satellite-land-cover", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q02", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Land Cover trend assessment for Spatial Planning and Land Management > Quality assessment statement", "title": "Satellite Land Cover trend assessment for Spatial Planning and Land Management", "chunk_index": 1, "token_count": 244, "text_raw": "These are the key outcomes of this assessment\n\nThe dataset maintains strong temporal continuity, with annual updates ensuring a smooth and reliable representation of land cover changes over time. While breakpoints were identified, they generally did not indicate major disruptions, reinforcing the dataset’s stability for long-term trend analysis.\n\nThe presence of breakpoints does not necessarily indicate abrupt landscape shifts but rather highlights the sensitivity of detection methods to gradual changes. This suggests that while breakpoints can help refine analysis, their impact on overall trends remains limited, emphasising the dataset’s resilience to minor variations.\n\nFor the specific land cover type analysed, the dataset exhibits a consistent ability to capture underlying trends. The similarity in results across segmented and total trends suggests that the data structure is well-calibrated, minimising distortions that could arise from classification inconsistencies or methodological biases.\n\n```\n\n![Urbanization_Map_Series.png](attachment:Urbanization_Map_Series.png)", "text_with_prefix": "EQC Quality Assessment: \"Satellite Land Cover trend assessment for Spatial Planning and Land Management\"\nDataset: satellite-land-cover [CDS]\nAspect: trend-assessment_q02 | Category: Satellite_ECVs\nSection: Satellite Land Cover trend assessment for Spatial Planning and Land Management > Quality assessment statement\n---\nThese are the key outcomes of this assessment\n\nThe dataset maintains strong temporal continuity, with annual updates ensuring a smooth and reliable representation of land cover changes over time. While breakpoints were identified, they generally did not indicate major disruptions, reinforcing the dataset’s stability for long-term trend analysis.\n\nThe presence of breakpoints does not necessarily indicate abrupt landscape shifts but rather highlights the sensitivity of detection methods to gradual changes. This suggests that while breakpoints can help refine analysis, their impact on overall trends remains limited, emphasising the dataset’s resilience to minor variations.\n\nFor the specific land cover type analysed, the dataset exhibits a consistent ability to capture underlying trends. The similarity in results across segmented and total trends suggests that the data structure is well-calibrated, minimising distortions that could arise from classification inconsistencies or methodological biases.\n\n```\n\n![Urbanization_Map_Series.png](attachment:Urbanization_Map_Series.png)"} {"chunk_id": "satellite_satellite-land-cover_trend-assessment_q02__54567dd939fb", "report_id": "satellite_satellite-land-cover_trend-assessment_q02", "dataset_id": "satellite-land-cover", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q02", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Land Cover trend assessment for Spatial Planning and Land Management > Methodology", "title": "Satellite Land Cover trend assessment for Spatial Planning and Land Management", "chunk_index": 2, "token_count": 113, "text_raw": "**This Use Case comprises the following steps:**\n\n**[](code-section-1)**\n\n**[](code-section-2)**\n\n**[](code-section-3)**\n\n**[](code-section-4)**\n\n**[](code-section-5)**", "text_with_prefix": "EQC Quality Assessment: \"Satellite Land Cover trend assessment for Spatial Planning and Land Management\"\nDataset: satellite-land-cover [CDS]\nAspect: trend-assessment_q02 | Category: Satellite_ECVs\nSection: Satellite Land Cover trend assessment for Spatial Planning and Land Management > Methodology\n---\n**This Use Case comprises the following steps:**\n\n**[](code-section-1)**\n\n**[](code-section-2)**\n\n**[](code-section-3)**\n\n**[](code-section-4)**\n\n**[](code-section-5)**"} {"chunk_id": "satellite_satellite-land-cover_trend-assessment_q02__48c1d910b09f", "report_id": "satellite_satellite-land-cover_trend-assessment_q02", "dataset_id": "satellite-land-cover", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q02", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 1. Define the AoI, search and download LC data", "title": "Satellite Land Cover trend assessment for Spatial Planning and Land Management", "chunk_index": 3, "token_count": 114, "text_raw": "Before we begin we must prepare our environment. This includes installing the Application Programming Interface (API) of the CDS, and importing the various python libraries that we will need.", "text_with_prefix": "EQC Quality Assessment: \"Satellite Land Cover trend assessment for Spatial Planning and Land Management\"\nDataset: satellite-land-cover [CDS]\nAspect: trend-assessment_q02 | Category: Satellite_ECVs\nSection: Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 1. Define the AoI, search and download LC data\n---\nBefore we begin we must prepare our environment. This includes installing the Application Programming Interface (API) of the CDS, and importing the various python libraries that we will need."} {"chunk_id": "satellite_satellite-land-cover_trend-assessment_q02__11d7d0f08de3", "report_id": "satellite_satellite-land-cover_trend-assessment_q02", "dataset_id": "satellite-land-cover", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q02", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 1. Define the AoI, search and download LC data > Install CDS API", "title": "Satellite Land Cover trend assessment for Spatial Planning and Land Management", "chunk_index": 4, "token_count": 671, "text_raw": "To install the CDS API, run the following command. We use an exclamation mark to pass the command to the shell (not to the Python interpreter).\nIf you already have the CDS API installed, you can skip or comment this step.\n\n##### Import all the libraries/packages\n\nWe will be working with data in NetCDF format. To best handle this type of data we will use libraries for working with multidimensional arrays, in particular Xarray. \nWe will also need libraries for plotting and viewing data.\n\nImport Standard Libraries\nImport Numerical & Statistical Libraries\nimport ruptures as rpt\nImport Geospatial Libraries\nImport Visualization Libraries\nImport External Tools\nSet Matplotlib Style\nSet the CDSAPI location\n\n##### Data Overview\n\nTo search for data, visit the CDS website: http://cds.climate.copernicus.eu Here you can search for 'Satellite observations' using the search bar. The data we need for this tutorial is the ***Land cover classification gridded maps from 1992 to present derived from satellite observations***. This catalogue entry provides global Land Cover Classification (LCC) maps with a very high spatial resolution, with a L4 processing level, on an annual basis with a one-year delay, following the [Global Climate Observing System (GCOS) convention requirements](https://library.wmo.int/records/item/58111-the-2022-gcos-ecvs-requirements-gcos-245). LULC maps correspond to a global classification scheme, encompassing 22 classes.\n\nThe dataset consists of 2 versions (v2.0.7 produced by the European Space Agency (ESA) Climate Change Initiative (CCI) and v2.1.1 produced by Copernicus Climate Change Service (C3S)).\n\nData specifications for this use case:\n* **Years:** 1992 to 2022\n* **Version:** v2.0.7 before 1992 and v2.1.1 after 2016\n* **Format:** Zip files\n\nAt the end of the download form, select “**Show API request**”. This will reveal a block of code, which you can simply copy and paste into a cell of your Jupyter Notebook. Having copied the API request, running it will retrieve and download the data you requested into your local directory. However, before you run it, the **terms and conditions** of this particular dataset need to have been accepted directly at the CDS website. The option to view and accept these conditions is given at the end of the download form, just above the “**Show API request**” option. In addition, it is also useful to define the time period and AoI parameters and edit the request accordingly, as exemplified in the cells below.\n\nYears to download\nList of requests to retrieve data\n\nDownload and regionalize by AoI\n\nInspect the database", "text_with_prefix": "EQC Quality Assessment: \"Satellite Land Cover trend assessment for Spatial Planning and Land Management\"\nDataset: satellite-land-cover [CDS]\nAspect: trend-assessment_q02 | Category: Satellite_ECVs\nSection: Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 1. Define the AoI, search and download LC data > Install CDS API\n---\nTo install the CDS API, run the following command. We use an exclamation mark to pass the command to the shell (not to the Python interpreter).\nIf you already have the CDS API installed, you can skip or comment this step.\n\n##### Import all the libraries/packages\n\nWe will be working with data in NetCDF format. To best handle this type of data we will use libraries for working with multidimensional arrays, in particular Xarray. \nWe will also need libraries for plotting and viewing data.\n\nImport Standard Libraries\nImport Numerical & Statistical Libraries\nimport ruptures as rpt\nImport Geospatial Libraries\nImport Visualization Libraries\nImport External Tools\nSet Matplotlib Style\nSet the CDSAPI location\n\n##### Data Overview\n\nTo search for data, visit the CDS website: http://cds.climate.copernicus.eu Here you can search for 'Satellite observations' using the search bar. The data we need for this tutorial is the ***Land cover classification gridded maps from 1992 to present derived from satellite observations***. This catalogue entry provides global Land Cover Classification (LCC) maps with a very high spatial resolution, with a L4 processing level, on an annual basis with a one-year delay, following the [Global Climate Observing System (GCOS) convention requirements](https://library.wmo.int/records/item/58111-the-2022-gcos-ecvs-requirements-gcos-245). LULC maps correspond to a global classification scheme, encompassing 22 classes.\n\nThe dataset consists of 2 versions (v2.0.7 produced by the European Space Agency (ESA) Climate Change Initiative (CCI) and v2.1.1 produced by Copernicus Climate Change Service (C3S)).\n\nData specifications for this use case:\n* **Years:** 1992 to 2022\n* **Version:** v2.0.7 before 1992 and v2.1.1 after 2016\n* **Format:** Zip files\n\nAt the end of the download form, select “**Show API request**”. This will reveal a block of code, which you can simply copy and paste into a cell of your Jupyter Notebook. Having copied the API request, running it will retrieve and download the data you requested into your local directory. However, before you run it, the **terms and conditions** of this particular dataset need to have been accepted directly at the CDS website. The option to view and accept these conditions is given at the end of the download form, just above the “**Show API request**” option. In addition, it is also useful to define the time period and AoI parameters and edit the request accordingly, as exemplified in the cells below.\n\nYears to download\nList of requests to retrieve data\n\nDownload and regionalize by AoI\n\nInspect the database"} {"chunk_id": "satellite_satellite-land-cover_trend-assessment_q02__9e6c7ab85526", "report_id": "satellite_satellite-land-cover_trend-assessment_q02", "dataset_id": "satellite-land-cover", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q02", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 1. Define the AoI, search and download LC data > Install CDS API", "title": "Satellite Land Cover trend assessment for Spatial Planning and Land Management", "chunk_index": 5, "token_count": 792, "text_raw": "exemplified in the cells below.\n\nYears to download\nList of requests to retrieve data\n\nDownload and regionalize by AoI\n\nInspect the database\n\n```text\n Size: 7GB\nDimensions: (year: 31, latitude: 3600, longitude: 5040, bounds: 2)\nCoordinates:\n * year (year) int64 248B 1992 1993 1994 ... 2020 2021 2022\n * latitude (latitude) float64 29kB 45.0 45.0 44.99 ... 35.0 35.0\n * longitude (longitude) float64 40kB -9.999 -9.996 ... 3.996 3.999\n lat_bounds (latitude, bounds, longitude) float64 290MB dask.array\n lon_bounds (longitude, bounds) float64 81kB dask.array\n time_bounds (year, bounds, longitude) datetime64[ns] 2MB dask.array\nDimensions without coordinates: bounds\nData variables:\n lccs_class (year, latitude, longitude) uint8 562MB dask.array\n processed_flag (year, latitude, longitude) float32 2GB dask.array\n current_pixel_state (year, latitude, longitude) float32 2GB dask.array\n observation_count (year, latitude, longitude) uint16 1GB dask.array\n change_count (year, latitude, longitude) uint8 562MB dask.array\n crs (year, longitude) int32 625kB dask.array\nAttributes: (12/38)\n id: ESACCI-LC-L4-LCCS-Map-300m-P1Y-1992-v2.0.7cds\n title: Land Cover Map of ESA CCI brokered by CDS\n summary: This dataset characterizes the land cover of ...\n type: ESACCI-LC-L4-LCCS-Map-300m-P1Y\n project: Climate Change Initiative - European Space Ag...\n references: http://www.esa-landcover-cci.org/\n ... ...\n geospatial_lon_max: 180\n spatial_resolution: 300m\n geospatial_lat_units: degrees_north\n geospatial_lat_resolution: 0.002778\n geospatial_lon_units: degrees_east\n geospatial_lon_resolution: 0.002778\n```\n\n(code-section-2)=", "text_with_prefix": "EQC Quality Assessment: \"Satellite Land Cover trend assessment for Spatial Planning and Land Management\"\nDataset: satellite-land-cover [CDS]\nAspect: trend-assessment_q02 | Category: Satellite_ECVs\nSection: Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 1. Define the AoI, search and download LC data > Install CDS API\n---\nexemplified in the cells below.\n\nYears to download\nList of requests to retrieve data\n\nDownload and regionalize by AoI\n\nInspect the database\n\n```text\n Size: 7GB\nDimensions: (year: 31, latitude: 3600, longitude: 5040, bounds: 2)\nCoordinates:\n * year (year) int64 248B 1992 1993 1994 ... 2020 2021 2022\n * latitude (latitude) float64 29kB 45.0 45.0 44.99 ... 35.0 35.0\n * longitude (longitude) float64 40kB -9.999 -9.996 ... 3.996 3.999\n lat_bounds (latitude, bounds, longitude) float64 290MB dask.array\n lon_bounds (longitude, bounds) float64 81kB dask.array\n time_bounds (year, bounds, longitude) datetime64[ns] 2MB dask.array\nDimensions without coordinates: bounds\nData variables:\n lccs_class (year, latitude, longitude) uint8 562MB dask.array\n processed_flag (year, latitude, longitude) float32 2GB dask.array\n current_pixel_state (year, latitude, longitude) float32 2GB dask.array\n observation_count (year, latitude, longitude) uint16 1GB dask.array\n change_count (year, latitude, longitude) uint8 562MB dask.array\n crs (year, longitude) int32 625kB dask.array\nAttributes: (12/38)\n id: ESACCI-LC-L4-LCCS-Map-300m-P1Y-1992-v2.0.7cds\n title: Land Cover Map of ESA CCI brokered by CDS\n summary: This dataset characterizes the land cover of ...\n type: ESACCI-LC-L4-LCCS-Map-300m-P1Y\n project: Climate Change Initiative - European Space Ag...\n references: http://www.esa-landcover-cci.org/\n ... ...\n geospatial_lon_max: 180\n spatial_resolution: 300m\n geospatial_lat_units: degrees_north\n geospatial_lat_resolution: 0.002778\n geospatial_lon_units: degrees_east\n geospatial_lon_resolution: 0.002778\n```\n\n(code-section-2)="} {"chunk_id": "satellite_satellite-land-cover_trend-assessment_q02__68a01471aaee", "report_id": "satellite_satellite-land-cover_trend-assessment_q02", "dataset_id": "satellite-land-cover", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q02", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 2. Inspect and view data for the defined AoI (Iberian Peninsula) > Compute urban area for each NUTS 2 region", "title": "Satellite Land Cover trend assessment for Spatial Planning and Land Management", "chunk_index": 6, "token_count": 302, "text_raw": "To identify changes in LC patterns, in this user question, NUTS 2 will be used, providing the information regarding the main regions of the Iberian Peninsula.\n\nThe [NUTS](https://ec.europa.eu/eurostat/web/nuts) are a hierarchical system divided into 3 levels. NUTS 1 correspond to major socio-economic regions, NUTS 2 correspond to basic regions for the application of regional policies, and NUTS 3 correspond to small regions for specific diagnoses. Additionally a NUTS 0 level, usually co-incident with national boundaries is also available. The NUTS legislation is periodically amended; therefore multiple years are available for download.\n\nThe step below masks the Land Cover data according to the NUTS 2 boundaries and calculate the area of each pixel (weighted by the cosine of Latitude). For each NUTS 2, we proceed with the analysis and visual inspection of Land Cover areas per class and corresponding percentages during the elected period.", "text_with_prefix": "EQC Quality Assessment: \"Satellite Land Cover trend assessment for Spatial Planning and Land Management\"\nDataset: satellite-land-cover [CDS]\nAspect: trend-assessment_q02 | Category: Satellite_ECVs\nSection: Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 2. Inspect and view data for the defined AoI (Iberian Peninsula) > Compute urban area for each NUTS 2 region\n---\nTo identify changes in LC patterns, in this user question, NUTS 2 will be used, providing the information regarding the main regions of the Iberian Peninsula.\n\nThe [NUTS](https://ec.europa.eu/eurostat/web/nuts) are a hierarchical system divided into 3 levels. NUTS 1 correspond to major socio-economic regions, NUTS 2 correspond to basic regions for the application of regional policies, and NUTS 3 correspond to small regions for specific diagnoses. Additionally a NUTS 0 level, usually co-incident with national boundaries is also available. The NUTS legislation is periodically amended; therefore multiple years are available for download.\n\nThe step below masks the Land Cover data according to the NUTS 2 boundaries and calculate the area of each pixel (weighted by the cosine of Latitude). For each NUTS 2, we proceed with the analysis and visual inspection of Land Cover areas per class and corresponding percentages during the elected period."} {"chunk_id": "satellite_satellite-land-cover_trend-assessment_q02__573c7b27259d", "report_id": "satellite_satellite-land-cover_trend-assessment_q02", "dataset_id": "satellite-land-cover", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q02", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 2. Inspect and view data for the defined AoI (Iberian Peninsula) > Mask regions", "title": "Satellite Land Cover trend assessment for Spatial Planning and Land Management", "chunk_index": 7, "token_count": 174, "text_raw": "First, we need to establish the geometry of the NUTS region (level 2) in order to make the corresponding statistics.\n\nConfigure Dask\nDefine CRS and bounding box for Iberian Peninsula\nLoad and filter GeoDataFrame\nEnsure dataset CRS is set\nEnsure dataset coordinates overlap with the filtered regions\nSubset dataset to valid ranges\nCheck subsetted dataset dimensions\nCreate the regionmask\nCreate a 2D mask", "text_with_prefix": "EQC Quality Assessment: \"Satellite Land Cover trend assessment for Spatial Planning and Land Management\"\nDataset: satellite-land-cover [CDS]\nAspect: trend-assessment_q02 | Category: Satellite_ECVs\nSection: Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 2. Inspect and view data for the defined AoI (Iberian Peninsula) > Mask regions\n---\nFirst, we need to establish the geometry of the NUTS region (level 2) in order to make the corresponding statistics.\n\nConfigure Dask\nDefine CRS and bounding box for Iberian Peninsula\nLoad and filter GeoDataFrame\nEnsure dataset CRS is set\nEnsure dataset coordinates overlap with the filtered regions\nSubset dataset to valid ranges\nCheck subsetted dataset dimensions\nCreate the regionmask\nCreate a 2D mask"} {"chunk_id": "satellite_satellite-land-cover_trend-assessment_q02__6b53a45f6245", "report_id": "satellite_satellite-land-cover_trend-assessment_q02", "dataset_id": "satellite-land-cover", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q02", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 2. Inspect and view data for the defined AoI (Iberian Peninsula) > Compute cell area", "title": "Satellite Land Cover trend assessment for Spatial Planning and Land Management", "chunk_index": 8, "token_count": 178, "text_raw": "Then, we can calculate the area of each pixel taking into consideration the curvature of the earth (i.e., weighted by the cosine of Latitude).\n\nScaling factor for conversion (constant longitude resolution)\nUse the latitude values directly from the dataset\nCalculate the difference between consecutive latitude values\nConvert latitude differences to kilometers\nCompute the grid cell area for each latitude\nAssign attributes to the grid cell area\nAdd the grid cell area as a coordinate to the dataset", "text_with_prefix": "EQC Quality Assessment: \"Satellite Land Cover trend assessment for Spatial Planning and Land Management\"\nDataset: satellite-land-cover [CDS]\nAspect: trend-assessment_q02 | Category: Satellite_ECVs\nSection: Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 2. Inspect and view data for the defined AoI (Iberian Peninsula) > Compute cell area\n---\nThen, we can calculate the area of each pixel taking into consideration the curvature of the earth (i.e., weighted by the cosine of Latitude).\n\nScaling factor for conversion (constant longitude resolution)\nUse the latitude values directly from the dataset\nCalculate the difference between consecutive latitude values\nConvert latitude differences to kilometers\nCompute the grid cell area for each latitude\nAssign attributes to the grid cell area\nAdd the grid cell area as a coordinate to the dataset"} {"chunk_id": "satellite_satellite-land-cover_trend-assessment_q02__ed9dfa0d5f35", "report_id": "satellite_satellite-land-cover_trend-assessment_q02", "dataset_id": "satellite-land-cover", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q02", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 2. Inspect and view data for the defined AoI (Iberian Peninsula) > Select Urban Classes and Prepare Dataset", "title": "Satellite Land Cover trend assessment for Spatial Planning and Land Management", "chunk_index": 9, "token_count": 146, "text_raw": "Define urban classes\nCreate a mask for urban areas\nMasked urban area using cell area\nStack latitude and longitude into a single dimension\nCreate a stacked mask and align dimensions\nAttach the stacked mask to the dataset\nGroup by regions and compute urban area for each year", "text_with_prefix": "EQC Quality Assessment: \"Satellite Land Cover trend assessment for Spatial Planning and Land Management\"\nDataset: satellite-land-cover [CDS]\nAspect: trend-assessment_q02 | Category: Satellite_ECVs\nSection: Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 2. Inspect and view data for the defined AoI (Iberian Peninsula) > Select Urban Classes and Prepare Dataset\n---\nDefine urban classes\nCreate a mask for urban areas\nMasked urban area using cell area\nStack latitude and longitude into a single dimension\nCreate a stacked mask and align dimensions\nAttach the stacked mask to the dataset\nGroup by regions and compute urban area for each year"} {"chunk_id": "satellite_satellite-land-cover_trend-assessment_q02__9d879bb943ea", "report_id": "satellite_satellite-land-cover_trend-assessment_q02", "dataset_id": "satellite-land-cover", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q02", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 2. Inspect and view data for the defined AoI (Iberian Peninsula) > Compute urban area per region and year", "title": "Satellite Land Cover trend assessment for Spatial Planning and Land Management", "chunk_index": 10, "token_count": 131, "text_raw": "Ensure valid region indexing and naming\nSave growth data by year\nAdd results to the list\nConvert results to a DataFrame\nAdd geometry to results_df\nEnsure results_df is a GeoDataFrame", "text_with_prefix": "EQC Quality Assessment: \"Satellite Land Cover trend assessment for Spatial Planning and Land Management\"\nDataset: satellite-land-cover [CDS]\nAspect: trend-assessment_q02 | Category: Satellite_ECVs\nSection: Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 2. Inspect and view data for the defined AoI (Iberian Peninsula) > Compute urban area per region and year\n---\nEnsure valid region indexing and naming\nSave growth data by year\nAdd results to the list\nConvert results to a DataFrame\nAdd geometry to results_df\nEnsure results_df is a GeoDataFrame"} {"chunk_id": "satellite_satellite-land-cover_trend-assessment_q02__053a24469f30", "report_id": "satellite_satellite-land-cover_trend-assessment_q02", "dataset_id": "satellite-land-cover", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q02", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 2. Inspect and view data for the defined AoI (Iberian Peninsula) > Map urban percentage coverage over-time by NUTS regions in the AoI", "title": "Satellite Land Cover trend assessment for Spatial Planning and Land Management", "chunk_index": 11, "token_count": 218, "text_raw": "Reproject the GeoDataFrame to a projected CRS (EPSG:3035 is a good choice for Europe)\nCalculate the total area of each region (in square kilometres)\nEnsure 'Year' is converted to string if needed\nCalculate Urban Percentage\nPivot the data to have years as columns\nExtract years for plotting\nNormalize color scale across all maps based on percentage coverage\nPlot settings\nPlot each year in a separate subplot\nHide unused subplots\nStep 8: Adjust layout\nAllocate space for the color bar\nCreate a single ScalarMappable object for the common colorbar", "text_with_prefix": "EQC Quality Assessment: \"Satellite Land Cover trend assessment for Spatial Planning and Land Management\"\nDataset: satellite-land-cover [CDS]\nAspect: trend-assessment_q02 | Category: Satellite_ECVs\nSection: Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 2. Inspect and view data for the defined AoI (Iberian Peninsula) > Map urban percentage coverage over-time by NUTS regions in the AoI\n---\nReproject the GeoDataFrame to a projected CRS (EPSG:3035 is a good choice for Europe)\nCalculate the total area of each region (in square kilometres)\nEnsure 'Year' is converted to string if needed\nCalculate Urban Percentage\nPivot the data to have years as columns\nExtract years for plotting\nNormalize color scale across all maps based on percentage coverage\nPlot settings\nPlot each year in a separate subplot\nHide unused subplots\nStep 8: Adjust layout\nAllocate space for the color bar\nCreate a single ScalarMappable object for the common colorbar"} {"chunk_id": "satellite_satellite-land-cover_trend-assessment_q02__48ebb42e8809", "report_id": "satellite_satellite-land-cover_trend-assessment_q02", "dataset_id": "satellite-land-cover", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q02", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 2. Inspect and view data for the defined AoI (Iberian Peninsula) > Map Analysis", "title": "Satellite Land Cover trend assessment for Spatial Planning and Land Management", "chunk_index": 12, "token_count": 153, "text_raw": "- Over the 28-year period, all regions maintain a consistent pattern of growth.\n\n- Regions such as Comunidad Madrid and Área Metropolitana de Lisboa consistently exhibit the highest levels of urban percentage coverage.\n\n- No negative fluctuations were observed\n\n- The biggest changes are observed between 2016-2017.", "text_with_prefix": "EQC Quality Assessment: \"Satellite Land Cover trend assessment for Spatial Planning and Land Management\"\nDataset: satellite-land-cover [CDS]\nAspect: trend-assessment_q02 | Category: Satellite_ECVs\nSection: Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 2. Inspect and view data for the defined AoI (Iberian Peninsula) > Map Analysis\n---\n- Over the 28-year period, all regions maintain a consistent pattern of growth.\n\n- Regions such as Comunidad Madrid and Área Metropolitana de Lisboa consistently exhibit the highest levels of urban percentage coverage.\n\n- No negative fluctuations were observed\n\n- The biggest changes are observed between 2016-2017."} {"chunk_id": "satellite_satellite-land-cover_trend-assessment_q02__7562f31d3fec", "report_id": "satellite_satellite-land-cover_trend-assessment_q02", "dataset_id": "satellite-land-cover", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q02", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 2. Inspect and view data for the defined AoI (Iberian Peninsula) > Potential Drivers and Methodological Considerations in Urbanisation Trends in the Iberian Peninsula (1992–2022)", "title": "Satellite Land Cover trend assessment for Spatial Planning and Land Management", "chunk_index": 13, "token_count": 1041, "text_raw": "Urbanisation in the Iberian Peninsula over the last three decades has been influenced by a combination of **economic shifts**, **demographic changes**, and **policy decisions**. Before analysing trends, it is essential to account for abrupt shifts that may distort trend estimates. These shifts, often referred to as breakpoints, can result from:\n\n**Economic Drivers**\n\n- Economic Growth and Real Estate Booms: The rapid expansion of urban areas in **Madrid, Barcelona, and Lisbon** was driven by economic growth and increased investment in housing and infrastructure, particularly during the **1996–2007 real estate boom** [(González & Leal, 2018)](https://doi.org/10.1016/j.landusepol.2018.05.023). The expansion of suburban areas led to the rise of peri-urban developments in regions like **Valencia and Porto** [(Silva et al., 2017)](https://doi.org/10.1016/j.cities.2017.04.002).\n\n- Economic Crisis (2008–2013): The 2008 financial crisis led to a slowdown in urban expansion, particularly in southern Spain, where unfinished developments and ghost towns became a common sight [(Martínez & García, 2015)](https://doi.org/10.1016/j.habitatint.2015.01.010). However, in some areas, reduced construction pressure allowed for the recovery of natural land cover [(Palmero-Iniesta et al., 2021)](https://api.semanticscholar.org/CorpusID:238829360).\n\n**Demographic and Social Changes**\n\n- Rural Depopulation: Migration from rural areas to urban centers, particularly among younger populations, has led to declining populations in **Extremadura, Castilla y León, and Alentejo**, accelerating the abandonment of agricultural lands and increasing the expansion of urban areas [(Sánchez et al., 2019)](https://doi.org/10.1016/j.geoforum.2019.06.015).\n\n- Tourism and Second-Home Development: The rise of tourism-driven urbanisation has significantly reshaped coastal regions like **the Costa del Sol, Algarve, and the Balearic Islands**, where seasonal housing developments have expanded rapidly [(Rullan, 2014)](https://doi.org/10.1016/j.landusepol.2014.02.009).\n\n- Aging Population: Aging populations in rural regions have led to increased urban migration, further depopulating countryside areas and concentrating populations in cities like **Seville, Valencia, and Bilbao** [(Gutiérrez et al., 2016)](https://doi.org/10.1016/j.jrurstud.2016.03.008).\n\n**Infrastructure and Policy Influences**\n\n- EU-Funded Infrastructure Projects: Investments in transport networks, such as the expansion of **high-speed rail (AVE) in Spain and Alfa Pendular in Portugal**, have facilitated suburbanisation and commuting from secondary urban centers [(López et al., 2017)](https://doi.org/10.1016/j.jtrangeo.2017.02.005).\n\n- Urban Planning Policies: The implementation of urban growth controls and greenbelt policies in cities like **Barcelona and Porto** has influenced spatial expansion patterns, sometimes leading to increased land prices and informal settlements [(Silva & Fernandes, 2019)](https://doi.org/10.1016/j.landusepol.2019.04.012).\n\n- Smart Cities and Sustainability Initiatives: Recent efforts to create more sustainable urban environments, such as Madrid’s **Madrid Central low-emission zone**, have shaped urbanisation patterns, promoting compact city models [(Delgado & Romero, 2021)](https://doi.org/10.1016/j.cities.2021.103120).\n\n**Methodology**\n\n- Certain shifts in urban land cover may be related to data collection and processing rather than actual urban expansion. For instance, transitions between different satellite sensors can introduce artificial breaks in the time series [(Chelali et al., 2019)](https://doi.org/10.1109/JURSE.2019.8808967).\n\n- **The LC dataset is generated from a multi-sensor surface reflectance time series derived from several satellite missions. As sensor technology evolved, different instruments were used to produce consistent global composites.**\n\n
", "text_with_prefix": "EQC Quality Assessment: \"Satellite Land Cover trend assessment for Spatial Planning and Land Management\"\nDataset: satellite-land-cover [CDS]\nAspect: trend-assessment_q02 | Category: Satellite_ECVs\nSection: Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 2. Inspect and view data for the defined AoI (Iberian Peninsula) > Potential Drivers and Methodological Considerations in Urbanisation Trends in the Iberian Peninsula (1992–2022)\n---\nUrbanisation in the Iberian Peninsula over the last three decades has been influenced by a combination of **economic shifts**, **demographic changes**, and **policy decisions**. Before analysing trends, it is essential to account for abrupt shifts that may distort trend estimates. These shifts, often referred to as breakpoints, can result from:\n\n**Economic Drivers**\n\n- Economic Growth and Real Estate Booms: The rapid expansion of urban areas in **Madrid, Barcelona, and Lisbon** was driven by economic growth and increased investment in housing and infrastructure, particularly during the **1996–2007 real estate boom** [(González & Leal, 2018)](https://doi.org/10.1016/j.landusepol.2018.05.023). The expansion of suburban areas led to the rise of peri-urban developments in regions like **Valencia and Porto** [(Silva et al., 2017)](https://doi.org/10.1016/j.cities.2017.04.002).\n\n- Economic Crisis (2008–2013): The 2008 financial crisis led to a slowdown in urban expansion, particularly in southern Spain, where unfinished developments and ghost towns became a common sight [(Martínez & García, 2015)](https://doi.org/10.1016/j.habitatint.2015.01.010). However, in some areas, reduced construction pressure allowed for the recovery of natural land cover [(Palmero-Iniesta et al., 2021)](https://api.semanticscholar.org/CorpusID:238829360).\n\n**Demographic and Social Changes**\n\n- Rural Depopulation: Migration from rural areas to urban centers, particularly among younger populations, has led to declining populations in **Extremadura, Castilla y León, and Alentejo**, accelerating the abandonment of agricultural lands and increasing the expansion of urban areas [(Sánchez et al., 2019)](https://doi.org/10.1016/j.geoforum.2019.06.015).\n\n- Tourism and Second-Home Development: The rise of tourism-driven urbanisation has significantly reshaped coastal regions like **the Costa del Sol, Algarve, and the Balearic Islands**, where seasonal housing developments have expanded rapidly [(Rullan, 2014)](https://doi.org/10.1016/j.landusepol.2014.02.009).\n\n- Aging Population: Aging populations in rural regions have led to increased urban migration, further depopulating countryside areas and concentrating populations in cities like **Seville, Valencia, and Bilbao** [(Gutiérrez et al., 2016)](https://doi.org/10.1016/j.jrurstud.2016.03.008).\n\n**Infrastructure and Policy Influences**\n\n- EU-Funded Infrastructure Projects: Investments in transport networks, such as the expansion of **high-speed rail (AVE) in Spain and Alfa Pendular in Portugal**, have facilitated suburbanisation and commuting from secondary urban centers [(López et al., 2017)](https://doi.org/10.1016/j.jtrangeo.2017.02.005).\n\n- Urban Planning Policies: The implementation of urban growth controls and greenbelt policies in cities like **Barcelona and Porto** has influenced spatial expansion patterns, sometimes leading to increased land prices and informal settlements [(Silva & Fernandes, 2019)](https://doi.org/10.1016/j.landusepol.2019.04.012).\n\n- Smart Cities and Sustainability Initiatives: Recent efforts to create more sustainable urban environments, such as Madrid’s **Madrid Central low-emission zone**, have shaped urbanisation patterns, promoting compact city models [(Delgado & Romero, 2021)](https://doi.org/10.1016/j.cities.2021.103120).\n\n**Methodology**\n\n- Certain shifts in urban land cover may be related to data collection and processing rather than actual urban expansion. For instance, transitions between different satellite sensors can introduce artificial breaks in the time series [(Chelali et al., 2019)](https://doi.org/10.1109/JURSE.2019.8808967).\n\n- **The LC dataset is generated from a multi-sensor surface reflectance time series derived from several satellite missions. As sensor technology evolved, different instruments were used to produce consistent global composites.**\n\n
"} {"chunk_id": "satellite_satellite-land-cover_trend-assessment_q02__845fccf32218", "report_id": "satellite_satellite-land-cover_trend-assessment_q02", "dataset_id": "satellite-land-cover", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q02", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 2. Inspect and view data for the defined AoI (Iberian Peninsula) > Potential Drivers and Methodological Considerations in Urbanisation Trends in the Iberian Peninsula (1992–2022)", "title": "Satellite Land Cover trend assessment for Spatial Planning and Land Management", "chunk_index": 14, "token_count": 1151, "text_raw": "generated from a multi-sensor surface reflectance time series derived from several satellite missions. As sensor technology evolved, different instruments were used to produce consistent global composites.**\n\n
\n\n| Surface Reflectance (SR) Input | Reference Period | Satellite Sensor / Mission | Main Characteristics |\n|---|---|---|---|\n| AVHRR SR composites | 1992–1999 | AVHRR-2 (NOAA-11, NOAA-14) | ~1 km spatial resolution, visible and near-infrared observations |\n| SPOT-VGT SR composites | 1999–2013 | SPOT-4 / SPOT-5 VEGETATION | 1 km spatial resolution, 4 spectral bands (blue, red, NIR, SWIR) |\n| MERIS SR composites | 2003–2012 | Envisat MERIS | ~300 m spatial resolution, 15 spectral bands in visible and near-infrared |\n| PROBA-V SR composites | 2013–2019 | PROBA-V | ~300 m spatial resolution, 4 spectral bands (blue, red, NIR, SWIR) |\n| Sentinel-3 SR composites | 2020 | Sentinel-3 OLCI | ~300 m spatial resolution, multispectral observations |\n| Sentinel-3 SR composites | 2021–2022 | Sentinel-3 OLCI + SLSTR | Optical and thermal observations supporting land-cover mapping |\n
\n\n\n- Due to the limited spatial resolution of early sensors (AVHRR and SPOT-VGT), reliable detection of annual urban change was challenging prior to 2003. During this period, the representation of urban areas is guided by reference urban footprints derived from the [**Global Human Settlement Layer (GHSL)**](https://doi.org/10.2788/656115) datasets for **1990** and **2000**. These datasets provide baseline constraints for the urban class within the **1992–1999** and **2000–2003** epochs, helping to stabilise the detection of urban areas while still allowing minor variations arising from classification adjustments and mixed-pixel effects. From **2003 onwards**, urban areas are derived from higher-resolution observations. To ensure temporal consistency, **urban areas are only allowed to expand over time**, and the resulting footprint is constrained between a **minimum extent defined by GHSL 2000** and a **maximum extent defined by GHSL 2014** [(C3S - LC Algorithm Theoretical Basis Document, 2024)](https://dast.copernicus-climate.eu/documents/satellite-land-cover/WP2-FDDP-LC-2021-2022-SENTINEL3-300m-v2.1.1_ATBD_v1.2_final.pdf).\n\n- In addition, the [**Global Urban Footprint (GUF)**](https://geoservice.dlr.de/web/datasets/guf) dataset is combined with **GHSL 2014** to reduce potential omissions of built-up areas [(C3S - LC Algorithm Theoretical Basis Document, 2024)](https://dast.copernicus-climate.eu/documents/satellite-land-cover/WP2-FDDP-LC-2021-2022-SENTINEL3-300m-v2.1.1_ATBD_v1.2_final.pdf).\n\n- In addition to optical surface reflectance observations, auxiliary datasets are used to improve land-cover mapping. In particular, **Envisat ASAR Wide Swath Mode (WSM)** radar observations (2005–2012) are incorporated as ancillary information [(C3S - LC Target Requirements and Gap Analysis Document, 2024)](https://dast.copernicus-climate.eu/documents/satellite-land-cover/WP3-TR-GAD-2023_LC_v1.1_final.pdf).\n\n- While the dataset relies on multiple satellite sensors, it is not generated through simple sensor-to-sensor transitions. Instead, the global land-cover time series is built around a baseline land-cover map derived from MERIS observations (2003–2012). Earlier periods (1992–2003) are reconstructed through back-dating using AVHRR and SPOT-VGT data, while later years are produced through incremental updates using SPOT-VGT, PROBA-V, and Sentinel-3 observations. During several periods, multiple sensors are used simultaneously, with different roles (e.g., temporal consistency versus spatial refinement). As a result, potential discontinuities in the time series are more likely to arise from changes in processing strategy, data fusion, and constraints, rather than from simple sensor replacements.", "text_with_prefix": "EQC Quality Assessment: \"Satellite Land Cover trend assessment for Spatial Planning and Land Management\"\nDataset: satellite-land-cover [CDS]\nAspect: trend-assessment_q02 | Category: Satellite_ECVs\nSection: Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 2. Inspect and view data for the defined AoI (Iberian Peninsula) > Potential Drivers and Methodological Considerations in Urbanisation Trends in the Iberian Peninsula (1992–2022)\n---\ngenerated from a multi-sensor surface reflectance time series derived from several satellite missions. As sensor technology evolved, different instruments were used to produce consistent global composites.**\n\n
\n\n| Surface Reflectance (SR) Input | Reference Period | Satellite Sensor / Mission | Main Characteristics |\n|---|---|---|---|\n| AVHRR SR composites | 1992–1999 | AVHRR-2 (NOAA-11, NOAA-14) | ~1 km spatial resolution, visible and near-infrared observations |\n| SPOT-VGT SR composites | 1999–2013 | SPOT-4 / SPOT-5 VEGETATION | 1 km spatial resolution, 4 spectral bands (blue, red, NIR, SWIR) |\n| MERIS SR composites | 2003–2012 | Envisat MERIS | ~300 m spatial resolution, 15 spectral bands in visible and near-infrared |\n| PROBA-V SR composites | 2013–2019 | PROBA-V | ~300 m spatial resolution, 4 spectral bands (blue, red, NIR, SWIR) |\n| Sentinel-3 SR composites | 2020 | Sentinel-3 OLCI | ~300 m spatial resolution, multispectral observations |\n| Sentinel-3 SR composites | 2021–2022 | Sentinel-3 OLCI + SLSTR | Optical and thermal observations supporting land-cover mapping |\n
\n\n\n- Due to the limited spatial resolution of early sensors (AVHRR and SPOT-VGT), reliable detection of annual urban change was challenging prior to 2003. During this period, the representation of urban areas is guided by reference urban footprints derived from the [**Global Human Settlement Layer (GHSL)**](https://doi.org/10.2788/656115) datasets for **1990** and **2000**. These datasets provide baseline constraints for the urban class within the **1992–1999** and **2000–2003** epochs, helping to stabilise the detection of urban areas while still allowing minor variations arising from classification adjustments and mixed-pixel effects. From **2003 onwards**, urban areas are derived from higher-resolution observations. To ensure temporal consistency, **urban areas are only allowed to expand over time**, and the resulting footprint is constrained between a **minimum extent defined by GHSL 2000** and a **maximum extent defined by GHSL 2014** [(C3S - LC Algorithm Theoretical Basis Document, 2024)](https://dast.copernicus-climate.eu/documents/satellite-land-cover/WP2-FDDP-LC-2021-2022-SENTINEL3-300m-v2.1.1_ATBD_v1.2_final.pdf).\n\n- In addition, the [**Global Urban Footprint (GUF)**](https://geoservice.dlr.de/web/datasets/guf) dataset is combined with **GHSL 2014** to reduce potential omissions of built-up areas [(C3S - LC Algorithm Theoretical Basis Document, 2024)](https://dast.copernicus-climate.eu/documents/satellite-land-cover/WP2-FDDP-LC-2021-2022-SENTINEL3-300m-v2.1.1_ATBD_v1.2_final.pdf).\n\n- In addition to optical surface reflectance observations, auxiliary datasets are used to improve land-cover mapping. In particular, **Envisat ASAR Wide Swath Mode (WSM)** radar observations (2005–2012) are incorporated as ancillary information [(C3S - LC Target Requirements and Gap Analysis Document, 2024)](https://dast.copernicus-climate.eu/documents/satellite-land-cover/WP3-TR-GAD-2023_LC_v1.1_final.pdf).\n\n- While the dataset relies on multiple satellite sensors, it is not generated through simple sensor-to-sensor transitions. Instead, the global land-cover time series is built around a baseline land-cover map derived from MERIS observations (2003–2012). Earlier periods (1992–2003) are reconstructed through back-dating using AVHRR and SPOT-VGT data, while later years are produced through incremental updates using SPOT-VGT, PROBA-V, and Sentinel-3 observations. During several periods, multiple sensors are used simultaneously, with different roles (e.g., temporal consistency versus spatial refinement). As a result, potential discontinuities in the time series are more likely to arise from changes in processing strategy, data fusion, and constraints, rather than from simple sensor replacements."} {"chunk_id": "satellite_satellite-land-cover_trend-assessment_q02__17212579e36b", "report_id": "satellite_satellite-land-cover_trend-assessment_q02", "dataset_id": "satellite-land-cover", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q02", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 2. Inspect and view data for the defined AoI (Iberian Peninsula) > Potential Drivers and Methodological Considerations in Urbanisation Trends in the Iberian Peninsula (1992–2022)", "title": "Satellite Land Cover trend assessment for Spatial Planning and Land Management", "chunk_index": 15, "token_count": 595, "text_raw": "with different roles (e.g., temporal consistency versus spatial refinement). As a result, potential discontinuities in the time series are more likely to arise from changes in processing strategy, data fusion, and constraints, rather than from simple sensor replacements.\n\n
\n\n| Global LC database | Reference period | Satellite data source |\n|---|---|---|\n| Baseline 10-year global LC map | 2003–2012 | • MERIS FR/RR global SR composites between 2003 and 2012 |\n| Global annual LC maps | 1992–1999 | • Baseline 10-year global LC map
• AVHRR global SR composites between 1992 and 1999 for back-dating the baseline |\n| | 1999–2013 | • Baseline 10-year global LC map
• SPOT-VGT global SR composites between 1999 and 2013 for up- and back-dating the baseline
• MERIS FR global SR composites between 2003 and 2012 to delineate the identified changes at 300 m spatial resolution
• PROBA-V global SR composites at 300 m for year 2013 to delineate the identified changes at 300 m spatial resolution |\n| | 2014–2019 | • PROBA-V global SR composites at 1 km for years 2014 to 2019 for updating the baseline
• PROBA-V time series at 300 m for 2014 to 2019 to delineate the identified changes at the LC map spatial resolution |\n| | 2020–2022 | • S3 global SR composites at 1 km for years 2020 to 2022 for updating the baseline
• S3 time series at 300 m for 2020 to 2022 to delineate the identified changes at the LC map spatial resolution |\n
\n\n\n(code-section-3)=", "text_with_prefix": "EQC Quality Assessment: \"Satellite Land Cover trend assessment for Spatial Planning and Land Management\"\nDataset: satellite-land-cover [CDS]\nAspect: trend-assessment_q02 | Category: Satellite_ECVs\nSection: Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 2. Inspect and view data for the defined AoI (Iberian Peninsula) > Potential Drivers and Methodological Considerations in Urbanisation Trends in the Iberian Peninsula (1992–2022)\n---\nwith different roles (e.g., temporal consistency versus spatial refinement). As a result, potential discontinuities in the time series are more likely to arise from changes in processing strategy, data fusion, and constraints, rather than from simple sensor replacements.\n\n
\n\n| Global LC database | Reference period | Satellite data source |\n|---|---|---|\n| Baseline 10-year global LC map | 2003–2012 | • MERIS FR/RR global SR composites between 2003 and 2012 |\n| Global annual LC maps | 1992–1999 | • Baseline 10-year global LC map
• AVHRR global SR composites between 1992 and 1999 for back-dating the baseline |\n| | 1999–2013 | • Baseline 10-year global LC map
• SPOT-VGT global SR composites between 1999 and 2013 for up- and back-dating the baseline
• MERIS FR global SR composites between 2003 and 2012 to delineate the identified changes at 300 m spatial resolution
• PROBA-V global SR composites at 300 m for year 2013 to delineate the identified changes at 300 m spatial resolution |\n| | 2014–2019 | • PROBA-V global SR composites at 1 km for years 2014 to 2019 for updating the baseline
• PROBA-V time series at 300 m for 2014 to 2019 to delineate the identified changes at the LC map spatial resolution |\n| | 2020–2022 | • S3 global SR composites at 1 km for years 2020 to 2022 for updating the baseline
• S3 time series at 300 m for 2020 to 2022 to delineate the identified changes at the LC map spatial resolution |\n
\n\n\n(code-section-3)="} {"chunk_id": "satellite_satellite-land-cover_trend-assessment_q02__74ded7afb5a7", "report_id": "satellite_satellite-land-cover_trend-assessment_q02", "dataset_id": "satellite-land-cover", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q02", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 3. Breakpoint Detection", "title": "Satellite Land Cover trend assessment for Spatial Planning and Land Management", "chunk_index": 16, "token_count": 990, "text_raw": "A first step in assessing temporal behaviour is to analyse the **rate of change** of the urban-area time series. This provides an intuitive view of how the pace of urban expansion evolves over time and helps highlight periods where the trajectory deviates from its typical pattern.\n\nHowever, while such exploratory diagnostics identify points of interest, they do not by themselves confirm whether the observed changes are structurally meaningful. Year-to-year variations may reflect noise, smoothing effects, or gradual adjustments in the underlying product. To formally assess structural changes, a **statistical breakpoint detection method** is applied.\n\nThe analysis therefore combines **rate-of-change diagnostics** with **statistical segmentation**, with particular attention to the monotonic nature of the data and the need to avoid over-interpreting gradual trends as discrete breakpoints.\n\n---\n\n**Rate of Change Calculation**\n\n- The **first derivative** of the urban-area time series is computed to quantify the **year-to-year change in urban extent**.\n- This highlights **periods where the pace of urban expansion deviates from its typical behaviour**, providing an initial indication of potential anomalies or shifts in the series.\n\n---\n\n**Statistical Thresholding of Derivative Spikes**\n\n- To identify unusually strong deviations, a dynamic threshold is defined as:\n\n**Threshold = mean(|rate of change|) + 1.5 × standard deviation**\n\n- Years where the **absolute rate of change exceeds this threshold** are flagged as **derivative spikes**.\n- These spikes represent periods where the magnitude of change is unusually large relative to the typical variability of the series.\n- This step serves as a **diagnostic indicator of anomalous growth behaviour**, but does not in itself define structural breakpoints.\n\n---\n\n**Breakpoint Detection Using the PELT Algorithm**\n\n- The **Pruned Exact Linear Time (PELT)** algorithm is used to detect **changes in the statistical behaviour of the growth rate**, rather than in the cumulative urban-area series itself.\n- This is important because the urban-area time series is **monotonic and non-decreasing**, meaning that applying segmentation directly to the cumulative values can lead to artificial breakpoints driven by gradual growth.\n\n- The algorithm is therefore applied to the **first derivative (growth rate)**, allowing detection of:\n - transitions between periods with different **rates of urban expansion**, rather than absolute levels.\n\n- The implementation uses:\n - the **L2 cost function** (`model=\"l2\"`) to minimise variance within segments,\n - a **fixed penalty parameter** (`pen = 2.5`) applied to a standardised growth-rate series,\n - and a **minimum segment length (`min_size = 3`)** to avoid short, unstable segments.\n\n- Prior to segmentation, the growth-rate series may be **standardised** to ensure comparability across regions with different magnitudes of urban change.\n\n---\n\n**Breakpoint Selection**\n\n- Breakpoints are identified directly from the segmentation of the growth-rate series.\n- Unlike earlier approaches, **no explicit filtering based on derivative spikes is applied**, as this was found to introduce additional assumptions and potentially remove meaningful structural changes.\n- Instead, derivative spikes and breakpoints are analysed **in parallel**, allowing independent but complementary perspectives on the temporal behaviour of the series.\n\n---\n\n**Methodological Considerations**\n\n- The urban-area time series is **strictly monotonic and cumulative by construction**, reflecting constraints in the land-cover product that prevent decreases in urban extent.\n- As a result:\n - segmentation applied to the raw series tends to produce **regularly spaced breakpoints**, even in the absence of true structural changes.\n - applying breakpoint detection to the **growth rate** mitigates this issue by focusing on changes in dynamics rather than accumulated values.\n\n- Even when applied to growth rates, breakpoint detection should be interpreted cautiously:\n - some detected changes may reflect **gradual adjustments in processing or classification**, rather than abrupt real-world events.\n - others may correspond to **genuine shifts in the pace of urban expansion**.\n\n- For this reason, breakpoint detection is treated as a **diagnostic tool**, not as definitive evidence of discrete events.\n\n---\n\n**Interpreting Breakpoint Timing**\n\nBreakpoint positions should be interpreted as **approximate indicators of when a change in growth dynamics begins**, rather than exact points of maximum change.\n\nBecause segmentation identifies **changes in statistical behaviour**, the detected breakpoint may:\n\n- precede the most visible increase in growth rate, or \n- occur within a broader transition period.\n\nAs a result, breakpoint timing should be considered **indicative rather than exact**, and interpreted in conjunction with derivative-based diagnostics.", "text_with_prefix": "EQC Quality Assessment: \"Satellite Land Cover trend assessment for Spatial Planning and Land Management\"\nDataset: satellite-land-cover [CDS]\nAspect: trend-assessment_q02 | Category: Satellite_ECVs\nSection: Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 3. Breakpoint Detection\n---\nA first step in assessing temporal behaviour is to analyse the **rate of change** of the urban-area time series. This provides an intuitive view of how the pace of urban expansion evolves over time and helps highlight periods where the trajectory deviates from its typical pattern.\n\nHowever, while such exploratory diagnostics identify points of interest, they do not by themselves confirm whether the observed changes are structurally meaningful. Year-to-year variations may reflect noise, smoothing effects, or gradual adjustments in the underlying product. To formally assess structural changes, a **statistical breakpoint detection method** is applied.\n\nThe analysis therefore combines **rate-of-change diagnostics** with **statistical segmentation**, with particular attention to the monotonic nature of the data and the need to avoid over-interpreting gradual trends as discrete breakpoints.\n\n---\n\n**Rate of Change Calculation**\n\n- The **first derivative** of the urban-area time series is computed to quantify the **year-to-year change in urban extent**.\n- This highlights **periods where the pace of urban expansion deviates from its typical behaviour**, providing an initial indication of potential anomalies or shifts in the series.\n\n---\n\n**Statistical Thresholding of Derivative Spikes**\n\n- To identify unusually strong deviations, a dynamic threshold is defined as:\n\n**Threshold = mean(|rate of change|) + 1.5 × standard deviation**\n\n- Years where the **absolute rate of change exceeds this threshold** are flagged as **derivative spikes**.\n- These spikes represent periods where the magnitude of change is unusually large relative to the typical variability of the series.\n- This step serves as a **diagnostic indicator of anomalous growth behaviour**, but does not in itself define structural breakpoints.\n\n---\n\n**Breakpoint Detection Using the PELT Algorithm**\n\n- The **Pruned Exact Linear Time (PELT)** algorithm is used to detect **changes in the statistical behaviour of the growth rate**, rather than in the cumulative urban-area series itself.\n- This is important because the urban-area time series is **monotonic and non-decreasing**, meaning that applying segmentation directly to the cumulative values can lead to artificial breakpoints driven by gradual growth.\n\n- The algorithm is therefore applied to the **first derivative (growth rate)**, allowing detection of:\n - transitions between periods with different **rates of urban expansion**, rather than absolute levels.\n\n- The implementation uses:\n - the **L2 cost function** (`model=\"l2\"`) to minimise variance within segments,\n - a **fixed penalty parameter** (`pen = 2.5`) applied to a standardised growth-rate series,\n - and a **minimum segment length (`min_size = 3`)** to avoid short, unstable segments.\n\n- Prior to segmentation, the growth-rate series may be **standardised** to ensure comparability across regions with different magnitudes of urban change.\n\n---\n\n**Breakpoint Selection**\n\n- Breakpoints are identified directly from the segmentation of the growth-rate series.\n- Unlike earlier approaches, **no explicit filtering based on derivative spikes is applied**, as this was found to introduce additional assumptions and potentially remove meaningful structural changes.\n- Instead, derivative spikes and breakpoints are analysed **in parallel**, allowing independent but complementary perspectives on the temporal behaviour of the series.\n\n---\n\n**Methodological Considerations**\n\n- The urban-area time series is **strictly monotonic and cumulative by construction**, reflecting constraints in the land-cover product that prevent decreases in urban extent.\n- As a result:\n - segmentation applied to the raw series tends to produce **regularly spaced breakpoints**, even in the absence of true structural changes.\n - applying breakpoint detection to the **growth rate** mitigates this issue by focusing on changes in dynamics rather than accumulated values.\n\n- Even when applied to growth rates, breakpoint detection should be interpreted cautiously:\n - some detected changes may reflect **gradual adjustments in processing or classification**, rather than abrupt real-world events.\n - others may correspond to **genuine shifts in the pace of urban expansion**.\n\n- For this reason, breakpoint detection is treated as a **diagnostic tool**, not as definitive evidence of discrete events.\n\n---\n\n**Interpreting Breakpoint Timing**\n\nBreakpoint positions should be interpreted as **approximate indicators of when a change in growth dynamics begins**, rather than exact points of maximum change.\n\nBecause segmentation identifies **changes in statistical behaviour**, the detected breakpoint may:\n\n- precede the most visible increase in growth rate, or \n- occur within a broader transition period.\n\nAs a result, breakpoint timing should be considered **indicative rather than exact**, and interpreted in conjunction with derivative-based diagnostics."} {"chunk_id": "satellite_satellite-land-cover_trend-assessment_q02__0fac6402f819", "report_id": "satellite_satellite-land-cover_trend-assessment_q02", "dataset_id": "satellite-land-cover", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q02", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 3. Breakpoint Detection", "title": "Satellite Land Cover trend assessment for Spatial Planning and Land Management", "chunk_index": 17, "token_count": 1026, "text_raw": "precede the most visible increase in growth rate, or \n- occur within a broader transition period.\n\nAs a result, breakpoint timing should be considered **indicative rather than exact**, and interpreted in conjunction with derivative-based diagnostics.\n\nCalculate breakpoints and derivative spikes for each region\nGHSL change years shown as hard-line transitions\nMidpoint used for before/after GHSL switch classification\nNumber of subplot columns in the regional figure layout\nThreshold multiplier used to flag unusually large derivative spikes:\nspike if |growth rate| > mean(|growth rate|) + spike_std_factor * std(growth rate)\nMinimum number of derivative observations allowed in each PELT segment.\nPenalty term for PELT breakpoint detection.\nStandardize the growth-rate series before applying PELT so that breakpoint\ndetection is comparable across regions with different magnitudes of urban change.\nremove invalid and terminal interval\n==================================================\n==================================================\n==================================================\n==================================================\n\nPlot Breakpoints Series\nStore spikes\nStore breakpoints\nBackground phases\nGHSL change years as hard lines\nRemove empty axes\nProcessing phases legend\nLine/signal legend\nCompact spacing between legends and plots\n\nTo assess whether detected discontinuities in the urban-area time series may be influenced by methodological factors, a targeted alignment analysis is performed. The aim is to evaluate whether **growth-rate breakpoints** and **derivative spikes** occur shortly after known **processing-chain changes**.\n\nFor each detected event, the time difference relative to each processing-change year is computed. An event is considered potentially related to a processing change only if it occurs **within the two years following that transition**. This forward-looking window reflects the temporal characteristics of the dataset, where land-cover changes are only confirmed if they persist over two consecutive years.\n\nIn addition to temporal alignment, the analysis also considers the **spatial consistency of detected signals**, quantified as the number of regions affected within a given interval. Methodological effects are expected to influence the dataset in a systematic way, and therefore to appear consistently across multiple regions. In contrast, signals affecting only a small number of regions are more likely to reflect **local dynamics** or region-specific variations rather than processing-related artefacts.\n\nThe results are summarised in **four complementary tables**, distinguishing between:\n\n- **Event type**: breakpoints vs derivative spikes \n- **Analytical scope**: aligned vs full set of identified breakpoints and derivative spikes\n\nThe first two tables report only events that occur **within two years after a processing transition**, while the two additional tables report **all detected events**.\n\nPlot the aligned breakpoints and derivative spikes with known methodological changes present in more than half of the regions\nBuild tables\n\n```text\n### Breakpoint alignment table ###\n```\n\n```text\n### Derivative spike alignment table ###\nNo derivative spike signals affect at least 10 regions within the known methodological changes transition windows\n```\n\nPlot all breakpoints and derivative spikes\nRename Interval column\nAdd relation to processing transition\nDrop helper column\nSort by number of regions, then by interval\nBuild tables\nDisplay nicely\n\n```text\n### Breakpoint Table ###\n```\n\n```text\nintervals number of regions \\\n0 2016-2017 13 \n1 2001-2002 12 \n2 2006-2007 6 \n3 1996-1997 3 \n4 2011-2012 2\n\nregions name \\\n0 Cantabria, Castilla y León, Castilla-La Mancha, Cataluña, Centro (PT), Comunidad Foral de Navarra, Comunitat Valenciana, Galicia, Norte, País Vasco, Principado de Asturias, Región de Murcia, Área Metropolitana de Lisboa \n1 Alentejo, Algarve, Aragón, Cantabria, Castilla y León, Castilla-La Mancha, Centro (PT), Comunidad Foral de Navarra, Extremadura, Galicia, Illes Balears, La Rioja \n2 Alentejo, Algarve, Andalucía, Cataluña, Illes Balears, País Vasco \n3 Cataluña, Comunitat Valenciana, País Vasco \n4 Castilla y León, Región de Murcia\n\nnearest processing changes \n0 None within 2 years \n1 GHSL_switch (2000) \n2 None within 2 years \n3 None within 2 years \n4 None within 2 years\n```\n\n```text\n### Derivative Spike Table ###\n```", "text_with_prefix": "EQC Quality Assessment: \"Satellite Land Cover trend assessment for Spatial Planning and Land Management\"\nDataset: satellite-land-cover [CDS]\nAspect: trend-assessment_q02 | Category: Satellite_ECVs\nSection: Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 3. Breakpoint Detection\n---\nprecede the most visible increase in growth rate, or \n- occur within a broader transition period.\n\nAs a result, breakpoint timing should be considered **indicative rather than exact**, and interpreted in conjunction with derivative-based diagnostics.\n\nCalculate breakpoints and derivative spikes for each region\nGHSL change years shown as hard-line transitions\nMidpoint used for before/after GHSL switch classification\nNumber of subplot columns in the regional figure layout\nThreshold multiplier used to flag unusually large derivative spikes:\nspike if |growth rate| > mean(|growth rate|) + spike_std_factor * std(growth rate)\nMinimum number of derivative observations allowed in each PELT segment.\nPenalty term for PELT breakpoint detection.\nStandardize the growth-rate series before applying PELT so that breakpoint\ndetection is comparable across regions with different magnitudes of urban change.\nremove invalid and terminal interval\n==================================================\n==================================================\n==================================================\n==================================================\n\nPlot Breakpoints Series\nStore spikes\nStore breakpoints\nBackground phases\nGHSL change years as hard lines\nRemove empty axes\nProcessing phases legend\nLine/signal legend\nCompact spacing between legends and plots\n\nTo assess whether detected discontinuities in the urban-area time series may be influenced by methodological factors, a targeted alignment analysis is performed. The aim is to evaluate whether **growth-rate breakpoints** and **derivative spikes** occur shortly after known **processing-chain changes**.\n\nFor each detected event, the time difference relative to each processing-change year is computed. An event is considered potentially related to a processing change only if it occurs **within the two years following that transition**. This forward-looking window reflects the temporal characteristics of the dataset, where land-cover changes are only confirmed if they persist over two consecutive years.\n\nIn addition to temporal alignment, the analysis also considers the **spatial consistency of detected signals**, quantified as the number of regions affected within a given interval. Methodological effects are expected to influence the dataset in a systematic way, and therefore to appear consistently across multiple regions. In contrast, signals affecting only a small number of regions are more likely to reflect **local dynamics** or region-specific variations rather than processing-related artefacts.\n\nThe results are summarised in **four complementary tables**, distinguishing between:\n\n- **Event type**: breakpoints vs derivative spikes \n- **Analytical scope**: aligned vs full set of identified breakpoints and derivative spikes\n\nThe first two tables report only events that occur **within two years after a processing transition**, while the two additional tables report **all detected events**.\n\nPlot the aligned breakpoints and derivative spikes with known methodological changes present in more than half of the regions\nBuild tables\n\n```text\n### Breakpoint alignment table ###\n```\n\n```text\n### Derivative spike alignment table ###\nNo derivative spike signals affect at least 10 regions within the known methodological changes transition windows\n```\n\nPlot all breakpoints and derivative spikes\nRename Interval column\nAdd relation to processing transition\nDrop helper column\nSort by number of regions, then by interval\nBuild tables\nDisplay nicely\n\n```text\n### Breakpoint Table ###\n```\n\n```text\nintervals number of regions \\\n0 2016-2017 13 \n1 2001-2002 12 \n2 2006-2007 6 \n3 1996-1997 3 \n4 2011-2012 2\n\nregions name \\\n0 Cantabria, Castilla y León, Castilla-La Mancha, Cataluña, Centro (PT), Comunidad Foral de Navarra, Comunitat Valenciana, Galicia, Norte, País Vasco, Principado de Asturias, Región de Murcia, Área Metropolitana de Lisboa \n1 Alentejo, Algarve, Aragón, Cantabria, Castilla y León, Castilla-La Mancha, Centro (PT), Comunidad Foral de Navarra, Extremadura, Galicia, Illes Balears, La Rioja \n2 Alentejo, Algarve, Andalucía, Cataluña, Illes Balears, País Vasco \n3 Cataluña, Comunitat Valenciana, País Vasco \n4 Castilla y León, Región de Murcia\n\nnearest processing changes \n0 None within 2 years \n1 GHSL_switch (2000) \n2 None within 2 years \n3 None within 2 years \n4 None within 2 years\n```\n\n```text\n### Derivative Spike Table ###\n```"} {"chunk_id": "satellite_satellite-land-cover_trend-assessment_q02__bb0e4ed4c64d", "report_id": "satellite_satellite-land-cover_trend-assessment_q02", "dataset_id": "satellite-land-cover", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q02", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 3. Breakpoint Detection", "title": "Satellite Land Cover trend assessment for Spatial Planning and Land Management", "chunk_index": 18, "token_count": 710, "text_raw": "GHSL_switch (2000) \n2 None within 2 years \n3 None within 2 years \n4 None within 2 years\n```\n\n```text\n### Derivative Spike Table ###\n```\n\n```text\nintervals number of regions \\\n0 2016-2017 8 \n1 2013-2014 7 \n2 2012-2013 5 \n3 2001-2002 4 \n4 2002-2003 4 \n5 2017-2018 4 \n6 2000-2001 2 \n7 2003-2004 2 \n8 2004-2005 2 \n9 2006-2007 2 \n10 2014-2015 2 \n11 2005-2006 1\n\nregions name \\\n0 Andalucía, Castilla y León, Castilla-La Mancha, Comunidad Foral de Navarra, Comunidad de Madrid, Comunitat Valenciana, La Rioja, Región de Murcia \n1 Andalucía, Castilla y León, Castilla-La Mancha, Centro (PT), Galicia, Principado de Asturias, Región de Murcia \n2 Andalucía, Aragón, Centro (PT), Región de Murcia, Área Metropolitana de Lisboa \n3 Algarve, Cataluña, Norte, Área Metropolitana de Lisboa \n4 Alentejo, Algarve, Cataluña, Illes Balears \n5 Andalucía, Aragón, Extremadura, La Rioja \n6 Cataluña, Área Metropolitana de Lisboa \n7 Algarve, Illes Balears \n8 Norte, País Vasco \n9 Comunidad Foral de Navarra, Extremadura \n10 Cantabria, Galicia \n11 País Vasco\n\nnearest processing changes \n0 None within 2 years \n1 SPOT-VGT_finish (2013); MERIS_baseline_finish (2012); PROBAV_start (2013) \n2 MERIS_baseline_finish (2012) \n3 GHSL_switch (2000) \n4 None within 2 years \n5 None within 2 years \n6 GHSL_switch (2000); AVHRR_finish (1999); SPOT-VGT_start (1999) \n7 GHSL_switch (2003); MERIS_baseline_start (2003) \n8 GHSL_switch (2003); MERIS_baseline_start (2003) \n9 None within 2 years \n10 SPOT-VGT_finish (2013); PROBAV_start (2013) \n11 None within 2 years\n```", "text_with_prefix": "EQC Quality Assessment: \"Satellite Land Cover trend assessment for Spatial Planning and Land Management\"\nDataset: satellite-land-cover [CDS]\nAspect: trend-assessment_q02 | Category: Satellite_ECVs\nSection: Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 3. Breakpoint Detection\n---\nGHSL_switch (2000) \n2 None within 2 years \n3 None within 2 years \n4 None within 2 years\n```\n\n```text\n### Derivative Spike Table ###\n```\n\n```text\nintervals number of regions \\\n0 2016-2017 8 \n1 2013-2014 7 \n2 2012-2013 5 \n3 2001-2002 4 \n4 2002-2003 4 \n5 2017-2018 4 \n6 2000-2001 2 \n7 2003-2004 2 \n8 2004-2005 2 \n9 2006-2007 2 \n10 2014-2015 2 \n11 2005-2006 1\n\nregions name \\\n0 Andalucía, Castilla y León, Castilla-La Mancha, Comunidad Foral de Navarra, Comunidad de Madrid, Comunitat Valenciana, La Rioja, Región de Murcia \n1 Andalucía, Castilla y León, Castilla-La Mancha, Centro (PT), Galicia, Principado de Asturias, Región de Murcia \n2 Andalucía, Aragón, Centro (PT), Región de Murcia, Área Metropolitana de Lisboa \n3 Algarve, Cataluña, Norte, Área Metropolitana de Lisboa \n4 Alentejo, Algarve, Cataluña, Illes Balears \n5 Andalucía, Aragón, Extremadura, La Rioja \n6 Cataluña, Área Metropolitana de Lisboa \n7 Algarve, Illes Balears \n8 Norte, País Vasco \n9 Comunidad Foral de Navarra, Extremadura \n10 Cantabria, Galicia \n11 País Vasco\n\nnearest processing changes \n0 None within 2 years \n1 SPOT-VGT_finish (2013); MERIS_baseline_finish (2012); PROBAV_start (2013) \n2 MERIS_baseline_finish (2012) \n3 GHSL_switch (2000) \n4 None within 2 years \n5 None within 2 years \n6 GHSL_switch (2000); AVHRR_finish (1999); SPOT-VGT_start (1999) \n7 GHSL_switch (2003); MERIS_baseline_start (2003) \n8 GHSL_switch (2003); MERIS_baseline_start (2003) \n9 None within 2 years \n10 SPOT-VGT_finish (2013); PROBAV_start (2013) \n11 None within 2 years\n```"} {"chunk_id": "satellite_satellite-land-cover_trend-assessment_q02__041e0e734970", "report_id": "satellite_satellite-land-cover_trend-assessment_q02", "dataset_id": "satellite-land-cover", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q02", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 3. Breakpoint Detection > Breakpoint Analysis", "title": "Satellite Land Cover trend assessment for Spatial Planning and Land Management", "chunk_index": 19, "token_count": 1077, "text_raw": "Breakpoint detection and derivative analysis capture two complementary aspects of temporal dynamics in land cover time series, a distinction widely recognised in change detection literature ([Chelali et al., 2019](https://doi.org/10.1109/JURSE.2019.8808967); [Chang et al., 2018](https://doi.org/10.1088/1755-1315/113/1/012087)):\n\n- **Breakpoints** identify the onset of structural changes in the time series \n- **Derivative spikes** highlight years of maximum year-to-year variation\n\nBecause of this, spikes may occur slightly after breakpoints, reflecting the difference between **initiation** and **peak expression** of change ([Chelali et al., 2019](https://doi.org/10.1109/JURSE.2019.8808967)).\n\nThe interpretation is guided by:\n\n- **temporal alignment with processing-chain transitions**, and \n- **spatial consistency across regions**\n\nThis is particularly important in satellite-derived land cover products, where methodological updates (e.g. sensor transitions, classification schemes, or reference layers) can introduce artificial discontinuities in time series.\n\nA signal affecting many regions simultaneously is more likely to reflect a **systematic driver** (either methodological or large-scale real dynamics), whereas signals affecting few regions are more likely **local effects** ([Chang et al., 2018](https://doi.org/10.1088/1755-1315/113/1/012087)).\n\n---\n\n**Breakpoint ~2016–2017**\n\nThis period exhibits the **highest spatial consistency**, with breakpoints detected across a large majority of regions. The signal is also supported by concurrent derivative spikes, indicating both **structural change** and **increased short-term variability** in the time series.\n\nImportantly, this breakpoint cluster shows **no temporal alignment with any identified processing-chain transition**, ruling out a methodological origin.\n\n**Interpretation** \nThis pattern provides strong evidence of a **genuine structural shift in urban dynamics**. The combination of high spatial coherence and lack of methodological alignment suggests a system-wide driver, consistent with the **post-2008 recovery phase**, during which urban growth resumed after a period of contraction ([Martínez & García, 2015](https://doi.org/10.1016/j.habitatint.2015.01.010); [Palmero-Iniesta et al., 2021](https://api.semanticscholar.org/CorpusID:238829360)).\n\nMore broadly, renewed urban expansion in the Iberian Peninsula during the 2010s has been linked to economic recovery, infrastructure investment, and renewed real estate activity ([González & Leal, 2018](https://doi.org/10.1016/j.landusepol.2018.05.023); [López et al., 2017](https://doi.org/10.1016/j.jtrangeo.2017.02.005)). The breakpoint reflects a sustained change in growth trajectory, while the associated spikes indicate short-term adjustments during this transition.\n\n---\n\n**Breakpoint ~2001–2002**\n\nA prominent breakpoint cluster is observed in this period, affecting a large number of regions and indicating a **highly coherent spatial signal**. This interval falls within the **transition window associated with the GHSL reference update (1999–2000)**.\n\n**Interpretation** \nThis signal is best understood as a **composite effect**, arising from the interaction of:\n\n- **Methodological influences**, particularly the constraint imposed by changes in reference datasets such as GSHL, which can introduce step-like adjustments in land cover products \n- **Underlying urban expansion dynamics**, corresponding to the early phase of the Iberian housing boom, characterised by rapid land consumption and development pressure ([González & Leal, 2018](https://doi.org/10.1016/j.landusepol.2018.05.023))\n\nThe persistence of the breakpoint signal across regions suggests a structural component; however, its temporal proximity to a major processing transition prevents a clear separation between **artefact and real change**.\n\n---\n\n**Breakpoint ~2006–2007**\n\nThis breakpoint cluster affects a **moderate number of regions** and is not accompanied by strong or synchronised derivative spikes, indicating the absence of abrupt year-to-year changes.\n\nNo temporal association is found with processing-chain transitions.\n\n**Interpretation** \nThis pattern reflects a **progressive structural adjustment** rather than a sudden shift. The breakpoint likely captures a **change in the underlying growth regime**, consistent with the late phase of the Iberian housing expansion cycle ([González & Leal, 2018](https://doi.org/10.1016/j.landusepol.2018.05.023)).", "text_with_prefix": "EQC Quality Assessment: \"Satellite Land Cover trend assessment for Spatial Planning and Land Management\"\nDataset: satellite-land-cover [CDS]\nAspect: trend-assessment_q02 | Category: Satellite_ECVs\nSection: Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 3. Breakpoint Detection > Breakpoint Analysis\n---\nBreakpoint detection and derivative analysis capture two complementary aspects of temporal dynamics in land cover time series, a distinction widely recognised in change detection literature ([Chelali et al., 2019](https://doi.org/10.1109/JURSE.2019.8808967); [Chang et al., 2018](https://doi.org/10.1088/1755-1315/113/1/012087)):\n\n- **Breakpoints** identify the onset of structural changes in the time series \n- **Derivative spikes** highlight years of maximum year-to-year variation\n\nBecause of this, spikes may occur slightly after breakpoints, reflecting the difference between **initiation** and **peak expression** of change ([Chelali et al., 2019](https://doi.org/10.1109/JURSE.2019.8808967)).\n\nThe interpretation is guided by:\n\n- **temporal alignment with processing-chain transitions**, and \n- **spatial consistency across regions**\n\nThis is particularly important in satellite-derived land cover products, where methodological updates (e.g. sensor transitions, classification schemes, or reference layers) can introduce artificial discontinuities in time series.\n\nA signal affecting many regions simultaneously is more likely to reflect a **systematic driver** (either methodological or large-scale real dynamics), whereas signals affecting few regions are more likely **local effects** ([Chang et al., 2018](https://doi.org/10.1088/1755-1315/113/1/012087)).\n\n---\n\n**Breakpoint ~2016–2017**\n\nThis period exhibits the **highest spatial consistency**, with breakpoints detected across a large majority of regions. The signal is also supported by concurrent derivative spikes, indicating both **structural change** and **increased short-term variability** in the time series.\n\nImportantly, this breakpoint cluster shows **no temporal alignment with any identified processing-chain transition**, ruling out a methodological origin.\n\n**Interpretation** \nThis pattern provides strong evidence of a **genuine structural shift in urban dynamics**. The combination of high spatial coherence and lack of methodological alignment suggests a system-wide driver, consistent with the **post-2008 recovery phase**, during which urban growth resumed after a period of contraction ([Martínez & García, 2015](https://doi.org/10.1016/j.habitatint.2015.01.010); [Palmero-Iniesta et al., 2021](https://api.semanticscholar.org/CorpusID:238829360)).\n\nMore broadly, renewed urban expansion in the Iberian Peninsula during the 2010s has been linked to economic recovery, infrastructure investment, and renewed real estate activity ([González & Leal, 2018](https://doi.org/10.1016/j.landusepol.2018.05.023); [López et al., 2017](https://doi.org/10.1016/j.jtrangeo.2017.02.005)). The breakpoint reflects a sustained change in growth trajectory, while the associated spikes indicate short-term adjustments during this transition.\n\n---\n\n**Breakpoint ~2001–2002**\n\nA prominent breakpoint cluster is observed in this period, affecting a large number of regions and indicating a **highly coherent spatial signal**. This interval falls within the **transition window associated with the GHSL reference update (1999–2000)**.\n\n**Interpretation** \nThis signal is best understood as a **composite effect**, arising from the interaction of:\n\n- **Methodological influences**, particularly the constraint imposed by changes in reference datasets such as GSHL, which can introduce step-like adjustments in land cover products \n- **Underlying urban expansion dynamics**, corresponding to the early phase of the Iberian housing boom, characterised by rapid land consumption and development pressure ([González & Leal, 2018](https://doi.org/10.1016/j.landusepol.2018.05.023))\n\nThe persistence of the breakpoint signal across regions suggests a structural component; however, its temporal proximity to a major processing transition prevents a clear separation between **artefact and real change**.\n\n---\n\n**Breakpoint ~2006–2007**\n\nThis breakpoint cluster affects a **moderate number of regions** and is not accompanied by strong or synchronised derivative spikes, indicating the absence of abrupt year-to-year changes.\n\nNo temporal association is found with processing-chain transitions.\n\n**Interpretation** \nThis pattern reflects a **progressive structural adjustment** rather than a sudden shift. The breakpoint likely captures a **change in the underlying growth regime**, consistent with the late phase of the Iberian housing expansion cycle ([González & Leal, 2018](https://doi.org/10.1016/j.landusepol.2018.05.023))."} {"chunk_id": "satellite_satellite-land-cover_trend-assessment_q02__9e768046a1a5", "report_id": "satellite_satellite-land-cover_trend-assessment_q02", "dataset_id": "satellite-land-cover", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q02", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 3. Breakpoint Detection > Breakpoint Analysis", "title": "Satellite Land Cover trend assessment for Spatial Planning and Land Management", "chunk_index": 20, "token_count": 468, "text_raw": "underlying growth regime**, consistent with the late phase of the Iberian housing expansion cycle ([González & Leal, 2018](https://doi.org/10.1016/j.landusepol.2018.05.023)).\n\nThe absence of strong derivative spikes suggests that the transition occurred through **gradual changes in growth rate**, rather than discrete jumps, supporting its interpretation as a **real, process-driven signal**.\n\n---\n\n**Minor breakpoint clusters**\n\nAdditional breakpoint detections are observed, each affecting a **limited number of regions** and lacking both strong spatial coherence and consistent alignment with processing transitions.\n\n**Interpretation** \nThese signals are interpreted as **localised or context-specific dynamics**, rather than manifestations of system-wide processes. Their limited spatial extent and lack of temporal consistency indicate that they do not reflect either:\n\n- large-scale methodological artefacts \n- coherent regional-scale urban transitions\n\nInstead, they likely correspond to **region-specific developments**, such as:\n\n- tourism-driven coastal urbanisation ([Rullan, 2014](https://doi.org/10.1016/j.landusepol.2014.02.009)) \n- peri-urban expansion processes ([Silva et al., 2017](https://doi.org/10.1016/j.cities.2017.04.002)) \n- planning policies and land-market dynamics ([Silva & Fernandes, 2019](https://doi.org/10.1016/j.landusepol.2019.04.012)) \n- demographic shifts, including rural depopulation and urban migration ([Sánchez et al., 2019](https://doi.org/10.1016/j.geoforum.2019.06.015); [Gutiérrez et al., 2016](https://doi.org/10.1016/j.jrurstud.2016.03.008))\n\n(code-section-4)=", "text_with_prefix": "EQC Quality Assessment: \"Satellite Land Cover trend assessment for Spatial Planning and Land Management\"\nDataset: satellite-land-cover [CDS]\nAspect: trend-assessment_q02 | Category: Satellite_ECVs\nSection: Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 3. Breakpoint Detection > Breakpoint Analysis\n---\nunderlying growth regime**, consistent with the late phase of the Iberian housing expansion cycle ([González & Leal, 2018](https://doi.org/10.1016/j.landusepol.2018.05.023)).\n\nThe absence of strong derivative spikes suggests that the transition occurred through **gradual changes in growth rate**, rather than discrete jumps, supporting its interpretation as a **real, process-driven signal**.\n\n---\n\n**Minor breakpoint clusters**\n\nAdditional breakpoint detections are observed, each affecting a **limited number of regions** and lacking both strong spatial coherence and consistent alignment with processing transitions.\n\n**Interpretation** \nThese signals are interpreted as **localised or context-specific dynamics**, rather than manifestations of system-wide processes. Their limited spatial extent and lack of temporal consistency indicate that they do not reflect either:\n\n- large-scale methodological artefacts \n- coherent regional-scale urban transitions\n\nInstead, they likely correspond to **region-specific developments**, such as:\n\n- tourism-driven coastal urbanisation ([Rullan, 2014](https://doi.org/10.1016/j.landusepol.2014.02.009)) \n- peri-urban expansion processes ([Silva et al., 2017](https://doi.org/10.1016/j.cities.2017.04.002)) \n- planning policies and land-market dynamics ([Silva & Fernandes, 2019](https://doi.org/10.1016/j.landusepol.2019.04.012)) \n- demographic shifts, including rural depopulation and urban migration ([Sánchez et al., 2019](https://doi.org/10.1016/j.geoforum.2019.06.015); [Gutiérrez et al., 2016](https://doi.org/10.1016/j.jrurstud.2016.03.008))\n\n(code-section-4)="} {"chunk_id": "satellite_satellite-land-cover_trend-assessment_q02__6946030f9f03", "report_id": "satellite_satellite-land-cover_trend-assessment_q02", "dataset_id": "satellite-land-cover", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q02", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 4.Trend Assessment", "title": "Satellite Land Cover trend assessment for Spatial Planning and Land Management", "chunk_index": 21, "token_count": 624, "text_raw": "The next and final step is to determine whether the trend should be computed for the entire period (total trend) or divided into segments based on detected breakpoints.\n\nThe decision between using the total trend or segmented trends is based on Sen’s Slope (Trend Magnitude) and Mann-Kendall p-values (Trend Significance). Sen’s Slope is a non-parametric estimator that calculates the median rate of change over time, making it robust to outliers and suitable for detecting monotonic trends. First, Sen’s Slope is computed for the total trend, capturing the overall rate of urban change. Then, Sen’s Slope is calculated for each segmented trend to assess whether breakpoints introduce significant shifts in trend magnitude. The Mann-Kendall test is a non-parametric statistical test used to assess the presence of a monotonic upward or downward trend in a time series without requiring the data to follow any particular distribution. In parallel, the Mann-Kendall p-value is evaluated for both the total and segmented trends to measure trend significance - lower p-values indicate stronger evidence of a significant trend.\n\nThe final decision is based on the following approach:\n\n- If segmented trends show substantially different Sen’s Slopes compared to the total trend and exhibit stronger statistical significance (lower p-values), segmentation is preferred.\n- If segmented trends closely resemble the total trend or do not provide a clear statistical advantage, the total trend is used to maintain a simpler and more robust interpretation.\n\nPrepare data for trend calculation\nKeep only valid rows\n\nCalculate trends\nCompute total + segmented trends\nTotal Sen slope\nConvert to % area per decade, as requested by reviewer\nMann-Kendall\nBreakpoints and segmented trends\n\nThe following summary table reports:\n\n- **Growth rate (%/decade)** → long-term rate of urban expansion \n- **Significance** → statistical robustness of the full trend \n - `***` → highly significant (p < 0.001) \n - `**` → significant (p < 0.01) \n - `*` → moderately significant (p < 0.05) \n - no symbol → not statistically significant \n- **Breakpoints** → number of structural changes detected \n- **Segmented trend needed?** → whether segmented trends change the interpretation \n- **Confidence** → degree of agreement between full and segmented trend significance\n\nBuild summary table\nUrban area for each region in the latest year\nSort by growth rate\nRound for display\nNormalize text values\nHTML table display\n\n```text\n21 regions were assessed. Segmented trends are not required in 21 of 21 regions. Confidence is High in 21 regions, Medium in 0, and Low in 0.\n```", "text_with_prefix": "EQC Quality Assessment: \"Satellite Land Cover trend assessment for Spatial Planning and Land Management\"\nDataset: satellite-land-cover [CDS]\nAspect: trend-assessment_q02 | Category: Satellite_ECVs\nSection: Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 4.Trend Assessment\n---\nThe next and final step is to determine whether the trend should be computed for the entire period (total trend) or divided into segments based on detected breakpoints.\n\nThe decision between using the total trend or segmented trends is based on Sen’s Slope (Trend Magnitude) and Mann-Kendall p-values (Trend Significance). Sen’s Slope is a non-parametric estimator that calculates the median rate of change over time, making it robust to outliers and suitable for detecting monotonic trends. First, Sen’s Slope is computed for the total trend, capturing the overall rate of urban change. Then, Sen’s Slope is calculated for each segmented trend to assess whether breakpoints introduce significant shifts in trend magnitude. The Mann-Kendall test is a non-parametric statistical test used to assess the presence of a monotonic upward or downward trend in a time series without requiring the data to follow any particular distribution. In parallel, the Mann-Kendall p-value is evaluated for both the total and segmented trends to measure trend significance - lower p-values indicate stronger evidence of a significant trend.\n\nThe final decision is based on the following approach:\n\n- If segmented trends show substantially different Sen’s Slopes compared to the total trend and exhibit stronger statistical significance (lower p-values), segmentation is preferred.\n- If segmented trends closely resemble the total trend or do not provide a clear statistical advantage, the total trend is used to maintain a simpler and more robust interpretation.\n\nPrepare data for trend calculation\nKeep only valid rows\n\nCalculate trends\nCompute total + segmented trends\nTotal Sen slope\nConvert to % area per decade, as requested by reviewer\nMann-Kendall\nBreakpoints and segmented trends\n\nThe following summary table reports:\n\n- **Growth rate (%/decade)** → long-term rate of urban expansion \n- **Significance** → statistical robustness of the full trend \n - `***` → highly significant (p < 0.001) \n - `**` → significant (p < 0.01) \n - `*` → moderately significant (p < 0.05) \n - no symbol → not statistically significant \n- **Breakpoints** → number of structural changes detected \n- **Segmented trend needed?** → whether segmented trends change the interpretation \n- **Confidence** → degree of agreement between full and segmented trend significance\n\nBuild summary table\nUrban area for each region in the latest year\nSort by growth rate\nRound for display\nNormalize text values\nHTML table display\n\n```text\n21 regions were assessed. Segmented trends are not required in 21 of 21 regions. Confidence is High in 21 regions, Medium in 0, and Low in 0.\n```"} {"chunk_id": "satellite_satellite-land-cover_trend-assessment_q02__acad1d20c1f0", "report_id": "satellite_satellite-land-cover_trend-assessment_q02", "dataset_id": "satellite-land-cover", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q02", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 4.Trend Assessment > Trend Analysis", "title": "Satellite Land Cover trend assessment for Spatial Planning and Land Management", "chunk_index": 22, "token_count": 351, "text_raw": "The results demonstrates that, although breakpoints were identified, they do not substantially alter the underlying trend estimates. In all cases, the total trend was preferred, as segmented alternatives did not exhibit significantly stronger slopes or lower p-values.\n\nThis suggests that the dataset provides a coherent temporal structure for monitoring urban land cover change. While localised variations exist, the broader trend of urban expansion remains largely continuous and statistically stable.\n\nFurthermore, the reliability of the total trend is evident even in regions with marked breakpoint clusters, such as Galicia, Andalucía, and Región de Murcia. Despite temporary changes in slope magnitude, the segmented trends did not offer a statistically significant improvement over the total series, suggesting that short-term accelerations or decelerations are well-integrated into the full-period estimate.\n\nGeographical characteristics modulate this effect. Larger metropolitan areas or regions with long-standing urban infrastructure exhibit smoother temporal evolution, while smaller or emerging urban zones may reflect sharper shifts in land cover classification. Nevertheless, these fluctuations did not systematically translate into divergent trend recommendations.\n\nIn summary, the analysis confirms that the dataset is suitable for deriving consistent urbanisation trends across multi-decadal periods. Segmentation remains a valuable diagnostic tool for identifying local deviations, but for most spatial planning and land management applications, the total trend offers a statistically supported and methodologically stable representation of urban growth dynamics.\n\n(code-section-5)=", "text_with_prefix": "EQC Quality Assessment: \"Satellite Land Cover trend assessment for Spatial Planning and Land Management\"\nDataset: satellite-land-cover [CDS]\nAspect: trend-assessment_q02 | Category: Satellite_ECVs\nSection: Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 4.Trend Assessment > Trend Analysis\n---\nThe results demonstrates that, although breakpoints were identified, they do not substantially alter the underlying trend estimates. In all cases, the total trend was preferred, as segmented alternatives did not exhibit significantly stronger slopes or lower p-values.\n\nThis suggests that the dataset provides a coherent temporal structure for monitoring urban land cover change. While localised variations exist, the broader trend of urban expansion remains largely continuous and statistically stable.\n\nFurthermore, the reliability of the total trend is evident even in regions with marked breakpoint clusters, such as Galicia, Andalucía, and Región de Murcia. Despite temporary changes in slope magnitude, the segmented trends did not offer a statistically significant improvement over the total series, suggesting that short-term accelerations or decelerations are well-integrated into the full-period estimate.\n\nGeographical characteristics modulate this effect. Larger metropolitan areas or regions with long-standing urban infrastructure exhibit smoother temporal evolution, while smaller or emerging urban zones may reflect sharper shifts in land cover classification. Nevertheless, these fluctuations did not systematically translate into divergent trend recommendations.\n\nIn summary, the analysis confirms that the dataset is suitable for deriving consistent urbanisation trends across multi-decadal periods. Segmentation remains a valuable diagnostic tool for identifying local deviations, but for most spatial planning and land management applications, the total trend offers a statistically supported and methodologically stable representation of urban growth dynamics.\n\n(code-section-5)="} {"chunk_id": "satellite_satellite-land-cover_trend-assessment_q02__ecec47b07f49", "report_id": "satellite_satellite-land-cover_trend-assessment_q02", "dataset_id": "satellite-land-cover", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q02", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 5. Discussion", "title": "Satellite Land Cover trend assessment for Spatial Planning and Land Management", "chunk_index": 23, "token_count": 432, "text_raw": "The results of this analysis demonstrate the suitability of the dataset for assessing long-term urbanisation trends. Despite the presence of breakpoints the overall consistency of the dataset supports its use for reliable trend estimation in urban land cover.\n\nFor spatial planning and land management applications, this temporal continuity is critical. The ability to trace urban expansion over a multi-decadal period enables stakeholders to evaluate the spatial footprint of urban growth, assess policy impacts, and inform future development strategies. The dataset captures the general trajectory of urbanisation effectively, even in the presence of short-term fluctuations or classification refinements.\n\nNonetheless, the analysis highlights the importance of methodological awareness. Some of the breakpoints occurre around years with known changes in the processing chain. These may introduce artificial discontinuities in urban extent, particularly in regions undergoing rapid development or where classification thresholds are more sensitive to spectral or spatial resolution. Such effects must be considered when interpreting apparent changes in urban dynamics.\n\nGeographic and demographic context also influence data interpretation. Larger metropolitan regions, such as Madrid or Lisbon, tend to exhibit smoother trends, as broader patterns of growth absorb minor classification inconsistencies. In contrast, smaller or fast-developing regions may show more pronounced breaks, either due to real shifts in urbanisation or amplified sensitivity to methodological updates.\n\nFrom a practical standpoint, the dataset supports robust urban trend detection, but effective application requires a cautious and informed approach. Users should:\n\n- Consider the potential influence of sensor transitions and processing updates on detected breakpoints\n- Prioritise total trends for general assessments, while using segmentation to highlight context-specific deviations\n- Interpret statistical outputs alongside relevant contextual knowledge, such as infrastructure investments, economic cycles, or policy milestones.\n\nIn conclusion, the dataset provides a strong foundation for monitoring urbanisation trends. Its temporal stability and spatial resolution are appropriate for regional-scale planning.", "text_with_prefix": "EQC Quality Assessment: \"Satellite Land Cover trend assessment for Spatial Planning and Land Management\"\nDataset: satellite-land-cover [CDS]\nAspect: trend-assessment_q02 | Category: Satellite_ECVs\nSection: Satellite Land Cover trend assessment for Spatial Planning and Land Management > Analysis and results > 5. Discussion\n---\nThe results of this analysis demonstrate the suitability of the dataset for assessing long-term urbanisation trends. Despite the presence of breakpoints the overall consistency of the dataset supports its use for reliable trend estimation in urban land cover.\n\nFor spatial planning and land management applications, this temporal continuity is critical. The ability to trace urban expansion over a multi-decadal period enables stakeholders to evaluate the spatial footprint of urban growth, assess policy impacts, and inform future development strategies. The dataset captures the general trajectory of urbanisation effectively, even in the presence of short-term fluctuations or classification refinements.\n\nNonetheless, the analysis highlights the importance of methodological awareness. Some of the breakpoints occurre around years with known changes in the processing chain. These may introduce artificial discontinuities in urban extent, particularly in regions undergoing rapid development or where classification thresholds are more sensitive to spectral or spatial resolution. Such effects must be considered when interpreting apparent changes in urban dynamics.\n\nGeographic and demographic context also influence data interpretation. Larger metropolitan regions, such as Madrid or Lisbon, tend to exhibit smoother trends, as broader patterns of growth absorb minor classification inconsistencies. In contrast, smaller or fast-developing regions may show more pronounced breaks, either due to real shifts in urbanisation or amplified sensitivity to methodological updates.\n\nFrom a practical standpoint, the dataset supports robust urban trend detection, but effective application requires a cautious and informed approach. Users should:\n\n- Consider the potential influence of sensor transitions and processing updates on detected breakpoints\n- Prioritise total trends for general assessments, while using segmentation to highlight context-specific deviations\n- Interpret statistical outputs alongside relevant contextual knowledge, such as infrastructure investments, economic cycles, or policy milestones.\n\nIn conclusion, the dataset provides a strong foundation for monitoring urbanisation trends. Its temporal stability and spatial resolution are appropriate for regional-scale planning."} {"chunk_id": "satellite_satellite-land-cover_trend-assessment_q02__748eeae3b912", "report_id": "satellite_satellite-land-cover_trend-assessment_q02", "dataset_id": "satellite-land-cover", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q02", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Land Cover trend assessment for Spatial Planning and Land Management > ℹ️ If you want to know more > Key Resources", "title": "Satellite Land Cover trend assessment for Spatial Planning and Land Management", "chunk_index": 24, "token_count": 346, "text_raw": "* The CDS catalogue entry for the data used was [Land cover classification gridded maps from 1992 to present derived from satellite observations](https://cds.climate.copernicus.eu/datasets/satellite-land-cover?tab=overview)\n\n* Product User Guide and Specification of the dataset [version 2.1](https://dast.copernicus-climate.eu/documents/satellite-land-cover/D5.3.1_PUGS_ICDR_LC_v2.1.x_PRODUCTS_v1.1.pdf) and [version 2.0](https://dast.copernicus-climate.eu/documents/satellite-land-cover/D3.3.11-v1.0_PUGS_CDR_LC-CCI_v2.0.7cds_Products_v1.0.1_APPROVED_Ver1.pdf)\n\n* [Eurostat NUTS](https://ec.europa.eu/eurostat/web/nuts) (Nomenclature of territorial units for statistics)\n\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), c3s_eqc_automatic_quality_control, prepared by [B-Open](https://www.bopen.eu/)", "text_with_prefix": "EQC Quality Assessment: \"Satellite Land Cover trend assessment for Spatial Planning and Land Management\"\nDataset: satellite-land-cover [CDS]\nAspect: trend-assessment_q02 | Category: Satellite_ECVs\nSection: Satellite Land Cover trend assessment for Spatial Planning and Land Management > ℹ️ If you want to know more > Key Resources\n---\n* The CDS catalogue entry for the data used was [Land cover classification gridded maps from 1992 to present derived from satellite observations](https://cds.climate.copernicus.eu/datasets/satellite-land-cover?tab=overview)\n\n* Product User Guide and Specification of the dataset [version 2.1](https://dast.copernicus-climate.eu/documents/satellite-land-cover/D5.3.1_PUGS_ICDR_LC_v2.1.x_PRODUCTS_v1.1.pdf) and [version 2.0](https://dast.copernicus-climate.eu/documents/satellite-land-cover/D3.3.11-v1.0_PUGS_CDR_LC-CCI_v2.0.7cds_Products_v1.0.1_APPROVED_Ver1.pdf)\n\n* [Eurostat NUTS](https://ec.europa.eu/eurostat/web/nuts) (Nomenclature of territorial units for statistics)\n\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), c3s_eqc_automatic_quality_control, prepared by [B-Open](https://www.bopen.eu/)"} {"chunk_id": "satellite_satellite-land-cover_trend-assessment_q02__4d515be1bdfb", "report_id": "satellite_satellite-land-cover_trend-assessment_q02", "dataset_id": "satellite-land-cover", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q02", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Land Cover trend assessment for Spatial Planning and Land Management > ℹ️ If you want to know more > References", "title": "Satellite Land Cover trend assessment for Spatial Planning and Land Management", "chunk_index": 25, "token_count": 993, "text_raw": "Vargo, J., Habeeb, D. M., & Stone, B. (2013). The importance of land cover change across urban-rural typologies for climate modeling. Journal of Environmental Management, 114, 243-252. https://doi.org/10.1016/j.jenvman.2012.10.007\n\nChang, Y., Hou, K., Li, X., Zhang, Y., & Chen, P. (2018). Review of land use and land cover change research progress. IOP Conference Series: Earth and Environmental Science, 113, 012087. https://doi.org/10.1088/1755-1315/113/1/012087\n\nChelali, A., Benediktsson, J. A., Chanussot, J., & Akbari, V. (2019). A review of artificial intelligence techniques for land cover change detection using remote sensing images. IEEE Transactions on Geoscience and Remote Sensing, 57(10), 7649–7668. https://doi.org/10.1109/JURSE.2019.8808967\n\nDelgado, L., & Romero, J. (2021). Evaluating the impact of Madrid Central on urban mobility: A step toward sustainable cities? Cities, 113, 103120. https://doi.org/10.1016/j.cities.2021.103120\n\nGonzález, M. J., & Leal, J. (2018). Urban growth and real estate development in the Iberian Peninsula: The boom of 1996–2007. Land Use Policy, 76, 648–657. https://doi.org/10.1016/j.landusepol.2018.05.023\n\nGutiérrez, F., Rodríguez, V., & Prieto, C. (2016). Aging population and urbanization: Implications for Spain’s demographic landscape. Journal of Rural Studies, 45, 85–98. https://doi.org/10.1016/j.jrurstud.2016.03.008\n\nLópez, F. A., Martín, J. C., & Albalate, D. (2017). High-speed rail and urban expansion in the Iberian Peninsula: The cases of AVE and Alfa Pendular. Journal of Transport Geography, 60, 57–68. https://doi.org/10.1016/j.jtrangeo.2017.02.005\n\nMartínez, P., & García, R. (2015). The impact of the 2008 financial crisis on urban expansion in southern Spain. Habitat International, 47, 42–50. https://doi.org/10.1016/j.habitatint.2015.01.010\n\nPalmero-Iniesta, H., Fernández, J. A., & Olcina, J. (2021). Land-use changes and natural land recovery after the 2008 crisis: Evidence from Spain. Environmental Science & Policy, 116, 72–83. https://api.semanticscholar.org/CorpusID:238829360\n\nRullan, O. (2014). The impact of tourism-driven urbanisation in Spain and Portugal: Coastal transformations and land use. Land Use Policy, 38, 139–150. https://doi.org/10.1016/j.landusepol.2014.02.009\n\nSánchez, M., Pérez, L., & Torres, D. (2019). Rural depopulation and urban migration trends in Spain and Portugal. Geoforum, 104, 78–91. https://doi.org/10.1016/j.geoforum.2019.06.015\n\nSilva, E., & Fernandes, J. (2019). The influence of urban planning policies on land prices and informal settlements in Iberian cities. Land Use Policy, 89, 104200. https://doi.org/10.1016/j.landusepol.2019.04.012\n\nSilva, V., Santos, R., & Moreira, J. (2017). The expansion of peri-urban areas in Lisbon and Porto: Drivers and consequences. Cities, 66, 101–112. https://doi.org/10.1016/j.cities.2017.04.002", "text_with_prefix": "EQC Quality Assessment: \"Satellite Land Cover trend assessment for Spatial Planning and Land Management\"\nDataset: satellite-land-cover [CDS]\nAspect: trend-assessment_q02 | Category: Satellite_ECVs\nSection: Satellite Land Cover trend assessment for Spatial Planning and Land Management > ℹ️ If you want to know more > References\n---\nVargo, J., Habeeb, D. M., & Stone, B. (2013). The importance of land cover change across urban-rural typologies for climate modeling. Journal of Environmental Management, 114, 243-252. https://doi.org/10.1016/j.jenvman.2012.10.007\n\nChang, Y., Hou, K., Li, X., Zhang, Y., & Chen, P. (2018). Review of land use and land cover change research progress. IOP Conference Series: Earth and Environmental Science, 113, 012087. https://doi.org/10.1088/1755-1315/113/1/012087\n\nChelali, A., Benediktsson, J. A., Chanussot, J., & Akbari, V. (2019). A review of artificial intelligence techniques for land cover change detection using remote sensing images. IEEE Transactions on Geoscience and Remote Sensing, 57(10), 7649–7668. https://doi.org/10.1109/JURSE.2019.8808967\n\nDelgado, L., & Romero, J. (2021). Evaluating the impact of Madrid Central on urban mobility: A step toward sustainable cities? Cities, 113, 103120. https://doi.org/10.1016/j.cities.2021.103120\n\nGonzález, M. J., & Leal, J. (2018). Urban growth and real estate development in the Iberian Peninsula: The boom of 1996–2007. Land Use Policy, 76, 648–657. https://doi.org/10.1016/j.landusepol.2018.05.023\n\nGutiérrez, F., Rodríguez, V., & Prieto, C. (2016). Aging population and urbanization: Implications for Spain’s demographic landscape. Journal of Rural Studies, 45, 85–98. https://doi.org/10.1016/j.jrurstud.2016.03.008\n\nLópez, F. A., Martín, J. C., & Albalate, D. (2017). High-speed rail and urban expansion in the Iberian Peninsula: The cases of AVE and Alfa Pendular. Journal of Transport Geography, 60, 57–68. https://doi.org/10.1016/j.jtrangeo.2017.02.005\n\nMartínez, P., & García, R. (2015). The impact of the 2008 financial crisis on urban expansion in southern Spain. Habitat International, 47, 42–50. https://doi.org/10.1016/j.habitatint.2015.01.010\n\nPalmero-Iniesta, H., Fernández, J. A., & Olcina, J. (2021). Land-use changes and natural land recovery after the 2008 crisis: Evidence from Spain. Environmental Science & Policy, 116, 72–83. https://api.semanticscholar.org/CorpusID:238829360\n\nRullan, O. (2014). The impact of tourism-driven urbanisation in Spain and Portugal: Coastal transformations and land use. Land Use Policy, 38, 139–150. https://doi.org/10.1016/j.landusepol.2014.02.009\n\nSánchez, M., Pérez, L., & Torres, D. (2019). Rural depopulation and urban migration trends in Spain and Portugal. Geoforum, 104, 78–91. https://doi.org/10.1016/j.geoforum.2019.06.015\n\nSilva, E., & Fernandes, J. (2019). The influence of urban planning policies on land prices and informal settlements in Iberian cities. Land Use Policy, 89, 104200. https://doi.org/10.1016/j.landusepol.2019.04.012\n\nSilva, V., Santos, R., & Moreira, J. (2017). The expansion of peri-urban areas in Lisbon and Porto: Drivers and consequences. Cities, 66, 101–112. https://doi.org/10.1016/j.cities.2017.04.002"} {"chunk_id": "satellite_satellite-lake-water-level_climate-and-weather-extremes_q01__8a048fe54437", "report_id": "satellite_satellite-lake-water-level_climate-and-weather-extremes_q01", "dataset_id": "satellite-lake-water-level", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q01", "aspect_base": "climate-and-weather-extremes", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Lake Victoria’s 2020 Flood Event Analysis > Quality assessment question", "title": "Lake Victoria’s 2020 Flood Event Analysis", "chunk_index": 0, "token_count": 319, "text_raw": "* **Can the C3S Lake Water Levels dataset be used to conduct extreme value analyses to detect and quantify flood events?**\n\nIn 2020, Lake Victoria experienced an extreme flood event with significant regional impacts. Accurate definition and characterization of such events are essential for understanding their causes and informing adaptation strategies. This notebook explores whether the [Lake water levels from 1992 to present derived from satellite observations](https://cds.climate.copernicus.eu/datasets/satellite-lake-water-level?tab=overview) (C3S-LWL v5.0) dataset, a freely available and harmonized satellite-derived product, is suitable for defining high-impact flood events through statistical analysis.\n\nInspired by methodologies such as those applied by Pietroiusti et al. (2024) [[1]](https://esd.copernicus.org/articles/15/225/2024/) using the DAHITI dataset, this analysis applies extreme value analysis (EVA) to assess the 2020 flood event. The goal is to evaluate whether C3S-LWL v5.0 can serve as a reliable resource for event definition in impact attribution workflows, particularly in data-scarce regions like the Lake Victoria basin.", "text_with_prefix": "EQC Quality Assessment: \"Lake Victoria’s 2020 Flood Event Analysis\"\nDataset: satellite-lake-water-level [CDS]\nAspect: climate-and-weather-extremes_q01 | Category: Satellite_ECVs\nSection: Lake Victoria’s 2020 Flood Event Analysis > Quality assessment question\n---\n* **Can the C3S Lake Water Levels dataset be used to conduct extreme value analyses to detect and quantify flood events?**\n\nIn 2020, Lake Victoria experienced an extreme flood event with significant regional impacts. Accurate definition and characterization of such events are essential for understanding their causes and informing adaptation strategies. This notebook explores whether the [Lake water levels from 1992 to present derived from satellite observations](https://cds.climate.copernicus.eu/datasets/satellite-lake-water-level?tab=overview) (C3S-LWL v5.0) dataset, a freely available and harmonized satellite-derived product, is suitable for defining high-impact flood events through statistical analysis.\n\nInspired by methodologies such as those applied by Pietroiusti et al. (2024) [[1]](https://esd.copernicus.org/articles/15/225/2024/) using the DAHITI dataset, this analysis applies extreme value analysis (EVA) to assess the 2020 flood event. The goal is to evaluate whether C3S-LWL v5.0 can serve as a reliable resource for event definition in impact attribution workflows, particularly in data-scarce regions like the Lake Victoria basin."} {"chunk_id": "satellite_satellite-lake-water-level_climate-and-weather-extremes_q01__9402cf40388a", "report_id": "satellite_satellite-lake-water-level_climate-and-weather-extremes_q01", "dataset_id": "satellite-lake-water-level", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q01", "aspect_base": "climate-and-weather-extremes", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Lake Victoria’s 2020 Flood Event Analysis > Quality assessment statement", "title": "Lake Victoria’s 2020 Flood Event Analysis", "chunk_index": 1, "token_count": 207, "text_raw": "These are the key outcomes of this assessment\n\n* The assessment indicates the C3S Lake Water Levels dataset is suitable to detect and characterise flood events, identifying the 2020 Lake Victoria flood event and ranking it as the third-highest 180-day lake level rise in the historical record (1948–2023), with only 1962 and 1998 recording greater increase.\n\n* Differences between DAHITI and C3S-LWL v5.0 are more pronounced in the early record (1992–2002), due to limited satellite coverage and distinct processing approaches, but these differences shrink substantially after 2002. In recent years, the two datasets show very close agreement.\n```", "text_with_prefix": "EQC Quality Assessment: \"Lake Victoria’s 2020 Flood Event Analysis\"\nDataset: satellite-lake-water-level [CDS]\nAspect: climate-and-weather-extremes_q01 | Category: Satellite_ECVs\nSection: Lake Victoria’s 2020 Flood Event Analysis > Quality assessment statement\n---\nThese are the key outcomes of this assessment\n\n* The assessment indicates the C3S Lake Water Levels dataset is suitable to detect and characterise flood events, identifying the 2020 Lake Victoria flood event and ranking it as the third-highest 180-day lake level rise in the historical record (1948–2023), with only 1962 and 1998 recording greater increase.\n\n* Differences between DAHITI and C3S-LWL v5.0 are more pronounced in the early record (1992–2002), due to limited satellite coverage and distinct processing approaches, but these differences shrink substantially after 2002. In recent years, the two datasets show very close agreement.\n```"} {"chunk_id": "satellite_satellite-lake-water-level_climate-and-weather-extremes_q01__6d6fccb67098", "report_id": "satellite_satellite-lake-water-level_climate-and-weather-extremes_q01", "dataset_id": "satellite-lake-water-level", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q01", "aspect_base": "climate-and-weather-extremes", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Lake Victoria’s 2020 Flood Event Analysis > Methodology", "title": "Lake Victoria’s 2020 Flood Event Analysis", "chunk_index": 2, "token_count": 591, "text_raw": "First, the C3S LWL and the DAHITI datasets of Lake Victoria’s water levels were compared. Both datasets are based on satellite altimetry but differ in their processing algorithms. We evaluated the two datasets in terms of temporal completeness and analysed the causes of their discrepancies. This comparison was motivated by the fact that DAHITI was used in Pietroiusti et al. (2024) [[1]](https://esd.copernicus.org/articles/15/225/2024/) for extreme event attribution (EEA) of the 2020 Lake Victoria flood. Our goal was to assess whether the C3S LWL dataset could serve as a suitable alternative to DAHITI for this type of application. To extend the record, a reconstructed lake level dataset was also built using HYDROMET data, which covers a period prior to satellite-based products. Finally, since lake level data is primarily required during the event definition step of the EEA process, we focused on this first step and tested its implementation using the C3S LWL dataset. The same procedure as on Pietroiusti et al. (2024) [[1]](https://esd.copernicus.org/articles/15/225/2024/) was followed for the event definition, which involved the calculation and ranking of the annual block maximum of the chosen variable, which was the rate of change in water levels over a time window.\n\nThe analysis and results are organised in the following steps, which are detailed in the sections below:\n\n**[](section-1)**\n * Download C3S–LWL v5.0 satellite-lake-water-level data for Lake Victoria.\n\n**[](section-2)**\n * Load and preprocess DAHITI data.\n * Plot monthly count of data recorded for both the DAHITI and C3S data.\n * Plot C3S and DAHITI’s Water Level and show the maximum difference.\n\n**[](section-3)**\n * Load and preprocess HYDROMET data.\n * Calculate the average difference between HYDROMET and C3S-LWL overlapping data and correct the HYDROMET dataset based on this.\n * Interpolate to fill missing gaps in HYDROMET.\n * Plot reconstructed time series.\n\n**[](section-4)**\n * Apply the annual block maxima to the reconstructed time series.\n * Compare with the results from Pietroiusti et al. (2024) [[1]](https://esd.copernicus.org/articles/15/225/2024/).", "text_with_prefix": "EQC Quality Assessment: \"Lake Victoria’s 2020 Flood Event Analysis\"\nDataset: satellite-lake-water-level [CDS]\nAspect: climate-and-weather-extremes_q01 | Category: Satellite_ECVs\nSection: Lake Victoria’s 2020 Flood Event Analysis > Methodology\n---\nFirst, the C3S LWL and the DAHITI datasets of Lake Victoria’s water levels were compared. Both datasets are based on satellite altimetry but differ in their processing algorithms. We evaluated the two datasets in terms of temporal completeness and analysed the causes of their discrepancies. This comparison was motivated by the fact that DAHITI was used in Pietroiusti et al. (2024) [[1]](https://esd.copernicus.org/articles/15/225/2024/) for extreme event attribution (EEA) of the 2020 Lake Victoria flood. Our goal was to assess whether the C3S LWL dataset could serve as a suitable alternative to DAHITI for this type of application. To extend the record, a reconstructed lake level dataset was also built using HYDROMET data, which covers a period prior to satellite-based products. Finally, since lake level data is primarily required during the event definition step of the EEA process, we focused on this first step and tested its implementation using the C3S LWL dataset. The same procedure as on Pietroiusti et al. (2024) [[1]](https://esd.copernicus.org/articles/15/225/2024/) was followed for the event definition, which involved the calculation and ranking of the annual block maximum of the chosen variable, which was the rate of change in water levels over a time window.\n\nThe analysis and results are organised in the following steps, which are detailed in the sections below:\n\n**[](section-1)**\n * Download C3S–LWL v5.0 satellite-lake-water-level data for Lake Victoria.\n\n**[](section-2)**\n * Load and preprocess DAHITI data.\n * Plot monthly count of data recorded for both the DAHITI and C3S data.\n * Plot C3S and DAHITI’s Water Level and show the maximum difference.\n\n**[](section-3)**\n * Load and preprocess HYDROMET data.\n * Calculate the average difference between HYDROMET and C3S-LWL overlapping data and correct the HYDROMET dataset based on this.\n * Interpolate to fill missing gaps in HYDROMET.\n * Plot reconstructed time series.\n\n**[](section-4)**\n * Apply the annual block maxima to the reconstructed time series.\n * Compare with the results from Pietroiusti et al. (2024) [[1]](https://esd.copernicus.org/articles/15/225/2024/)."} {"chunk_id": "satellite_satellite-lake-water-level_climate-and-weather-extremes_q01__21cb959b9277", "report_id": "satellite_satellite-lake-water-level_climate-and-weather-extremes_q01", "dataset_id": "satellite-lake-water-level", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q01", "aspect_base": "climate-and-weather-extremes", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Lake Victoria’s 2020 Flood Event Analysis > Analysis and results > 1. Data request and download > Download data", "title": "Lake Victoria’s 2020 Flood Event Analysis", "chunk_index": 3, "token_count": 112, "text_raw": "```text\n100%|██████████| 1/1 [00:00<00:00, 20.35it/s]\n```\n\n(section-2)=", "text_with_prefix": "EQC Quality Assessment: \"Lake Victoria’s 2020 Flood Event Analysis\"\nDataset: satellite-lake-water-level [CDS]\nAspect: climate-and-weather-extremes_q01 | Category: Satellite_ECVs\nSection: Lake Victoria’s 2020 Flood Event Analysis > Analysis and results > 1. Data request and download > Download data\n---\n```text\n100%|██████████| 1/1 [00:00<00:00, 20.35it/s]\n```\n\n(section-2)="} {"chunk_id": "satellite_satellite-lake-water-level_climate-and-weather-extremes_q01__3bccc161ac8d", "report_id": "satellite_satellite-lake-water-level_climate-and-weather-extremes_q01", "dataset_id": "satellite-lake-water-level", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q01", "aspect_base": "climate-and-weather-extremes", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Lake Victoria’s 2020 Flood Event Analysis > Analysis and results > 2. Comparison of C3S LWL with the DAHITI dataset > Load and preprocess DAHITI data", "title": "Lake Victoria’s 2020 Flood Event Analysis", "chunk_index": 4, "token_count": 629, "text_raw": "This assessment applies the methods developed by Pietroiusti et al. (2024) [[1]](https://esd.copernicus.org/articles/15/225/2024/), whose data are openly available via [Zenodo](https://zenodo.org/records/10793917). In that study, the DAHITI dataset [[2]](https://hess.copernicus.org/articles/19/4345/2015/hess-19-4345-2015.html) was employed. Accordingly, a comparison of Lake Victoria’s water level time series from C3S-LWL v5.0 and DAHITI will be carried out. The DAHITI time series for Lake Victoria can be accessed and downloaded from the [DAHITI portal](https://dahiti.dgfi.tum.de/en/2/water-level-altimetry/).\n\nFor reproducibility, the specific DAHITI dataset used in this assessment is located in the [Zenodo](https://zenodo.org/records/10793917) repository at: lakevic-eea-data-zenodo\\lakevic-eea-data-zenodo\\lakevic-eea-analysis\\lakelevels\\DAHITI_lakelevels_070322.csv.\n\nConvert date column to datetime and set as index\nCreate scaled integer water level\nKeep original as meters\nError as integer\nReorder columns\n\n```text\nwater_level error water_level_m\ndatetime \n1992-09-27 1135000 0 1135.000\n1992-10-07 1135023 0 1135.023\n1992-10-17 1135032 0 1135.032\n1992-10-27 1134958 0 1134.958\n1992-11-06 1135035 0 1135.035\n... ... ... ...\n2022-01-22 1136143 0 1136.143\n2022-02-01 1136145 0 1136.145\n2022-02-11 1136199 0 1136.199\n2022-02-21 1136227 0 1136.227\n2022-03-03 1136229 0 1136.229\n\n[1058 rows x 3 columns]\n```", "text_with_prefix": "EQC Quality Assessment: \"Lake Victoria’s 2020 Flood Event Analysis\"\nDataset: satellite-lake-water-level [CDS]\nAspect: climate-and-weather-extremes_q01 | Category: Satellite_ECVs\nSection: Lake Victoria’s 2020 Flood Event Analysis > Analysis and results > 2. Comparison of C3S LWL with the DAHITI dataset > Load and preprocess DAHITI data\n---\nThis assessment applies the methods developed by Pietroiusti et al. (2024) [[1]](https://esd.copernicus.org/articles/15/225/2024/), whose data are openly available via [Zenodo](https://zenodo.org/records/10793917). In that study, the DAHITI dataset [[2]](https://hess.copernicus.org/articles/19/4345/2015/hess-19-4345-2015.html) was employed. Accordingly, a comparison of Lake Victoria’s water level time series from C3S-LWL v5.0 and DAHITI will be carried out. The DAHITI time series for Lake Victoria can be accessed and downloaded from the [DAHITI portal](https://dahiti.dgfi.tum.de/en/2/water-level-altimetry/).\n\nFor reproducibility, the specific DAHITI dataset used in this assessment is located in the [Zenodo](https://zenodo.org/records/10793917) repository at: lakevic-eea-data-zenodo\\lakevic-eea-data-zenodo\\lakevic-eea-analysis\\lakelevels\\DAHITI_lakelevels_070322.csv.\n\nConvert date column to datetime and set as index\nCreate scaled integer water level\nKeep original as meters\nError as integer\nReorder columns\n\n```text\nwater_level error water_level_m\ndatetime \n1992-09-27 1135000 0 1135.000\n1992-10-07 1135023 0 1135.023\n1992-10-17 1135032 0 1135.032\n1992-10-27 1134958 0 1134.958\n1992-11-06 1135035 0 1135.035\n... ... ... ...\n2022-01-22 1136143 0 1136.143\n2022-02-01 1136145 0 1136.145\n2022-02-11 1136199 0 1136.199\n2022-02-21 1136227 0 1136.227\n2022-03-03 1136229 0 1136.229\n\n[1058 rows x 3 columns]\n```"} {"chunk_id": "satellite_satellite-lake-water-level_climate-and-weather-extremes_q01__36aca75e1c85", "report_id": "satellite_satellite-lake-water-level_climate-and-weather-extremes_q01", "dataset_id": "satellite-lake-water-level", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q01", "aspect_base": "climate-and-weather-extremes", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Lake Victoria’s 2020 Flood Event Analysis > Analysis and results > 2. Comparison of C3S LWL with the DAHITI dataset > Monthly temporal completeness", "title": "Lake Victoria’s 2020 Flood Event Analysis", "chunk_index": 5, "token_count": 579, "text_raw": "First dataset: xarray object\nSecond dataset: Pandas DataFrame\nCreate subplots with 1 row and 2 columns\nPlot first bar chart\nPlot second bar chart\nAdjust layout\n\n*Figure 1. Comparison of count of monthly values over time available in the C3S-LWL v5.0 dataset (on the left) and in DAHITI (on the right).*\n\nThe DAHITI dataset exhibits a relatively consistent number of monthly observations, typically ranging from 1 to 4 throughout its entire temporal span. In contrast, the C3S-LWL v5.0 dataset shows variability in the number of monthly values, depending on the availability of satellite missions over time.\nUp until the end of 2010, only one value per month was recorded. From 2011 to 2015, this number increased to mostly three observations per month, likely reflecting additional unfiltered data from the Jason-2 mission. A significant increase in observation frequency is evident from 2016 onward, corresponding to the introduction of Jason-3 and Sentinel-3A satellites. The C3S-LWL v5.0 dataset contains only one gap, in 2006, indicating strong overall temporal completeness.\n\nThe differing number of monthly observations between the datasets can be explained by the fact that DAHITI applies interpolation using a Kalman filter, a statistical technique that predicts the water level at the next time step based on previous observations and associated uncertainties [[1]](https://esd.copernicus.org/articles/15/225/2024/). This allows DAHITI to maintain a nearly uniform temporal resolution, achieving approximately three observations per month even during periods with limited satellite overpasses. On the other hand, the C3S-LWL v5.0 dataset does not interpolate. The observations from the different satellite missions are processed in three sequential steps, each applying thresholds and filtering criteria to remove bad-quality data. As a result, only high-quality data are retained in the final dataset. Because older satellite sensors generally had lower accuracy, a larger fraction of early observations is filtered out, whereas later missions contribute more usable data. More in depth information on the processing steps to make the C3S-LWL v5.0 dataset can be found in the Algorithm Theoretical Basis Document [(ATBD)](https://confluence.ecmwf.int/pages/viewpage.action?pageId=425328942).", "text_with_prefix": "EQC Quality Assessment: \"Lake Victoria’s 2020 Flood Event Analysis\"\nDataset: satellite-lake-water-level [CDS]\nAspect: climate-and-weather-extremes_q01 | Category: Satellite_ECVs\nSection: Lake Victoria’s 2020 Flood Event Analysis > Analysis and results > 2. Comparison of C3S LWL with the DAHITI dataset > Monthly temporal completeness\n---\nFirst dataset: xarray object\nSecond dataset: Pandas DataFrame\nCreate subplots with 1 row and 2 columns\nPlot first bar chart\nPlot second bar chart\nAdjust layout\n\n*Figure 1. Comparison of count of monthly values over time available in the C3S-LWL v5.0 dataset (on the left) and in DAHITI (on the right).*\n\nThe DAHITI dataset exhibits a relatively consistent number of monthly observations, typically ranging from 1 to 4 throughout its entire temporal span. In contrast, the C3S-LWL v5.0 dataset shows variability in the number of monthly values, depending on the availability of satellite missions over time.\nUp until the end of 2010, only one value per month was recorded. From 2011 to 2015, this number increased to mostly three observations per month, likely reflecting additional unfiltered data from the Jason-2 mission. A significant increase in observation frequency is evident from 2016 onward, corresponding to the introduction of Jason-3 and Sentinel-3A satellites. The C3S-LWL v5.0 dataset contains only one gap, in 2006, indicating strong overall temporal completeness.\n\nThe differing number of monthly observations between the datasets can be explained by the fact that DAHITI applies interpolation using a Kalman filter, a statistical technique that predicts the water level at the next time step based on previous observations and associated uncertainties [[1]](https://esd.copernicus.org/articles/15/225/2024/). This allows DAHITI to maintain a nearly uniform temporal resolution, achieving approximately three observations per month even during periods with limited satellite overpasses. On the other hand, the C3S-LWL v5.0 dataset does not interpolate. The observations from the different satellite missions are processed in three sequential steps, each applying thresholds and filtering criteria to remove bad-quality data. As a result, only high-quality data are retained in the final dataset. Because older satellite sensors generally had lower accuracy, a larger fraction of early observations is filtered out, whereas later missions contribute more usable data. More in depth information on the processing steps to make the C3S-LWL v5.0 dataset can be found in the Algorithm Theoretical Basis Document [(ATBD)](https://confluence.ecmwf.int/pages/viewpage.action?pageId=425328942)."} {"chunk_id": "satellite_satellite-lake-water-level_climate-and-weather-extremes_q01__b8e360192cb6", "report_id": "satellite_satellite-lake-water-level_climate-and-weather-extremes_q01", "dataset_id": "satellite-lake-water-level", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q01", "aspect_base": "climate-and-weather-extremes", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Lake Victoria’s 2020 Flood Event Analysis > Analysis and results > 2. Comparison of C3S LWL with the DAHITI dataset > Plot C3S and DAHITI's Water Level", "title": "Lake Victoria’s 2020 Flood Event Analysis", "chunk_index": 6, "token_count": 718, "text_raw": "Extract time series from xarray (da)\nFinal touches\n\n*Figure 2. Lake Victoria's water level from 1992 to present from C3S-LWL v5.0 and DAHITI.*\n\nExtract time series from xarray (da)\nAlign by nearest timestamp\nDrop rows where no match was found\nCompute difference\nEnsure the datetime column is in datetime format\nDefine periods\nCalculate average differences using the datetime column\n\n```text\nAverage difference for 1992-2002 (TOPEX/Poseidon): 0.334 m\nAverage difference for 2002-2016: 0.043 m\nAverage difference for 2016-2023: 0.002 m\n```\n\nThe difference in water level estimates for Lake Victoria between the DAHITI and C3S-LWL v5.0 datasets varies noticeably over time (see *Figure 2*). A larger discrepancy is observed during the first decade of the time series (1992-2002), when the only satellite data available originated from the TOPEX/Poseidon mission. During this period, the average difference between the two datasets is 33.2 cm in average, with C3S values consistently higher than those from DAHITI.\n\nThis early discrepancy appears to stem from differences in the algorithms and retracking methods used by the two datasets. DAHITI, in some cases, applies an improved 10% threshold retracker for better perfomance on inland water bodies, depending on the specific case. However, specific information on whether the 10% retracker was applied to Lake Victoria could not be confirmed, since available documentation only details its application for lakes in the Americas. Additionally, DAHITI employs a Kalman filter to interpolate and smooth its time series, further influencing the results [[2]](https://hess.copernicus.org/articles/19/4345/2015/hess-19-4345-2015.html). In contrast, according to the [ATBD](https://confluence.ecmwf.int/pages/viewpage.action?pageId=425328942), the C3S-LWL v5.0 dataset uses a ocean model retracking for large lakes such as Lake Victoria.\n\nHowever, this difference diminishes noticeably after 2002 (less than 10 cm in average), as newer satellite missions introduced more frequent and accurate measurements. The influence of processing seems to be much smaller in later years, likely because the improvements in sensor technology and data availability reduce the impact of such methodological differences.\n\nThese findings are supported by the [PQAR](https://confluence.ecmwf.int/pages/viewpage.action?pageId=428248112), which includes Lake Victoria under the Southern Africa region. According to the PQAR, the Pearson correlation coefficient between DAHITI and C3S in this region is very close to 1, and the Unbiased Root Mean Square Error (URMSE) remains below 25%, indicating strong overall agreement in this region.\n\n(section-3)=", "text_with_prefix": "EQC Quality Assessment: \"Lake Victoria’s 2020 Flood Event Analysis\"\nDataset: satellite-lake-water-level [CDS]\nAspect: climate-and-weather-extremes_q01 | Category: Satellite_ECVs\nSection: Lake Victoria’s 2020 Flood Event Analysis > Analysis and results > 2. Comparison of C3S LWL with the DAHITI dataset > Plot C3S and DAHITI's Water Level\n---\nExtract time series from xarray (da)\nFinal touches\n\n*Figure 2. Lake Victoria's water level from 1992 to present from C3S-LWL v5.0 and DAHITI.*\n\nExtract time series from xarray (da)\nAlign by nearest timestamp\nDrop rows where no match was found\nCompute difference\nEnsure the datetime column is in datetime format\nDefine periods\nCalculate average differences using the datetime column\n\n```text\nAverage difference for 1992-2002 (TOPEX/Poseidon): 0.334 m\nAverage difference for 2002-2016: 0.043 m\nAverage difference for 2016-2023: 0.002 m\n```\n\nThe difference in water level estimates for Lake Victoria between the DAHITI and C3S-LWL v5.0 datasets varies noticeably over time (see *Figure 2*). A larger discrepancy is observed during the first decade of the time series (1992-2002), when the only satellite data available originated from the TOPEX/Poseidon mission. During this period, the average difference between the two datasets is 33.2 cm in average, with C3S values consistently higher than those from DAHITI.\n\nThis early discrepancy appears to stem from differences in the algorithms and retracking methods used by the two datasets. DAHITI, in some cases, applies an improved 10% threshold retracker for better perfomance on inland water bodies, depending on the specific case. However, specific information on whether the 10% retracker was applied to Lake Victoria could not be confirmed, since available documentation only details its application for lakes in the Americas. Additionally, DAHITI employs a Kalman filter to interpolate and smooth its time series, further influencing the results [[2]](https://hess.copernicus.org/articles/19/4345/2015/hess-19-4345-2015.html). In contrast, according to the [ATBD](https://confluence.ecmwf.int/pages/viewpage.action?pageId=425328942), the C3S-LWL v5.0 dataset uses a ocean model retracking for large lakes such as Lake Victoria.\n\nHowever, this difference diminishes noticeably after 2002 (less than 10 cm in average), as newer satellite missions introduced more frequent and accurate measurements. The influence of processing seems to be much smaller in later years, likely because the improvements in sensor technology and data availability reduce the impact of such methodological differences.\n\nThese findings are supported by the [PQAR](https://confluence.ecmwf.int/pages/viewpage.action?pageId=428248112), which includes Lake Victoria under the Southern Africa region. According to the PQAR, the Pearson correlation coefficient between DAHITI and C3S in this region is very close to 1, and the Unbiased Root Mean Square Error (URMSE) remains below 25%, indicating strong overall agreement in this region.\n\n(section-3)="} {"chunk_id": "satellite_satellite-lake-water-level_climate-and-weather-extremes_q01__ae2e94a94ab7", "report_id": "satellite_satellite-lake-water-level_climate-and-weather-extremes_q01", "dataset_id": "satellite-lake-water-level", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q01", "aspect_base": "climate-and-weather-extremes", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Lake Victoria’s 2020 Flood Event Analysis > Analysis and results > 3. Time series reconstruction", "title": "Lake Victoria’s 2020 Flood Event Analysis", "chunk_index": 7, "token_count": 180, "text_raw": "The temporal coverage of the C3S-LWL dataset varies by lake, with a target record length of more than 25 years and a minimum threshold of 10 years. To increase statistical confidence and improve robustness, other available datasets can be merged with C3S-LWL to form a single continuous record.\n\nThe code on this section is based on the work of Pietroiusti et al. (2024) [[1]](https://esd.copernicus.org/articles/15/225/2024/).", "text_with_prefix": "EQC Quality Assessment: \"Lake Victoria’s 2020 Flood Event Analysis\"\nDataset: satellite-lake-water-level [CDS]\nAspect: climate-and-weather-extremes_q01 | Category: Satellite_ECVs\nSection: Lake Victoria’s 2020 Flood Event Analysis > Analysis and results > 3. Time series reconstruction\n---\nThe temporal coverage of the C3S-LWL dataset varies by lake, with a target record length of more than 25 years and a minimum threshold of 10 years. To increase statistical confidence and improve robustness, other available datasets can be merged with C3S-LWL to form a single continuous record.\n\nThe code on this section is based on the work of Pietroiusti et al. (2024) [[1]](https://esd.copernicus.org/articles/15/225/2024/)."} {"chunk_id": "satellite_satellite-lake-water-level_climate-and-weather-extremes_q01__ff2b263f9cf4", "report_id": "satellite_satellite-lake-water-level_climate-and-weather-extremes_q01", "dataset_id": "satellite-lake-water-level", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q01", "aspect_base": "climate-and-weather-extremes", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Lake Victoria’s 2020 Flood Event Analysis > Analysis and results > 3. Time series reconstruction > Load and preprocess HYDROMET data", "title": "Lake Victoria’s 2020 Flood Event Analysis", "chunk_index": 8, "token_count": 545, "text_raw": "From January 1, 1948 to August 1, 1996, daily in situ water level measurements at the Jinja station were obtained from the WMO Hydrometeorological Survey (hereafter referred to as HYDROMET) [[3]](https://library.wmo.int/records/item/54830-hydrometeorological-survey-of-the-catchments-of-lakes-victoria-kyoga-and-albert?offset=32). These measurements, originally recorded as water depth at the lake’s outflow, were later converted to meters above sea level by adding a geoid correction of 1122.887 m, following the approach of Vanderkelen et al. (2018) [[4]](https://hess.copernicus.org/articles/22/5509/2018/).\n\nThe HYDROMET dataset is located in the [Zenodo](https://zenodo.org/records/10793917) repository at: lakevic-eea-data-zenodo\\lakevic-eea-data-zenodo\\lakevic-eea-analysis\\lakelevels\\Jinja_lakelevels_Van.txt.\n\nPre-process data\nReplace HYDROMET zeros with NaN, so that Matplotlib won't connect the points\n\n```text\nwater_level meas\ndate \n1948-01-01 1134.097 11.210\n1948-01-02 1134.102 11.215\n1948-01-03 1134.062 11.175\n1948-01-04 1134.052 11.165\n1948-01-05 1134.077 11.190\n... ... ...\n1996-07-28 1134.777 11.890\n1996-07-29 1134.757 11.870\n1996-07-30 1134.752 11.865\n1996-07-31 1134.717 11.830\n1996-08-01 1134.747 11.860\n\n[17746 rows x 2 columns]\n```", "text_with_prefix": "EQC Quality Assessment: \"Lake Victoria’s 2020 Flood Event Analysis\"\nDataset: satellite-lake-water-level [CDS]\nAspect: climate-and-weather-extremes_q01 | Category: Satellite_ECVs\nSection: Lake Victoria’s 2020 Flood Event Analysis > Analysis and results > 3. Time series reconstruction > Load and preprocess HYDROMET data\n---\nFrom January 1, 1948 to August 1, 1996, daily in situ water level measurements at the Jinja station were obtained from the WMO Hydrometeorological Survey (hereafter referred to as HYDROMET) [[3]](https://library.wmo.int/records/item/54830-hydrometeorological-survey-of-the-catchments-of-lakes-victoria-kyoga-and-albert?offset=32). These measurements, originally recorded as water depth at the lake’s outflow, were later converted to meters above sea level by adding a geoid correction of 1122.887 m, following the approach of Vanderkelen et al. (2018) [[4]](https://hess.copernicus.org/articles/22/5509/2018/).\n\nThe HYDROMET dataset is located in the [Zenodo](https://zenodo.org/records/10793917) repository at: lakevic-eea-data-zenodo\\lakevic-eea-data-zenodo\\lakevic-eea-analysis\\lakelevels\\Jinja_lakelevels_Van.txt.\n\nPre-process data\nReplace HYDROMET zeros with NaN, so that Matplotlib won't connect the points\n\n```text\nwater_level meas\ndate \n1948-01-01 1134.097 11.210\n1948-01-02 1134.102 11.215\n1948-01-03 1134.062 11.175\n1948-01-04 1134.052 11.165\n1948-01-05 1134.077 11.190\n... ... ...\n1996-07-28 1134.777 11.890\n1996-07-29 1134.757 11.870\n1996-07-30 1134.752 11.865\n1996-07-31 1134.717 11.830\n1996-08-01 1134.747 11.860\n\n[17746 rows x 2 columns]\n```"} {"chunk_id": "satellite_satellite-lake-water-level_climate-and-weather-extremes_q01__fecc9b73b236", "report_id": "satellite_satellite-lake-water-level_climate-and-weather-extremes_q01", "dataset_id": "satellite-lake-water-level", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q01", "aspect_base": "climate-and-weather-extremes", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Lake Victoria’s 2020 Flood Event Analysis > Analysis and results > 3. Time series reconstruction > Merging the datasets together", "title": "Lake Victoria’s 2020 Flood Event Analysis", "chunk_index": 9, "token_count": 609, "text_raw": "*Figure 3. HYDROMET (1948-1996) and C3S (1992-2023) datasets timeseries before correction.*\n\nTo obtain the reconstructed dataseries of Lake Victoria's water levels it is necessary to perform a correction to HYDROMET data based on the overlaping period (1992-1996).\n\nMerge the two time-series, key on date, keep all observations (outer)\nGet only the overlapping observations\n\n```text\nHYDROMET C3S\ndate \n1992-09-28 1134.567 1135.32\n1992-10-18 1134.572 1135.30\n1992-11-15 1134.597 1135.36\n1992-12-16 1134.672 1135.43\n1993-01-14 1134.767 1135.54\n```\n\n```text\nAverage Difference (C3S - HYDROMET) = 77.8 ± 3.0 cm\n```\n\nSnap the two timeseries together, add avg diff to HYDROMET\n\nLeft Plot: Full time series\nRight Plot: Zoomed section\n\n*Figure 4. HYDROMET (1948-1996) and C3S (1992-2023) datasets timeseries after correction.*\n\nThe average difference between the two datasets is 77.8 ± 3.0 cm [[1]](https://esd.copernicus.org/articles/15/225/2024/). The HYDROMET dataset was corrected using this value.\n\nCorrect for difference\nOverwriting time-series with C3S where available\n\nTo have only one value per day\nMake only one list with HYDROMET+C3S data\n\nRound to 2 sig figs\n\nCalculate basic statistics for the 'water_level' column\nCalculate 10% trimmed mean (removes lowest and highest 10% of values)\nCalculate percentiles\n\nInterpolate to daily res and round to 2 sig figs\n\nPlot the complete water levels timeseries with the statistics calculated\n\n*Figure 5. Reconstructed Lake Victoria's water levels timeseries (1948-2023).*\n\nA single dataset combining both sources was created. For the overlap period between 1992 and 1997, C3S LWL-v5.0 data were used when available. The gaps were filled by interpolation to have a daily resolution.\n\n(section-4)=", "text_with_prefix": "EQC Quality Assessment: \"Lake Victoria’s 2020 Flood Event Analysis\"\nDataset: satellite-lake-water-level [CDS]\nAspect: climate-and-weather-extremes_q01 | Category: Satellite_ECVs\nSection: Lake Victoria’s 2020 Flood Event Analysis > Analysis and results > 3. Time series reconstruction > Merging the datasets together\n---\n*Figure 3. HYDROMET (1948-1996) and C3S (1992-2023) datasets timeseries before correction.*\n\nTo obtain the reconstructed dataseries of Lake Victoria's water levels it is necessary to perform a correction to HYDROMET data based on the overlaping period (1992-1996).\n\nMerge the two time-series, key on date, keep all observations (outer)\nGet only the overlapping observations\n\n```text\nHYDROMET C3S\ndate \n1992-09-28 1134.567 1135.32\n1992-10-18 1134.572 1135.30\n1992-11-15 1134.597 1135.36\n1992-12-16 1134.672 1135.43\n1993-01-14 1134.767 1135.54\n```\n\n```text\nAverage Difference (C3S - HYDROMET) = 77.8 ± 3.0 cm\n```\n\nSnap the two timeseries together, add avg diff to HYDROMET\n\nLeft Plot: Full time series\nRight Plot: Zoomed section\n\n*Figure 4. HYDROMET (1948-1996) and C3S (1992-2023) datasets timeseries after correction.*\n\nThe average difference between the two datasets is 77.8 ± 3.0 cm [[1]](https://esd.copernicus.org/articles/15/225/2024/). The HYDROMET dataset was corrected using this value.\n\nCorrect for difference\nOverwriting time-series with C3S where available\n\nTo have only one value per day\nMake only one list with HYDROMET+C3S data\n\nRound to 2 sig figs\n\nCalculate basic statistics for the 'water_level' column\nCalculate 10% trimmed mean (removes lowest and highest 10% of values)\nCalculate percentiles\n\nInterpolate to daily res and round to 2 sig figs\n\nPlot the complete water levels timeseries with the statistics calculated\n\n*Figure 5. Reconstructed Lake Victoria's water levels timeseries (1948-2023).*\n\nA single dataset combining both sources was created. For the overlap period between 1992 and 1997, C3S LWL-v5.0 data were used when available. The gaps were filled by interpolation to have a daily resolution.\n\n(section-4)="} {"chunk_id": "satellite_satellite-lake-water-level_climate-and-weather-extremes_q01__587856510cf6", "report_id": "satellite_satellite-lake-water-level_climate-and-weather-extremes_q01", "dataset_id": "satellite-lake-water-level", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q01", "aspect_base": "climate-and-weather-extremes", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Lake Victoria’s 2020 Flood Event Analysis > Analysis and results > 4. Probabilistic extreme event analysis", "title": "Lake Victoria’s 2020 Flood Event Analysis", "chunk_index": 10, "token_count": 378, "text_raw": "Pietroiusti et al. (2024) [[1]](https://esd.copernicus.org/articles/15/225/2024/) assessed the influence of anthropogenic climate change on the 2020 Lake Victoria floods using the extreme event attribution framework outlined by Philip et al. (2020) [[5]](https://ascmo.copernicus.org/articles/6/177/2020/) and van Oldenborgh et al. (2021) [[6]](https://doi.org/10.1007/s10584-021-03071-7). The methodology follows a structured sequence of steps: (i) defining the event, (ii) estimating probabilities and trends based on observational data, (iii) validating climate models, (iv) conducting attribution using multiple models and methods, and (v) synthesizing the findings into clear attribution statements.\n\nIn this assessment, only the first step, event definition, is performed, using reconstructed Lake Victoria water level time series from the HYDROMET dataset and the C3S LWL-v5.0 product. The aim is to assess whether the C3S LWL-v5.0 dataset is suitable for use in attribution analyses of this kind. While the remaining steps involve model simulations and additional data sources that are beyond the scope of this assessment, the assumption is that if the observational component is consistent, the full attribution framework could, in principle, be applied using this dataset as well.", "text_with_prefix": "EQC Quality Assessment: \"Lake Victoria’s 2020 Flood Event Analysis\"\nDataset: satellite-lake-water-level [CDS]\nAspect: climate-and-weather-extremes_q01 | Category: Satellite_ECVs\nSection: Lake Victoria’s 2020 Flood Event Analysis > Analysis and results > 4. Probabilistic extreme event analysis\n---\nPietroiusti et al. (2024) [[1]](https://esd.copernicus.org/articles/15/225/2024/) assessed the influence of anthropogenic climate change on the 2020 Lake Victoria floods using the extreme event attribution framework outlined by Philip et al. (2020) [[5]](https://ascmo.copernicus.org/articles/6/177/2020/) and van Oldenborgh et al. (2021) [[6]](https://doi.org/10.1007/s10584-021-03071-7). The methodology follows a structured sequence of steps: (i) defining the event, (ii) estimating probabilities and trends based on observational data, (iii) validating climate models, (iv) conducting attribution using multiple models and methods, and (v) synthesizing the findings into clear attribution statements.\n\nIn this assessment, only the first step, event definition, is performed, using reconstructed Lake Victoria water level time series from the HYDROMET dataset and the C3S LWL-v5.0 product. The aim is to assess whether the C3S LWL-v5.0 dataset is suitable for use in attribution analyses of this kind. While the remaining steps involve model simulations and additional data sources that are beyond the scope of this assessment, the assumption is that if the observational component is consistent, the full attribution framework could, in principle, be applied using this dataset as well."} {"chunk_id": "satellite_satellite-lake-water-level_climate-and-weather-extremes_q01__90c3dd1daf9b", "report_id": "satellite_satellite-lake-water-level_climate-and-weather-extremes_q01", "dataset_id": "satellite-lake-water-level", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q01", "aspect_base": "climate-and-weather-extremes", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Lake Victoria’s 2020 Flood Event Analysis > Analysis and results > 4. Probabilistic extreme event analysis > Event definition", "title": "Lake Victoria’s 2020 Flood Event Analysis", "chunk_index": 11, "token_count": 1026, "text_raw": "The variable chosen for analysis is the rate of change in water levels (ΔL) over a time window (Δt), with Δt set to 180 days since most of the level rise in the 2020 Lake Victoria flooding occurred during the six-month period between November 2019 and May 2020 [[1]](https://esd.copernicus.org/articles/15/225/2024/). The code computes ΔL for every possible 180-day interval in the dataset, groups these values by year, and identifies the single largest 180-day change for each year, this is the the annual block maximum. The annual block maxima are then ranked to compare the magnitude of events across years. Finally, the ranking and maximum ΔL/Δt for 2020 are reported.\n\nSet the time window (dt)\nExtract annual block maxima\nKeep only years with full data\nCalculate rank of each year\nPrint 2020 event details\n\n```text\n🔹 2020 ΔL/Δt (180 days): 1.210 m\n🔹 2020 Rank: 3 out of 74 years\n```\n\nEnsure rolling mean exists\nGet top 3 values\nStep plot for annual maxima (red)\nRolling mean (green)\nTop 3 events (black markers + labels)\nLabels, title, grid\n\n*Figure 6. Annual block maxima time series $(\\Delta L / \\Delta t)_{\\text{max}}$ with Δt=180 d for the period 1897–2021 and 10-year rolling mean of the time series (using C3S-LWL v5.0 data).*\n\nattachment:e4f5e3f5-af36-4613-9df5-03a4f99fde20.png\n```\n*Figure 7. Annual block maxima time series $(\\Delta L / \\Delta t)_{\\text{max}}$ with Δt=180 d for the period 1897–2021 and 10-year rolling mean of the time series (using DAHITI data). Source: Pietroiusti et al. (2022) [[7]](https://cris.vub.be/ws/portalfiles/portal/95073120/Geography_MSc_thesis_Rosa_Pietroiusti_final.pdf)*\n\nThe results of this analysis are consistent with those reported by Pietroiusti et al. (2024) [[1]](https://esd.copernicus.org/articles/15/225/2024/). According to the C3S-LWL v5.0 dataset, the year 2020 recorded the third-largest 180-day lake level rise, surpassed only by 1962 and 1998. The magnitude of the 2020 rise was 1.21 meters, which matches the value obtained by Pietroiusti et al. using the DAHITI dataset.\n\nHowever, some discrepancies emerge when comparing other years. For instance, in 2000, the C3S-LWL v5.0 data shows a negative value, indicating a consistent decrease in lake levels throughout the year (see *Figure 6*). Conversely, the corresponding figure in Pietroiusti et al. (2022) [[7]](https://cris.vub.be/ws/portalfiles/portal/95073120/Geography_MSc_thesis_Rosa_Pietroiusti_final.pdf) (see *Figure 7*) shows a slightly positive value.\n\nThese differences likely stem from variations in the underlying datasets, as discussed in [](section-2). In 2000, TOPEX/Poseidon was still the only mission providing data for the C3S-LWL, with only one value per month. Meanwhile, DAHITI applied Kalman-filtered interpolation, resulting in a denser time series. Because no reliable in situ data are available for this period, it is difficult to determine which dataset more accurately reflects reality.\n\nNevertheless, the agreement between both datasets on the 2020 block maximum strengthens confidence in the reliability of this result. Although C3S-LWL and DAHITI use different processing algorithms, the increased availability of satellite observations in recent years (particularly after 2016) has resulted in greater similarity between their outputs. With more frequent and higher-quality measurements from multiple missions, the differences between the datasets diminish, making the 1.21 meter increase observed in 2020 not only consistent across sources, but also giving more confidence in the result.", "text_with_prefix": "EQC Quality Assessment: \"Lake Victoria’s 2020 Flood Event Analysis\"\nDataset: satellite-lake-water-level [CDS]\nAspect: climate-and-weather-extremes_q01 | Category: Satellite_ECVs\nSection: Lake Victoria’s 2020 Flood Event Analysis > Analysis and results > 4. Probabilistic extreme event analysis > Event definition\n---\nThe variable chosen for analysis is the rate of change in water levels (ΔL) over a time window (Δt), with Δt set to 180 days since most of the level rise in the 2020 Lake Victoria flooding occurred during the six-month period between November 2019 and May 2020 [[1]](https://esd.copernicus.org/articles/15/225/2024/). The code computes ΔL for every possible 180-day interval in the dataset, groups these values by year, and identifies the single largest 180-day change for each year, this is the the annual block maximum. The annual block maxima are then ranked to compare the magnitude of events across years. Finally, the ranking and maximum ΔL/Δt for 2020 are reported.\n\nSet the time window (dt)\nExtract annual block maxima\nKeep only years with full data\nCalculate rank of each year\nPrint 2020 event details\n\n```text\n🔹 2020 ΔL/Δt (180 days): 1.210 m\n🔹 2020 Rank: 3 out of 74 years\n```\n\nEnsure rolling mean exists\nGet top 3 values\nStep plot for annual maxima (red)\nRolling mean (green)\nTop 3 events (black markers + labels)\nLabels, title, grid\n\n*Figure 6. Annual block maxima time series $(\\Delta L / \\Delta t)_{\\text{max}}$ with Δt=180 d for the period 1897–2021 and 10-year rolling mean of the time series (using C3S-LWL v5.0 data).*\n\nattachment:e4f5e3f5-af36-4613-9df5-03a4f99fde20.png\n```\n*Figure 7. Annual block maxima time series $(\\Delta L / \\Delta t)_{\\text{max}}$ with Δt=180 d for the period 1897–2021 and 10-year rolling mean of the time series (using DAHITI data). Source: Pietroiusti et al. (2022) [[7]](https://cris.vub.be/ws/portalfiles/portal/95073120/Geography_MSc_thesis_Rosa_Pietroiusti_final.pdf)*\n\nThe results of this analysis are consistent with those reported by Pietroiusti et al. (2024) [[1]](https://esd.copernicus.org/articles/15/225/2024/). According to the C3S-LWL v5.0 dataset, the year 2020 recorded the third-largest 180-day lake level rise, surpassed only by 1962 and 1998. The magnitude of the 2020 rise was 1.21 meters, which matches the value obtained by Pietroiusti et al. using the DAHITI dataset.\n\nHowever, some discrepancies emerge when comparing other years. For instance, in 2000, the C3S-LWL v5.0 data shows a negative value, indicating a consistent decrease in lake levels throughout the year (see *Figure 6*). Conversely, the corresponding figure in Pietroiusti et al. (2022) [[7]](https://cris.vub.be/ws/portalfiles/portal/95073120/Geography_MSc_thesis_Rosa_Pietroiusti_final.pdf) (see *Figure 7*) shows a slightly positive value.\n\nThese differences likely stem from variations in the underlying datasets, as discussed in [](section-2). In 2000, TOPEX/Poseidon was still the only mission providing data for the C3S-LWL, with only one value per month. Meanwhile, DAHITI applied Kalman-filtered interpolation, resulting in a denser time series. Because no reliable in situ data are available for this period, it is difficult to determine which dataset more accurately reflects reality.\n\nNevertheless, the agreement between both datasets on the 2020 block maximum strengthens confidence in the reliability of this result. Although C3S-LWL and DAHITI use different processing algorithms, the increased availability of satellite observations in recent years (particularly after 2016) has resulted in greater similarity between their outputs. With more frequent and higher-quality measurements from multiple missions, the differences between the datasets diminish, making the 1.21 meter increase observed in 2020 not only consistent across sources, but also giving more confidence in the result."} {"chunk_id": "satellite_satellite-lake-water-level_climate-and-weather-extremes_q01__a350eef12477", "report_id": "satellite_satellite-lake-water-level_climate-and-weather-extremes_q01", "dataset_id": "satellite-lake-water-level", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q01", "aspect_base": "climate-and-weather-extremes", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Lake Victoria’s 2020 Flood Event Analysis > Analysis and results > 4. Probabilistic extreme event analysis > Event definition", "title": "Lake Victoria’s 2020 Flood Event Analysis", "chunk_index": 12, "token_count": 214, "text_raw": ". With more frequent and higher-quality measurements from multiple missions, the differences between the datasets diminish, making the 1.21 meter increase observed in 2020 not only consistent across sources, but also giving more confidence in the result.\n\nThis assessment supports the suitability of the C3S-LWL v5.0 dataset for extreme event attribution applications, as the event-definition results closely match those from Pietroiusti et al. (2024) [[1]](https://esd.copernicus.org/articles/15/225/2024/). Since this step feeds directly into the rest of the workflow, similar outcomes would be expected in a full attribution analysis.", "text_with_prefix": "EQC Quality Assessment: \"Lake Victoria’s 2020 Flood Event Analysis\"\nDataset: satellite-lake-water-level [CDS]\nAspect: climate-and-weather-extremes_q01 | Category: Satellite_ECVs\nSection: Lake Victoria’s 2020 Flood Event Analysis > Analysis and results > 4. Probabilistic extreme event analysis > Event definition\n---\n. With more frequent and higher-quality measurements from multiple missions, the differences between the datasets diminish, making the 1.21 meter increase observed in 2020 not only consistent across sources, but also giving more confidence in the result.\n\nThis assessment supports the suitability of the C3S-LWL v5.0 dataset for extreme event attribution applications, as the event-definition results closely match those from Pietroiusti et al. (2024) [[1]](https://esd.copernicus.org/articles/15/225/2024/). Since this step feeds directly into the rest of the workflow, similar outcomes would be expected in a full attribution analysis."} {"chunk_id": "satellite_satellite-lake-water-level_climate-and-weather-extremes_q01__b1fb50772bcb", "report_id": "satellite_satellite-lake-water-level_climate-and-weather-extremes_q01", "dataset_id": "satellite-lake-water-level", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q01", "aspect_base": "climate-and-weather-extremes", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Lake Victoria’s 2020 Flood Event Analysis > ℹ️ If you want to know more", "title": "Lake Victoria’s 2020 Flood Event Analysis", "chunk_index": 13, "token_count": 211, "text_raw": "* NASA (2021). [Lake Victoria’s Rising Waters](https://earthobservatory.nasa.gov/images/148414/lake-victorias-rising-waters)\n* ACSA Uganda & Uganda Coalition for Sustainble Development (UCSD) (2020). [The Implication of Floods to Food Security During and the Aftermath of COVID-19 Pandemic in Uganda](https://acsa-ug.org/wordpress/wp-content/uploads/2020/06/THE-IMPLICATION-OF-FLOODS-TO-FOOD-SECURITY-DURING-AND-THE-AFTERMATH-OF-COVID19-PANDEMIC-IN-UGANDA.pdf)", "text_with_prefix": "EQC Quality Assessment: \"Lake Victoria’s 2020 Flood Event Analysis\"\nDataset: satellite-lake-water-level [CDS]\nAspect: climate-and-weather-extremes_q01 | Category: Satellite_ECVs\nSection: Lake Victoria’s 2020 Flood Event Analysis > ℹ️ If you want to know more\n---\n* NASA (2021). [Lake Victoria’s Rising Waters](https://earthobservatory.nasa.gov/images/148414/lake-victorias-rising-waters)\n* ACSA Uganda & Uganda Coalition for Sustainble Development (UCSD) (2020). [The Implication of Floods to Food Security During and the Aftermath of COVID-19 Pandemic in Uganda](https://acsa-ug.org/wordpress/wp-content/uploads/2020/06/THE-IMPLICATION-OF-FLOODS-TO-FOOD-SECURITY-DURING-AND-THE-AFTERMATH-OF-COVID19-PANDEMIC-IN-UGANDA.pdf)"} {"chunk_id": "satellite_satellite-lake-water-level_climate-and-weather-extremes_q01__f677a9be5a3f", "report_id": "satellite_satellite-lake-water-level_climate-and-weather-extremes_q01", "dataset_id": "satellite-lake-water-level", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q01", "aspect_base": "climate-and-weather-extremes", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Lake Victoria’s 2020 Flood Event Analysis > ℹ️ If you want to know more > Key resources", "title": "Lake Victoria’s 2020 Flood Event Analysis", "chunk_index": 14, "token_count": 283, "text_raw": "Code libraries used:\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)\n\nDataset documentation:\n\n* [LWL v5.0 and LWL-S v1.0: Product User Guide and Specification (PUGS)](https://confluence.ecmwf.int/pages/viewpage.action?pageId=425328964)\n* [LWL v5.0 and LWL-S v1.0: Algorithm Theoretical Basis Document (ATBD)](https://confluence.ecmwf.int/pages/viewpage.action?pageId=425328942)\n* [LWL v5.0 and LWL-S v1.0: Product Quality Assessment Report (PQAR)](https://confluence.ecmwf.int/pages/viewpage.action?pageId=428248112)", "text_with_prefix": "EQC Quality Assessment: \"Lake Victoria’s 2020 Flood Event Analysis\"\nDataset: satellite-lake-water-level [CDS]\nAspect: climate-and-weather-extremes_q01 | Category: Satellite_ECVs\nSection: Lake Victoria’s 2020 Flood Event Analysis > ℹ️ If you want to know more > Key resources\n---\nCode libraries used:\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)\n\nDataset documentation:\n\n* [LWL v5.0 and LWL-S v1.0: Product User Guide and Specification (PUGS)](https://confluence.ecmwf.int/pages/viewpage.action?pageId=425328964)\n* [LWL v5.0 and LWL-S v1.0: Algorithm Theoretical Basis Document (ATBD)](https://confluence.ecmwf.int/pages/viewpage.action?pageId=425328942)\n* [LWL v5.0 and LWL-S v1.0: Product Quality Assessment Report (PQAR)](https://confluence.ecmwf.int/pages/viewpage.action?pageId=428248112)"} {"chunk_id": "satellite_satellite-lake-water-level_climate-and-weather-extremes_q01__30ae3250d21a", "report_id": "satellite_satellite-lake-water-level_climate-and-weather-extremes_q01", "dataset_id": "satellite-lake-water-level", "store": "CDS", "doc_type": "EQC_QA", "aspect": "climate-and-weather-extremes_q01", "aspect_base": "climate-and-weather-extremes", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Lake Victoria’s 2020 Flood Event Analysis > ℹ️ If you want to know more > References", "title": "Lake Victoria’s 2020 Flood Event Analysis", "chunk_index": 15, "token_count": 901, "text_raw": "[[1]](https://esd.copernicus.org/articles/15/225/2024/) Pietroiusti, R., Vanderkelen, I., Otto, F. E. L., Barnes, C., Temple, L., Akurut, M., Bally, P., van Lipzig, N. P. M., and Thiery, W. (2024). Possible role of anthropogenic climate change in the record-breaking 2020 Lake Victoria levels and floods, Earth Syst. Dynam., 15, 225–264.\n\n[[2]](https://hess.copernicus.org/articles/19/4345/2015/hess-19-4345-2015.html) Schwatke, C., Dettmering, D., Bosch, W., and Seitz, F. (2015). DAHITI - an innovative approach for estimating water level time series over inland waters using multi-mission satellite altimetry: , Hydrol. Earth Syst. Sci., 19, 4345-4364.\n\n[[3]](https://library.wmo.int/records/item/54830-hydrometeorological-survey-of-the-catchments-of-lakes-victoria-kyoga-and-albert?offset=32) WMO-UNPD (1974). Hydrometeorological Survey of the Catchments of Lakes Victoria, Kyoga\nand Albert: Vol 1 Meteorology and Hydrology of the Basin.\n\n[[4]](https://hess.copernicus.org/articles/22/5509/2018/) Vanderkelen, I., Van Lipzig, N. P., and Thiery, W. (2018). Modelling the water balance of Lake Victoria (East Africa)-Part 1: Observational analysis. Hydrology and Earth System Sciences, 22(10):5509–5525.\n\n[[5]](https://ascmo.copernicus.org/articles/6/177/2020/) Philip, S., Kew, S., van Oldenborgh, G. J., Otto, F., Vautard, R., van der Wiel, K., King, A., Lott, F., Arrighi, J., Singh, R., and van Aalst, M. (2020). A protocol for probabilistic extreme event attribution analyses, Adv. Stat. Clim. Meteorol. Oceanogr., 6, 177–203.\n\n[[6]](https://doi.org/10.1007/s10584-021-03071-7) G. J. van Oldenborgh, K. van der Wiel, S. Kew, S. Philip, F. Otto, R. Vautard, A. King, F. Lott, J. Arrighi, R. Singh, and M. van Aalst. Pathways and pitfalls in extreme event attribution. Climatic Change, vol. 166, no. 1, p. 13, 2021.\n\n[[7]](https://cris.vub.be/ws/portalfiles/portal/95073120/Geography_MSc_thesis_Rosa_Pietroiusti_final.pdf) Pietroiusti, R., Vanderkelen, I., van Lipzig, N. P. M., and Thiery, W. (2022). Was the 2020 Lake Victoria flooding ‘caused’ by anthropogenic climate change? An event attribution study. M.Sc. thesis, Dept. of Hydrology and Climate, Vrije Universiteit Brussel and KU Leuven.\n\n[[8]](https://link.springer.com/book/10.1007/978-1-4471-3675-0) Coles, S. (2001). An Introduction to Statistical Modeling of Extreme Values. Springer Series in Statistics. Springer-Verlag London. ISBN: 978-1-85233-459-8.", "text_with_prefix": "EQC Quality Assessment: \"Lake Victoria’s 2020 Flood Event Analysis\"\nDataset: satellite-lake-water-level [CDS]\nAspect: climate-and-weather-extremes_q01 | Category: Satellite_ECVs\nSection: Lake Victoria’s 2020 Flood Event Analysis > ℹ️ If you want to know more > References\n---\n[[1]](https://esd.copernicus.org/articles/15/225/2024/) Pietroiusti, R., Vanderkelen, I., Otto, F. E. L., Barnes, C., Temple, L., Akurut, M., Bally, P., van Lipzig, N. P. M., and Thiery, W. (2024). Possible role of anthropogenic climate change in the record-breaking 2020 Lake Victoria levels and floods, Earth Syst. Dynam., 15, 225–264.\n\n[[2]](https://hess.copernicus.org/articles/19/4345/2015/hess-19-4345-2015.html) Schwatke, C., Dettmering, D., Bosch, W., and Seitz, F. (2015). DAHITI - an innovative approach for estimating water level time series over inland waters using multi-mission satellite altimetry: , Hydrol. Earth Syst. Sci., 19, 4345-4364.\n\n[[3]](https://library.wmo.int/records/item/54830-hydrometeorological-survey-of-the-catchments-of-lakes-victoria-kyoga-and-albert?offset=32) WMO-UNPD (1974). Hydrometeorological Survey of the Catchments of Lakes Victoria, Kyoga\nand Albert: Vol 1 Meteorology and Hydrology of the Basin.\n\n[[4]](https://hess.copernicus.org/articles/22/5509/2018/) Vanderkelen, I., Van Lipzig, N. P., and Thiery, W. (2018). Modelling the water balance of Lake Victoria (East Africa)-Part 1: Observational analysis. Hydrology and Earth System Sciences, 22(10):5509–5525.\n\n[[5]](https://ascmo.copernicus.org/articles/6/177/2020/) Philip, S., Kew, S., van Oldenborgh, G. J., Otto, F., Vautard, R., van der Wiel, K., King, A., Lott, F., Arrighi, J., Singh, R., and van Aalst, M. (2020). A protocol for probabilistic extreme event attribution analyses, Adv. Stat. Clim. Meteorol. Oceanogr., 6, 177–203.\n\n[[6]](https://doi.org/10.1007/s10584-021-03071-7) G. J. van Oldenborgh, K. van der Wiel, S. Kew, S. Philip, F. Otto, R. Vautard, A. King, F. Lott, J. Arrighi, R. Singh, and M. van Aalst. Pathways and pitfalls in extreme event attribution. Climatic Change, vol. 166, no. 1, p. 13, 2021.\n\n[[7]](https://cris.vub.be/ws/portalfiles/portal/95073120/Geography_MSc_thesis_Rosa_Pietroiusti_final.pdf) Pietroiusti, R., Vanderkelen, I., van Lipzig, N. P. M., and Thiery, W. (2022). Was the 2020 Lake Victoria flooding ‘caused’ by anthropogenic climate change? An event attribution study. M.Sc. thesis, Dept. of Hydrology and Climate, Vrije Universiteit Brussel and KU Leuven.\n\n[[8]](https://link.springer.com/book/10.1007/978-1-4471-3675-0) Coles, S. (2001). An Introduction to Statistical Modeling of Extreme Values. Springer Series in Statistics. Springer-Verlag London. ISBN: 978-1-85233-459-8."} {"chunk_id": "satellite_satellite-lake-water-level_completeness_q02__c811029a47e3", "report_id": "satellite_satellite-lake-water-level_completeness_q02", "dataset_id": "satellite-lake-water-level", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q02", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Assessing Lake Titicaca’s water levels in support of water management > Quality assessment question(s)", "title": "Assessing Lake Titicaca’s water levels in support of water management", "chunk_index": 0, "token_count": 563, "text_raw": "* **Is the temporal completeness of the satellite-derived lake water level dataset for Lake Titicaca sufficient to support water resources management in the Titicaca–Desaguadero–Poopó–Salar de Coipasa (TDPS) System?**\n\nThe [lake water levels from 1992 to present derived from satellite observations](https://cds.climate.copernicus.eu/datasets/satellite-lake-water-level?tab=overview) (C3S-LWL v5.0) dataset provides long-term water level observations for numerous lakes worldwide. The Climate Data Record (CDR) of the C3S-LWL v5.0 dataset is updated annually. The dataset provides at least 10 years of temporal coverage. However, the specific temporal coverage and average time step vary by lake (time steps are irregular). The dataset provides one value of water level per time step.\n\nThis dataset can be useful for Integrated Water Resources Management (IWRM) applications such as understanding lake responses to seasonal dynamics, climate variability, and long-term climate change-factors which directly influence water availability for ecosystems and dependent human populations. To illustrate the potential applicability of the dataset, Lake Titicaca was selected as a case study. The lake forms part of the Titicaca–Desaguadero–Poopó–Salar de Coipasa (TDPS) hydrological system, which spans Peru and Bolivia. This system provides critical freshwater resources to millions of people, supporting agriculture, domestic use, and ecosystem services across the Altiplano. Lake Titicaca feeds the Desaguadero River, which flows into Lake Poopó. This hydrological system is complex and subject to multiple drivers (e.g., climate variability, land use change, population growth), which exert pressures on water availability and quality. These pressures affect the state of the system, such as lake levels, which in turn have impacts on ecosystems and human livelihoods [[1]](https://doi.org/10.3390/w8040144) [[2]](https://doi.org/10.1016/j.ejrh.2021.100927).\n\nThe objective of this assessment is to examine the temporal completeness of Lake Titicaca's water level data in the C3S-LWL v5.0 dataset and assess its ability to support Integrated Water Resources Management (IWRM) in the context of the TDPS system.", "text_with_prefix": "EQC Quality Assessment: \"Assessing Lake Titicaca’s water levels in support of water management\"\nDataset: satellite-lake-water-level [CDS]\nAspect: completeness_q02 | Category: Satellite_ECVs\nSection: Assessing Lake Titicaca’s water levels in support of water management > Quality assessment question(s)\n---\n* **Is the temporal completeness of the satellite-derived lake water level dataset for Lake Titicaca sufficient to support water resources management in the Titicaca–Desaguadero–Poopó–Salar de Coipasa (TDPS) System?**\n\nThe [lake water levels from 1992 to present derived from satellite observations](https://cds.climate.copernicus.eu/datasets/satellite-lake-water-level?tab=overview) (C3S-LWL v5.0) dataset provides long-term water level observations for numerous lakes worldwide. The Climate Data Record (CDR) of the C3S-LWL v5.0 dataset is updated annually. The dataset provides at least 10 years of temporal coverage. However, the specific temporal coverage and average time step vary by lake (time steps are irregular). The dataset provides one value of water level per time step.\n\nThis dataset can be useful for Integrated Water Resources Management (IWRM) applications such as understanding lake responses to seasonal dynamics, climate variability, and long-term climate change-factors which directly influence water availability for ecosystems and dependent human populations. To illustrate the potential applicability of the dataset, Lake Titicaca was selected as a case study. The lake forms part of the Titicaca–Desaguadero–Poopó–Salar de Coipasa (TDPS) hydrological system, which spans Peru and Bolivia. This system provides critical freshwater resources to millions of people, supporting agriculture, domestic use, and ecosystem services across the Altiplano. Lake Titicaca feeds the Desaguadero River, which flows into Lake Poopó. This hydrological system is complex and subject to multiple drivers (e.g., climate variability, land use change, population growth), which exert pressures on water availability and quality. These pressures affect the state of the system, such as lake levels, which in turn have impacts on ecosystems and human livelihoods [[1]](https://doi.org/10.3390/w8040144) [[2]](https://doi.org/10.1016/j.ejrh.2021.100927).\n\nThe objective of this assessment is to examine the temporal completeness of Lake Titicaca's water level data in the C3S-LWL v5.0 dataset and assess its ability to support Integrated Water Resources Management (IWRM) in the context of the TDPS system."} {"chunk_id": "satellite_satellite-lake-water-level_completeness_q02__5b4b43edb710", "report_id": "satellite_satellite-lake-water-level_completeness_q02", "dataset_id": "satellite-lake-water-level", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q02", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Assessing Lake Titicaca’s water levels in support of water management > Quality assessment statement", "title": "Assessing Lake Titicaca’s water levels in support of water management", "chunk_index": 1, "token_count": 345, "text_raw": "These are the key outcomes of this assessment\n\n* The temporal completeness of the C3S–LWL v5.0 dataset improves significantly after 2016, following the deployment of Sentinel-3A and Jason-3, and is further enhanced by the launches of Sentinel-3B (2018) and Sentinel-6A (2020).\n\n* The C3S-LWL dataset captures the seasonality of Lake Titicaca effectively, except in years with extended gaps in observations, such as 2012. If many months in a year have no data in the C3S-LWL dataset the results are visibly different to those of in-situ data.\n\n* The C3S–LWL timeseries captures the major flooding events of 2003 and 2004 in Lake Titicaca, demonstrating its ability to represent cumulative seasonal peaks. However, its limited temporal resolution (specially before 2019) makes it unsuitable for detecting short-lived flash floods caused by isolated extreme precipitation events.\n\n* The cumulative impact of major droughts, such as the 2023 event, on Lake Titicaca’s water level is evident in the C3S–LWL timeseries.\n\n* The current temporal span and gaps in the dataset limit its ability to reliably detect and analyse ENSO-related effects on Titicaca lake's water levels.\n\n```", "text_with_prefix": "EQC Quality Assessment: \"Assessing Lake Titicaca’s water levels in support of water management\"\nDataset: satellite-lake-water-level [CDS]\nAspect: completeness_q02 | Category: Satellite_ECVs\nSection: Assessing Lake Titicaca’s water levels in support of water management > Quality assessment statement\n---\nThese are the key outcomes of this assessment\n\n* The temporal completeness of the C3S–LWL v5.0 dataset improves significantly after 2016, following the deployment of Sentinel-3A and Jason-3, and is further enhanced by the launches of Sentinel-3B (2018) and Sentinel-6A (2020).\n\n* The C3S-LWL dataset captures the seasonality of Lake Titicaca effectively, except in years with extended gaps in observations, such as 2012. If many months in a year have no data in the C3S-LWL dataset the results are visibly different to those of in-situ data.\n\n* The C3S–LWL timeseries captures the major flooding events of 2003 and 2004 in Lake Titicaca, demonstrating its ability to represent cumulative seasonal peaks. However, its limited temporal resolution (specially before 2019) makes it unsuitable for detecting short-lived flash floods caused by isolated extreme precipitation events.\n\n* The cumulative impact of major droughts, such as the 2023 event, on Lake Titicaca’s water level is evident in the C3S–LWL timeseries.\n\n* The current temporal span and gaps in the dataset limit its ability to reliably detect and analyse ENSO-related effects on Titicaca lake's water levels.\n\n```"} {"chunk_id": "satellite_satellite-lake-water-level_completeness_q02__5a4a5f52d3b9", "report_id": "satellite_satellite-lake-water-level_completeness_q02", "dataset_id": "satellite-lake-water-level", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q02", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Assessing Lake Titicaca’s water levels in support of water management > Methodology", "title": "Assessing Lake Titicaca’s water levels in support of water management", "chunk_index": 2, "token_count": 512, "text_raw": "C3S-LWL data for Lake Titicaca were downloaded and evaluated for temporal coverage, including median and maximum time steps and the occurrence of data gaps. Next, water level time series were plotted to examine seasonality and historically known extreme events, with visual comparisons made to in-situ observations from the Bolivian National Meteorological and Hydrological Service (SENAMHI) . The relationship between ENSO phases and lake levels was also explored following established approaches. Finally, percentile ranks were calculated for key performance indicators to assess data completeness, and the suitability of the dataset for similar analyses in other lakes, supporting Integrated Water Resources Management, was evaluated.\n\nThe analysis and results are organised in the following steps, which are detailed in the sections below:\n\n**[](section-1)**\n * Download all available C3S–LWL v5.0 satellite-lake-water-level data for Lake Titicaca.\n\n**[](section-2)**\n * Calculate performance indicators related to temporal resolution: median timestep and max. timestep for different periods of time.\n * Plot the number of water level records per month across the dataset’s full temporal span.\n * Analyse the causes of data gaps and changes in data availability over time using information from the dataset’s Documentation.\n\n**[](section-3)**\n * Plot the water level time series, highlighting points relevant for analysing seasonality as well as historically reported extreme events.\n * Compare CS3-LWL and SENAMHI data visually. \n * Evaluate whether the dataset captures seasonality and extreme events.\n * Analyse the relationship between ENSO phases and Lake Titicaca’s water levels using an approach adapted from Gutierrez-Villarreal et al. (2024) [[3]](https://doi.org/10.1016/j.wace.2024.100710), and assess whether this relationship is evident in the C3S–LWL v5.0 dataset.\n\n**[](section-4)**\n * Calculate the percentile ranks of Lake Titicaca across key performance indicators to assess its relative data completeness within the dataset.\n * Analysed the suitability of the dataset for applying similar assessments to other lakes in support of IWRM.", "text_with_prefix": "EQC Quality Assessment: \"Assessing Lake Titicaca’s water levels in support of water management\"\nDataset: satellite-lake-water-level [CDS]\nAspect: completeness_q02 | Category: Satellite_ECVs\nSection: Assessing Lake Titicaca’s water levels in support of water management > Methodology\n---\nC3S-LWL data for Lake Titicaca were downloaded and evaluated for temporal coverage, including median and maximum time steps and the occurrence of data gaps. Next, water level time series were plotted to examine seasonality and historically known extreme events, with visual comparisons made to in-situ observations from the Bolivian National Meteorological and Hydrological Service (SENAMHI) . The relationship between ENSO phases and lake levels was also explored following established approaches. Finally, percentile ranks were calculated for key performance indicators to assess data completeness, and the suitability of the dataset for similar analyses in other lakes, supporting Integrated Water Resources Management, was evaluated.\n\nThe analysis and results are organised in the following steps, which are detailed in the sections below:\n\n**[](section-1)**\n * Download all available C3S–LWL v5.0 satellite-lake-water-level data for Lake Titicaca.\n\n**[](section-2)**\n * Calculate performance indicators related to temporal resolution: median timestep and max. timestep for different periods of time.\n * Plot the number of water level records per month across the dataset’s full temporal span.\n * Analyse the causes of data gaps and changes in data availability over time using information from the dataset’s Documentation.\n\n**[](section-3)**\n * Plot the water level time series, highlighting points relevant for analysing seasonality as well as historically reported extreme events.\n * Compare CS3-LWL and SENAMHI data visually. \n * Evaluate whether the dataset captures seasonality and extreme events.\n * Analyse the relationship between ENSO phases and Lake Titicaca’s water levels using an approach adapted from Gutierrez-Villarreal et al. (2024) [[3]](https://doi.org/10.1016/j.wace.2024.100710), and assess whether this relationship is evident in the C3S–LWL v5.0 dataset.\n\n**[](section-4)**\n * Calculate the percentile ranks of Lake Titicaca across key performance indicators to assess its relative data completeness within the dataset.\n * Analysed the suitability of the dataset for applying similar assessments to other lakes in support of IWRM."} {"chunk_id": "satellite_satellite-lake-water-level_completeness_q02__90ee836137be", "report_id": "satellite_satellite-lake-water-level_completeness_q02", "dataset_id": "satellite-lake-water-level", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q02", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Assessing Lake Titicaca’s water levels in support of water management > Analysis and results > 1. Request and download data > Download data", "title": "Assessing Lake Titicaca’s water levels in support of water management", "chunk_index": 3, "token_count": 142, "text_raw": "```text\n100%|██████████| 1/1 [00:00<00:00, 3.98it/s]\n```\n\nExtract the DataArray for water_surface_height_above_reference_datum\nExtract the DataArray for water_surface_height_uncertainty\n\n(section-2)=", "text_with_prefix": "EQC Quality Assessment: \"Assessing Lake Titicaca’s water levels in support of water management\"\nDataset: satellite-lake-water-level [CDS]\nAspect: completeness_q02 | Category: Satellite_ECVs\nSection: Assessing Lake Titicaca’s water levels in support of water management > Analysis and results > 1. Request and download data > Download data\n---\n```text\n100%|██████████| 1/1 [00:00<00:00, 3.98it/s]\n```\n\nExtract the DataArray for water_surface_height_above_reference_datum\nExtract the DataArray for water_surface_height_uncertainty\n\n(section-2)="} {"chunk_id": "satellite_satellite-lake-water-level_completeness_q02__cf1224ff5425", "report_id": "satellite_satellite-lake-water-level_completeness_q02", "dataset_id": "satellite-lake-water-level", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q02", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Assessing Lake Titicaca’s water levels in support of water management > Analysis and results > 2. Temporal completeness > Performance indicators", "title": "Assessing Lake Titicaca’s water levels in support of water management", "chunk_index": 4, "token_count": 321, "text_raw": "The Product Quality Assessment Report ([PQAR](https://confluence.ecmwf.int/pages/viewpage.action?pageId=428248112)) for the dataset includes a table in Annex A presenting the performance indicators calculated for all lakes. These indicators are computed for the entire available period (1992–2023) and for the most recent 10-year period (2014–2023). The median and maximum timesteps are useful indicators for assessing temporal completeness. In this section, these indicators will be calculated for Lake Titicaca and compared with the information provided in the PQAR. The maximum timestep represents the largest gap between two observations, expressed as the maximum number of days without data.\n\nCalculate time differences\nSubset for 2014 to 2023\nPrint results\n\n```text\nAll Data:\nMedian Timestep (days): 28\nMax Timestep (days): 142\nTimeseries Duration (years): 28.5972602739726\n\nSubset 2014-2023:\nMedian Timestep (days): 14\nMax Timestep (days): 93\n```\n\nThe integer part of the results are the same as the ones presented in the PQAR.", "text_with_prefix": "EQC Quality Assessment: \"Assessing Lake Titicaca’s water levels in support of water management\"\nDataset: satellite-lake-water-level [CDS]\nAspect: completeness_q02 | Category: Satellite_ECVs\nSection: Assessing Lake Titicaca’s water levels in support of water management > Analysis and results > 2. Temporal completeness > Performance indicators\n---\nThe Product Quality Assessment Report ([PQAR](https://confluence.ecmwf.int/pages/viewpage.action?pageId=428248112)) for the dataset includes a table in Annex A presenting the performance indicators calculated for all lakes. These indicators are computed for the entire available period (1992–2023) and for the most recent 10-year period (2014–2023). The median and maximum timesteps are useful indicators for assessing temporal completeness. In this section, these indicators will be calculated for Lake Titicaca and compared with the information provided in the PQAR. The maximum timestep represents the largest gap between two observations, expressed as the maximum number of days without data.\n\nCalculate time differences\nSubset for 2014 to 2023\nPrint results\n\n```text\nAll Data:\nMedian Timestep (days): 28\nMax Timestep (days): 142\nTimeseries Duration (years): 28.5972602739726\n\nSubset 2014-2023:\nMedian Timestep (days): 14\nMax Timestep (days): 93\n```\n\nThe integer part of the results are the same as the ones presented in the PQAR."} {"chunk_id": "satellite_satellite-lake-water-level_completeness_q02__42e4db2a6370", "report_id": "satellite_satellite-lake-water-level_completeness_q02", "dataset_id": "satellite-lake-water-level", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q02", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Assessing Lake Titicaca’s water levels in support of water management > Analysis and results > 2. Temporal completeness > Records per month", "title": "Assessing Lake Titicaca’s water levels in support of water management", "chunk_index": 5, "token_count": 327, "text_raw": "Resample to monthly frequency and count non-NaN values\nConvert to Pandas for easier plotting\nPlot as bars\nCustomize the x-axis labels to show only years\n\n*Figure 1. Count of values available in the CDS water level dataset per month over time.*\n\nA clear increase in the number of measurements per month can be observed from 2020 onwards. Therefore, the median and maximum timestep indicators were calculated again for two periods: 1995-2019 and 2020-2023.\n\nCalculate time differences\nSubset for 1995 to 2019\nPrint results\nSubset for 2020 to 2023\n\n```text\nSubset 1995-2019:\nMedian Timestep (days): 31\nMax Timestep (days): 142\n\nSubset 2020-2023:\nMedian Timestep (days): 12\nMax Timestep (days): 93\n```\n\nThe results show that for Lake Titicaca, from 1995 to 2019, there is approximately one measurement per month (median time step of 31 days). From 2020 onwards, the number of measurements per month increases to a median time step of 12 days.", "text_with_prefix": "EQC Quality Assessment: \"Assessing Lake Titicaca’s water levels in support of water management\"\nDataset: satellite-lake-water-level [CDS]\nAspect: completeness_q02 | Category: Satellite_ECVs\nSection: Assessing Lake Titicaca’s water levels in support of water management > Analysis and results > 2. Temporal completeness > Records per month\n---\nResample to monthly frequency and count non-NaN values\nConvert to Pandas for easier plotting\nPlot as bars\nCustomize the x-axis labels to show only years\n\n*Figure 1. Count of values available in the CDS water level dataset per month over time.*\n\nA clear increase in the number of measurements per month can be observed from 2020 onwards. Therefore, the median and maximum timestep indicators were calculated again for two periods: 1995-2019 and 2020-2023.\n\nCalculate time differences\nSubset for 1995 to 2019\nPrint results\nSubset for 2020 to 2023\n\n```text\nSubset 1995-2019:\nMedian Timestep (days): 31\nMax Timestep (days): 142\n\nSubset 2020-2023:\nMedian Timestep (days): 12\nMax Timestep (days): 93\n```\n\nThe results show that for Lake Titicaca, from 1995 to 2019, there is approximately one measurement per month (median time step of 31 days). From 2020 onwards, the number of measurements per month increases to a median time step of 12 days."} {"chunk_id": "satellite_satellite-lake-water-level_completeness_q02__fe89a6eb3e9b", "report_id": "satellite_satellite-lake-water-level_completeness_q02", "dataset_id": "satellite-lake-water-level", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q02", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Assessing Lake Titicaca’s water levels in support of water management > Analysis and results > 2. Temporal completeness > Satellite missions over time", "title": "Assessing Lake Titicaca’s water levels in support of water management", "chunk_index": 6, "token_count": 643, "text_raw": "Define the satellite missions and their operational periods\nConvert the mission data into a DataFrame\nSort the DataFrame by start date\nCalculate durations in days\nCreate the figure and axes\nPlot the bars (same as before)\nCreate a range of years for ticks\nSet x-axis ticks to January of each year\nEnable grid for x-axis\nAxis labels and title\n\n*Figure 2. Span of the satellite missions used to estimate water levels in the C3S–LWL v5.0 dataset. The time periods reflect the operational availability of each mission as described in the Algorithm Theoretical Basis Document ([ATBD](https://confluence.ecmwf.int/pages/viewpage.action?pageId=425328942)). A correction was applied to the end year of the TOPEX/Poseidon mission to reflect its actual data availability.*\n\nAccording to the [PQAR](https://confluence.ecmwf.int/pages/viewpage.action?pageId=428248112), the median time step in the lake water level dataset decreases after 2016, indicating an increase in the number of recorded values. This improvement is attributed to the introduction of Sentinel-3A and Jason-3 missions, which provided more frequent overpasses with higher-quality sensors.\n\nFor Lake Titicaca, this enhancement is evident post-2016, as months with two records became more common. Following the launch of Sentinel-3B and Sentinel-6A, a further increase in data frequency was observed, with most months after late 2019 containing two records, and some months even having three or four (see *Figure 1*). This indicates a significant improvement in dataset completeness due to these newer missions.\n\nPrior to 2016, the dataset exhibits more gaps, with several months lacking data. This can be explained by the limitations of earlier missions: TOPEX/Poseidon, Jason-1, Envisat, Jason-2, and SARAL. These missions had less precise instruments, making their measurements more susceptible to filtering. Although satellite coverage and sensor quality improved after 2016, a few gaps—such as those seen in 2021—may still occur. The observations from the different satellite missions are processed in three sequential steps, each applying thresholds and filtering criteria to remove bad-quality data. As a result, only high-quality data are retained in the final dataset. Because older satellite sensors generally had lower accuracy, a larger fraction of early observations is filtered out, whereas later missions contribute more usable data. More in depth information on the processing steps to make the C3S-LWL v5.0 dataset can be found in the Algorithm Theoretical Basis Document [(ATBD)](https://confluence.ecmwf.int/pages/viewpage.action?pageId=425328942).\n\n(section-3)=", "text_with_prefix": "EQC Quality Assessment: \"Assessing Lake Titicaca’s water levels in support of water management\"\nDataset: satellite-lake-water-level [CDS]\nAspect: completeness_q02 | Category: Satellite_ECVs\nSection: Assessing Lake Titicaca’s water levels in support of water management > Analysis and results > 2. Temporal completeness > Satellite missions over time\n---\nDefine the satellite missions and their operational periods\nConvert the mission data into a DataFrame\nSort the DataFrame by start date\nCalculate durations in days\nCreate the figure and axes\nPlot the bars (same as before)\nCreate a range of years for ticks\nSet x-axis ticks to January of each year\nEnable grid for x-axis\nAxis labels and title\n\n*Figure 2. Span of the satellite missions used to estimate water levels in the C3S–LWL v5.0 dataset. The time periods reflect the operational availability of each mission as described in the Algorithm Theoretical Basis Document ([ATBD](https://confluence.ecmwf.int/pages/viewpage.action?pageId=425328942)). A correction was applied to the end year of the TOPEX/Poseidon mission to reflect its actual data availability.*\n\nAccording to the [PQAR](https://confluence.ecmwf.int/pages/viewpage.action?pageId=428248112), the median time step in the lake water level dataset decreases after 2016, indicating an increase in the number of recorded values. This improvement is attributed to the introduction of Sentinel-3A and Jason-3 missions, which provided more frequent overpasses with higher-quality sensors.\n\nFor Lake Titicaca, this enhancement is evident post-2016, as months with two records became more common. Following the launch of Sentinel-3B and Sentinel-6A, a further increase in data frequency was observed, with most months after late 2019 containing two records, and some months even having three or four (see *Figure 1*). This indicates a significant improvement in dataset completeness due to these newer missions.\n\nPrior to 2016, the dataset exhibits more gaps, with several months lacking data. This can be explained by the limitations of earlier missions: TOPEX/Poseidon, Jason-1, Envisat, Jason-2, and SARAL. These missions had less precise instruments, making their measurements more susceptible to filtering. Although satellite coverage and sensor quality improved after 2016, a few gaps—such as those seen in 2021—may still occur. The observations from the different satellite missions are processed in three sequential steps, each applying thresholds and filtering criteria to remove bad-quality data. As a result, only high-quality data are retained in the final dataset. Because older satellite sensors generally had lower accuracy, a larger fraction of early observations is filtered out, whereas later missions contribute more usable data. More in depth information on the processing steps to make the C3S-LWL v5.0 dataset can be found in the Algorithm Theoretical Basis Document [(ATBD)](https://confluence.ecmwf.int/pages/viewpage.action?pageId=425328942).\n\n(section-3)="} {"chunk_id": "satellite_satellite-lake-water-level_completeness_q02__2f8a5aa99d2b", "report_id": "satellite_satellite-lake-water-level_completeness_q02", "dataset_id": "satellite-lake-water-level", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q02", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Assessing Lake Titicaca’s water levels in support of water management > Analysis and results > 3. IWRM applications", "title": "Assessing Lake Titicaca’s water levels in support of water management", "chunk_index": 7, "token_count": 224, "text_raw": "Due to the irregular time steps, annual update frequency, and current absence of Interim Climate Data Records (ICDRs), the C3S–LWL v5.0 dataset is not well-suited for applications requiring high temporal resolution or real-time monitoring, such as flood early warning systems or rapid drought response mechanisms. With the upcoming release of ICDRs on the C3S platform, these limitations are likely to be overcome.\n\nThe Climate Data Records (CDRs) of the dataset have a long temporal span, making it valuable for applications that benefit from historical context and trend analysis. This includes assessing interannual variability, long-term climate influences, and understanding hydrological responses to large-scale drivers such as ENSO.", "text_with_prefix": "EQC Quality Assessment: \"Assessing Lake Titicaca’s water levels in support of water management\"\nDataset: satellite-lake-water-level [CDS]\nAspect: completeness_q02 | Category: Satellite_ECVs\nSection: Assessing Lake Titicaca’s water levels in support of water management > Analysis and results > 3. IWRM applications\n---\nDue to the irregular time steps, annual update frequency, and current absence of Interim Climate Data Records (ICDRs), the C3S–LWL v5.0 dataset is not well-suited for applications requiring high temporal resolution or real-time monitoring, such as flood early warning systems or rapid drought response mechanisms. With the upcoming release of ICDRs on the C3S platform, these limitations are likely to be overcome.\n\nThe Climate Data Records (CDRs) of the dataset have a long temporal span, making it valuable for applications that benefit from historical context and trend analysis. This includes assessing interannual variability, long-term climate influences, and understanding hydrological responses to large-scale drivers such as ENSO."} {"chunk_id": "satellite_satellite-lake-water-level_completeness_q02__ccb570e0d4ad", "report_id": "satellite_satellite-lake-water-level_completeness_q02", "dataset_id": "satellite-lake-water-level", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q02", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Assessing Lake Titicaca’s water levels in support of water management > Analysis and results > 3. IWRM applications > Detect seasonality and extreme events", "title": "Assessing Lake Titicaca’s water levels in support of water management", "chunk_index": 8, "token_count": 1003, "text_raw": "A time series of Lake Titicaca water levels was plotted, highlighting April and December values to illustrate seasonality, as well as the most severe historically reported extreme events.\n\nExtract April and December data\nConvert DataArray time coordinate to a pandas datetime index\n--- Identify extremes ---\nPlot time series\nAdd points for April and December with subtler colors\n--- Add highlighted extremes with safe annotations ---\nRemove duplicate labels in the legend\nForce x-axis to show every year\nRotate ticks if needed\n\n*Figure 3. Water level in the Titicaca Lake in the period 1995-2023 according to the C3S–LWL v5.0.*\n\nIn-situ water level data for Lake Titicaca are not openly available for direct download. Access to this information must be formally requested from the relevant governmental institutions in Peru and Bolivia. However, the Bolivian National Meteorological and Hydrological Service (SENAMHI) publishes a weekly bulletin that reports information on Lake Titicaca. The figure below, extracted from the SENAMHI bulletin of August 1, 2025, presents the time series of monthly average water levels recorded at the Huatajata Station [[4]](https://senamhi.gob.bo/meteorologia/boletines/hidrologico/LagoTiticaca/2025/Agosto/LAGO%20TITICACA%20-%20Informe%20Semanal%20%28Actualizado%2001_08_2025%29.pdf).\n\n![LAKE LEVEL SENAMHI.jpg](attachment:188a8952-b0db-4164-aa44-a20c221455d2.jpg)\n\n*Figure 4. Monthly averaged water level Lake Titicaca (1974-2025). Source: SENAMHI Bolivia (2025) [[4]](https://senamhi.gob.bo/meteorologia/boletines/hidrologico/LagoTiticaca/2025/Agosto/LAGO%20TITICACA%20-%20Informe%20Semanal%20%28Actualizado%2001_08_2025%29.pdf)*\n\nAn approximate comparison between the C3S-LWL data and the in-situ measurements was performed using Figures 3 and 4. To achieve this, the C3S-LWL figure was placed on top of SENAMHI’s figure with transparency, and the vertical and horizontal scales were adjusted to be identical. The time (x) axis was aligned to match SENAMHI’s figure, and the graph was then shifted vertically to best fit the most recent years of SENAMHI’s time series. This approach was necessary because the two datasets use different datums (approx. 1.5 m difference); since the most recent C3S data are considered more reliable, priority was given to visually aligning the later years. The result is presented in *Figure 5* below.\n\n![LAKE LEVEL COMPARISON.jpg](attachment:b97f0a2a-4d26-48e1-9152-fb6c459b49bd.jpg)\n\n*Figure 5. Overlap of SENAMHI's time series [[4]](https://senamhi.gob.bo/meteorologia/boletines/hidrologico/LagoTiticaca/2025/Agosto/LAGO%20TITICACA%20-%20Informe%20Semanal%20%28Actualizado%2001_08_2025%29.pdf) and Figure 3 using the C3S-LWL data.*\n\nIt can be observed in *Figure 5* that C3S–LWL values tend to be lower than those of SENAMHI prior to 2008. However, this observation is based on visual inspection only and cannot be corroborated numerically, as no quantitative comparison was applied on.\n\nConvert time to a PeriodIndex (monthly)\nSelect only the chosen year\nCount how many entries exist per month\nAll months for that year\nFind missing ones\n\n```text\nMissing months in 2011: ['January 2011', 'April 2011']\nMissing months in 2012: ['March 2012', 'April 2012', 'May 2012', 'August 2012', 'September 2012', 'December 2012']\nMissing months in 2013: ['April 2013', 'December 2013']\n```", "text_with_prefix": "EQC Quality Assessment: \"Assessing Lake Titicaca’s water levels in support of water management\"\nDataset: satellite-lake-water-level [CDS]\nAspect: completeness_q02 | Category: Satellite_ECVs\nSection: Assessing Lake Titicaca’s water levels in support of water management > Analysis and results > 3. IWRM applications > Detect seasonality and extreme events\n---\nA time series of Lake Titicaca water levels was plotted, highlighting April and December values to illustrate seasonality, as well as the most severe historically reported extreme events.\n\nExtract April and December data\nConvert DataArray time coordinate to a pandas datetime index\n--- Identify extremes ---\nPlot time series\nAdd points for April and December with subtler colors\n--- Add highlighted extremes with safe annotations ---\nRemove duplicate labels in the legend\nForce x-axis to show every year\nRotate ticks if needed\n\n*Figure 3. Water level in the Titicaca Lake in the period 1995-2023 according to the C3S–LWL v5.0.*\n\nIn-situ water level data for Lake Titicaca are not openly available for direct download. Access to this information must be formally requested from the relevant governmental institutions in Peru and Bolivia. However, the Bolivian National Meteorological and Hydrological Service (SENAMHI) publishes a weekly bulletin that reports information on Lake Titicaca. The figure below, extracted from the SENAMHI bulletin of August 1, 2025, presents the time series of monthly average water levels recorded at the Huatajata Station [[4]](https://senamhi.gob.bo/meteorologia/boletines/hidrologico/LagoTiticaca/2025/Agosto/LAGO%20TITICACA%20-%20Informe%20Semanal%20%28Actualizado%2001_08_2025%29.pdf).\n\n![LAKE LEVEL SENAMHI.jpg](attachment:188a8952-b0db-4164-aa44-a20c221455d2.jpg)\n\n*Figure 4. Monthly averaged water level Lake Titicaca (1974-2025). Source: SENAMHI Bolivia (2025) [[4]](https://senamhi.gob.bo/meteorologia/boletines/hidrologico/LagoTiticaca/2025/Agosto/LAGO%20TITICACA%20-%20Informe%20Semanal%20%28Actualizado%2001_08_2025%29.pdf)*\n\nAn approximate comparison between the C3S-LWL data and the in-situ measurements was performed using Figures 3 and 4. To achieve this, the C3S-LWL figure was placed on top of SENAMHI’s figure with transparency, and the vertical and horizontal scales were adjusted to be identical. The time (x) axis was aligned to match SENAMHI’s figure, and the graph was then shifted vertically to best fit the most recent years of SENAMHI’s time series. This approach was necessary because the two datasets use different datums (approx. 1.5 m difference); since the most recent C3S data are considered more reliable, priority was given to visually aligning the later years. The result is presented in *Figure 5* below.\n\n![LAKE LEVEL COMPARISON.jpg](attachment:b97f0a2a-4d26-48e1-9152-fb6c459b49bd.jpg)\n\n*Figure 5. Overlap of SENAMHI's time series [[4]](https://senamhi.gob.bo/meteorologia/boletines/hidrologico/LagoTiticaca/2025/Agosto/LAGO%20TITICACA%20-%20Informe%20Semanal%20%28Actualizado%2001_08_2025%29.pdf) and Figure 3 using the C3S-LWL data.*\n\nIt can be observed in *Figure 5* that C3S–LWL values tend to be lower than those of SENAMHI prior to 2008. However, this observation is based on visual inspection only and cannot be corroborated numerically, as no quantitative comparison was applied on.\n\nConvert time to a PeriodIndex (monthly)\nSelect only the chosen year\nCount how many entries exist per month\nAll months for that year\nFind missing ones\n\n```text\nMissing months in 2011: ['January 2011', 'April 2011']\nMissing months in 2012: ['March 2012', 'April 2012', 'May 2012', 'August 2012', 'September 2012', 'December 2012']\nMissing months in 2013: ['April 2013', 'December 2013']\n```"} {"chunk_id": "satellite_satellite-lake-water-level_completeness_q02__c4e042fda149", "report_id": "satellite_satellite-lake-water-level_completeness_q02", "dataset_id": "satellite-lake-water-level", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q02", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Assessing Lake Titicaca’s water levels in support of water management > Analysis and results > 3. IWRM applications > Detect seasonality and extreme events", "title": "Assessing Lake Titicaca’s water levels in support of water management", "chunk_index": 9, "token_count": 949, "text_raw": "', 'May 2012', 'August 2012', 'September 2012', 'December 2012']\nMissing months in 2013: ['April 2013', 'December 2013']\n```\n\n**Seasonality:** The annual cycle of rainfall over the TDPS system exhibits a marked seasonality, with a rainy season from December to March that accounts for over 75% of the annual precipitation, a dry season from May to August, and a pre-wet season from September to December [[3]](https://doi.org/10.1016/j.wace.2024.100710). Lake Titicaca's level is primarily influenced by variations in precipitation and high evaporation rates [[5]](https://doi.org/10.5194/hess-29-655-2025). Consequently, December (just before the onset of the rainy season) typically corresponds to the lake’s lowest water level in the hydrological year, while the highest level usually occurs around April, at the end of the rainy season and the start of the dry season. *Figure 3* shows that this seasonality is captured on years with data for April and December. Years with more complete records capture this seasonal cycle more reliably. The largest discrepancies between SENAMHI and C3S-LWL occur in 1996, likely due to lower-quality measurement instruments, and in 2011 and 2012, probably because many months lack observations in the C3S-LWL dataset (see *Figure 5*). In 2013, although April and December are missing, the remaining months still reproduce the seasonal pattern well. Overall, C3S-LWL captures the seasonality of Lake Titicaca effectively, except in years with extended gaps in observations, such as 2012.\n\n**Floods of 2003 and 2004:** The C3S-LWL timeseries reflects the intense precipitation events that occurred in 2003 and 2004 (see *Figure 5*). At the beginning of 2003, extreme rainfall caused Lake Titicaca’s water level to rise significantly. By the end of January 2003, the Peruvian government declared a two-month state of emergency to support affected communities. The most impacted areas in Puno (the region where the Peruvian side of Lake Titicaca is located) included Huancané, Putina, Azángaro, Asillo, and San Antón. By late January, seven deaths and severe damage to infrastructure had been reported [[6]](https://reliefweb.int/report/peru/act-peru-12003-heavy-rainfall) [[7]](https://reliefweb.int/report/peru/flash-floods-leave-peruvian-farmers-devastated).\n\nAccording to the National Institute of Statistics and Informatics of Peru (INEI), the mean inflow from tributary rivers into Lake Titicaca in January and February 2004 was 158.2% and 33.6% higher than the long-term average, respectively. These surges caused flooding in surrounding areas of the lake (Table 25 in [[8]](https://www.inei.gob.pe/media/MenuRecursivo/boletines/5340.pdf)).\n\n**Limitations in flood detection:** The C3S–LWL dataset contains only one observation per month for most months prior to late 2019. Therefore, although the flooding events of 2003 and 2004 are visible in the peaks of *Figure 3*, short-lived flash floods that cause rapid water-level changes over shorter time windows are not captured. Similarly, SENAMHI’s data in *Figure 4* consist of monthly averages, which may also smooth out short-duration peaks. This averaging provides a clearer view of long-term lake levels with reduced noise but limits the ability to detect flash floods. The peaks observed in April 2003 and 2004 reflect the cumulative effect of increased inflows during the entire rainy season, whereas shorter-lived flooding events triggered by isolated extreme precipitation are likely underrepresented. A more in-depth analysis of extremes would require data with higher temporal resolution, which is beyond the scope of this study.", "text_with_prefix": "EQC Quality Assessment: \"Assessing Lake Titicaca’s water levels in support of water management\"\nDataset: satellite-lake-water-level [CDS]\nAspect: completeness_q02 | Category: Satellite_ECVs\nSection: Assessing Lake Titicaca’s water levels in support of water management > Analysis and results > 3. IWRM applications > Detect seasonality and extreme events\n---\n', 'May 2012', 'August 2012', 'September 2012', 'December 2012']\nMissing months in 2013: ['April 2013', 'December 2013']\n```\n\n**Seasonality:** The annual cycle of rainfall over the TDPS system exhibits a marked seasonality, with a rainy season from December to March that accounts for over 75% of the annual precipitation, a dry season from May to August, and a pre-wet season from September to December [[3]](https://doi.org/10.1016/j.wace.2024.100710). Lake Titicaca's level is primarily influenced by variations in precipitation and high evaporation rates [[5]](https://doi.org/10.5194/hess-29-655-2025). Consequently, December (just before the onset of the rainy season) typically corresponds to the lake’s lowest water level in the hydrological year, while the highest level usually occurs around April, at the end of the rainy season and the start of the dry season. *Figure 3* shows that this seasonality is captured on years with data for April and December. Years with more complete records capture this seasonal cycle more reliably. The largest discrepancies between SENAMHI and C3S-LWL occur in 1996, likely due to lower-quality measurement instruments, and in 2011 and 2012, probably because many months lack observations in the C3S-LWL dataset (see *Figure 5*). In 2013, although April and December are missing, the remaining months still reproduce the seasonal pattern well. Overall, C3S-LWL captures the seasonality of Lake Titicaca effectively, except in years with extended gaps in observations, such as 2012.\n\n**Floods of 2003 and 2004:** The C3S-LWL timeseries reflects the intense precipitation events that occurred in 2003 and 2004 (see *Figure 5*). At the beginning of 2003, extreme rainfall caused Lake Titicaca’s water level to rise significantly. By the end of January 2003, the Peruvian government declared a two-month state of emergency to support affected communities. The most impacted areas in Puno (the region where the Peruvian side of Lake Titicaca is located) included Huancané, Putina, Azángaro, Asillo, and San Antón. By late January, seven deaths and severe damage to infrastructure had been reported [[6]](https://reliefweb.int/report/peru/act-peru-12003-heavy-rainfall) [[7]](https://reliefweb.int/report/peru/flash-floods-leave-peruvian-farmers-devastated).\n\nAccording to the National Institute of Statistics and Informatics of Peru (INEI), the mean inflow from tributary rivers into Lake Titicaca in January and February 2004 was 158.2% and 33.6% higher than the long-term average, respectively. These surges caused flooding in surrounding areas of the lake (Table 25 in [[8]](https://www.inei.gob.pe/media/MenuRecursivo/boletines/5340.pdf)).\n\n**Limitations in flood detection:** The C3S–LWL dataset contains only one observation per month for most months prior to late 2019. Therefore, although the flooding events of 2003 and 2004 are visible in the peaks of *Figure 3*, short-lived flash floods that cause rapid water-level changes over shorter time windows are not captured. Similarly, SENAMHI’s data in *Figure 4* consist of monthly averages, which may also smooth out short-duration peaks. This averaging provides a clearer view of long-term lake levels with reduced noise but limits the ability to detect flash floods. The peaks observed in April 2003 and 2004 reflect the cumulative effect of increased inflows during the entire rainy season, whereas shorter-lived flooding events triggered by isolated extreme precipitation are likely underrepresented. A more in-depth analysis of extremes would require data with higher temporal resolution, which is beyond the scope of this study."} {"chunk_id": "satellite_satellite-lake-water-level_completeness_q02__bb971f7da7f4", "report_id": "satellite_satellite-lake-water-level_completeness_q02", "dataset_id": "satellite-lake-water-level", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q02", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Assessing Lake Titicaca’s water levels in support of water management > Analysis and results > 3. IWRM applications > Detect seasonality and extreme events", "title": "Assessing Lake Titicaca’s water levels in support of water management", "chunk_index": 10, "token_count": 317, "text_raw": ", whereas shorter-lived flooding events triggered by isolated extreme precipitation are likely underrepresented. A more in-depth analysis of extremes would require data with higher temporal resolution, which is beyond the scope of this study.\n\n**Drought of 2023:** Conversely, the lowest point on the dataset reflects the severe drought experienced in 2023 in the TDPS, which led to a substantial drop in lake levels, reaching a minimum by the end of the year [[9]](https://www.copernicus.eu/fr/node/40141). This event had system-wide consequences, in Puno (Peru) and La Paz (Bolivia), up to 80% of potato and sweet potato and 90% of Andean grains were lost due to water scarcity [[3]](https://doi.org/10.1016/j.wace.2024.100710). From an IWRM perspective, capturing and understanding these extremes is critical for guiding investment in adaptive infrastructure and preparing multisectoral response strategies. It also underscores the need for transboundary coordination, as hydrological impacts in one part of the system inevitably propagate throughout the TDPS basin.", "text_with_prefix": "EQC Quality Assessment: \"Assessing Lake Titicaca’s water levels in support of water management\"\nDataset: satellite-lake-water-level [CDS]\nAspect: completeness_q02 | Category: Satellite_ECVs\nSection: Assessing Lake Titicaca’s water levels in support of water management > Analysis and results > 3. IWRM applications > Detect seasonality and extreme events\n---\n, whereas shorter-lived flooding events triggered by isolated extreme precipitation are likely underrepresented. A more in-depth analysis of extremes would require data with higher temporal resolution, which is beyond the scope of this study.\n\n**Drought of 2023:** Conversely, the lowest point on the dataset reflects the severe drought experienced in 2023 in the TDPS, which led to a substantial drop in lake levels, reaching a minimum by the end of the year [[9]](https://www.copernicus.eu/fr/node/40141). This event had system-wide consequences, in Puno (Peru) and La Paz (Bolivia), up to 80% of potato and sweet potato and 90% of Andean grains were lost due to water scarcity [[3]](https://doi.org/10.1016/j.wace.2024.100710). From an IWRM perspective, capturing and understanding these extremes is critical for guiding investment in adaptive infrastructure and preparing multisectoral response strategies. It also underscores the need for transboundary coordination, as hydrological impacts in one part of the system inevitably propagate throughout the TDPS basin."} {"chunk_id": "satellite_satellite-lake-water-level_completeness_q02__a4614817f2b0", "report_id": "satellite_satellite-lake-water-level_completeness_q02", "dataset_id": "satellite-lake-water-level", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q02", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Assessing Lake Titicaca’s water levels in support of water management > Analysis and results > 3. IWRM applications > Relation to ENSO", "title": "Assessing Lake Titicaca’s water levels in support of water management", "chunk_index": 11, "token_count": 1034, "text_raw": "In the TDPS region, El Niño events are typically associated with drier conditions, while La Niña events often bring increased rainfall, particularly during the wet season (December-March). These precipitation patterns directly influence Lake Titicaca's hydrology. El Niño years tend to reduce inflows and lower lake levels, whereas La Niña years usually promote higher inflows and rising water levels. However, the 2022-2023 drought illustrates that La Niña can also be linked to severe dry conditions, particularly during the pre-wet season (September-November). Unlike its wet-season effects, La Niña during this early period is associated with decreased precipitation, which can delay the onset of the rainy season and heighten the risk of drought. Gutierrez-Villarreal et al. (2024) explain that the 2022-23 pre-wet season was one of the driest on record, with November 2022 marking the driest month in 58 years. As a result, Lake Titicaca’s seasonal rise was significantly suppressed, its water level increased by only 0.09 meters between December and April, one of the weakest rises since 1940. Notably, the lake level even declined between December and January, a rare occurrence during a period when levels typically rise [[3]](https://doi.org/10.1016/j.wace.2024.100710).\n\n![Gutierrez2024.jpg](attachment:6ef06ea7-4202-4a54-9d6d-05c3936c0542.jpg)\n\n*Figure 6. Lake Titicaca's geographical location and hydrograms. a) The Lake Titicaca, Desaguadero River, and Lake Poopó basins (TDPS system) boundary is delimited by a black line, and the lake area is represented by a black polygon. Red lines in the Andean cordillera denote altitudes of 3 000 m.a.s.l. Interannual increment of lake levels of b) April, c) February, and d) January, with respect to December. The dataset spans between 1914 and 2023. The purple horizontal lines represent the long-term (1914–2023) average of lake level differences. “n” refers to the calendar year. Selected El Niño (La Niña) years are tagged with red (blue) bars if the [El Niño 3.4 index from the HadiSST](https://psl.noaa.gov/data/timeseries/month/Nino34/) [[10]](https://doi.org/10.1029/2002jd002670). dataset is higher (lower) than 0.75 (−0.75) during at least 4 months between October n-1 and April n. Reproduced from Gutierrez-Villarreal et al. (2024), licensed under CC BY 4 [[3]](https://doi.org/10.1016/j.wace.2024.100710).*\n\nThe water levels used by Gutierrez-Villarreal et al. (2024) to make *Figure 6* are monthly lake level data from 1914 to 2023, measured at the Muelle Enafer station (3,800 m a.s.l.) on the Peruvian side of Lake Titicaca, and provided by Peru’s National Meteorology and Hydrology Service (SENAMHI) [[3]](https://doi.org/10.1016/j.wace.2024.100710). The figure illustrates how water levels in Lake Titicaca can be a valuable indicator for analysing the relationship between ENSO phases and extreme events in the TDPS system, an insight that is highly relevant for IWRM due to its implications for drought preparedness and long-term planning. A similar analysis was conducted here using the C3S–LWL v5.0 dataset, to assess whether the patterns and anomalies described in the original study can also be detected with this alternative data source.\n\nLoad Niño 3.4 SST Index from the HadISST1.1 (https://psl.noaa.gov/data/timeseries/month/Nino34/)\n\nExtract April and December data\n\nRename ENSO column\nFunction to compute differences and ENSO-colored bar data\nMean per year\nCalculate difference\nCompute ENSO-based bar colors\nIdentify missing years\nCompute all three\nOnly for the April plot, add marker and text\nHorizontal line from y-axis to the bar at y_val\nAdd text above the 2023 bar\nShared x-label", "text_with_prefix": "EQC Quality Assessment: \"Assessing Lake Titicaca’s water levels in support of water management\"\nDataset: satellite-lake-water-level [CDS]\nAspect: completeness_q02 | Category: Satellite_ECVs\nSection: Assessing Lake Titicaca’s water levels in support of water management > Analysis and results > 3. IWRM applications > Relation to ENSO\n---\nIn the TDPS region, El Niño events are typically associated with drier conditions, while La Niña events often bring increased rainfall, particularly during the wet season (December-March). These precipitation patterns directly influence Lake Titicaca's hydrology. El Niño years tend to reduce inflows and lower lake levels, whereas La Niña years usually promote higher inflows and rising water levels. However, the 2022-2023 drought illustrates that La Niña can also be linked to severe dry conditions, particularly during the pre-wet season (September-November). Unlike its wet-season effects, La Niña during this early period is associated with decreased precipitation, which can delay the onset of the rainy season and heighten the risk of drought. Gutierrez-Villarreal et al. (2024) explain that the 2022-23 pre-wet season was one of the driest on record, with November 2022 marking the driest month in 58 years. As a result, Lake Titicaca’s seasonal rise was significantly suppressed, its water level increased by only 0.09 meters between December and April, one of the weakest rises since 1940. Notably, the lake level even declined between December and January, a rare occurrence during a period when levels typically rise [[3]](https://doi.org/10.1016/j.wace.2024.100710).\n\n![Gutierrez2024.jpg](attachment:6ef06ea7-4202-4a54-9d6d-05c3936c0542.jpg)\n\n*Figure 6. Lake Titicaca's geographical location and hydrograms. a) The Lake Titicaca, Desaguadero River, and Lake Poopó basins (TDPS system) boundary is delimited by a black line, and the lake area is represented by a black polygon. Red lines in the Andean cordillera denote altitudes of 3 000 m.a.s.l. Interannual increment of lake levels of b) April, c) February, and d) January, with respect to December. The dataset spans between 1914 and 2023. The purple horizontal lines represent the long-term (1914–2023) average of lake level differences. “n” refers to the calendar year. Selected El Niño (La Niña) years are tagged with red (blue) bars if the [El Niño 3.4 index from the HadiSST](https://psl.noaa.gov/data/timeseries/month/Nino34/) [[10]](https://doi.org/10.1029/2002jd002670). dataset is higher (lower) than 0.75 (−0.75) during at least 4 months between October n-1 and April n. Reproduced from Gutierrez-Villarreal et al. (2024), licensed under CC BY 4 [[3]](https://doi.org/10.1016/j.wace.2024.100710).*\n\nThe water levels used by Gutierrez-Villarreal et al. (2024) to make *Figure 6* are monthly lake level data from 1914 to 2023, measured at the Muelle Enafer station (3,800 m a.s.l.) on the Peruvian side of Lake Titicaca, and provided by Peru’s National Meteorology and Hydrology Service (SENAMHI) [[3]](https://doi.org/10.1016/j.wace.2024.100710). The figure illustrates how water levels in Lake Titicaca can be a valuable indicator for analysing the relationship between ENSO phases and extreme events in the TDPS system, an insight that is highly relevant for IWRM due to its implications for drought preparedness and long-term planning. A similar analysis was conducted here using the C3S–LWL v5.0 dataset, to assess whether the patterns and anomalies described in the original study can also be detected with this alternative data source.\n\nLoad Niño 3.4 SST Index from the HadISST1.1 (https://psl.noaa.gov/data/timeseries/month/Nino34/)\n\nExtract April and December data\n\nRename ENSO column\nFunction to compute differences and ENSO-colored bar data\nMean per year\nCalculate difference\nCompute ENSO-based bar colors\nIdentify missing years\nCompute all three\nOnly for the April plot, add marker and text\nHorizontal line from y-axis to the bar at y_val\nAdd text above the 2023 bar\nShared x-label"} {"chunk_id": "satellite_satellite-lake-water-level_completeness_q02__bf986c9e808c", "report_id": "satellite_satellite-lake-water-level_completeness_q02", "dataset_id": "satellite-lake-water-level", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q02", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Assessing Lake Titicaca’s water levels in support of water management > Analysis and results > 3. IWRM applications > Relation to ENSO", "title": "Assessing Lake Titicaca’s water levels in support of water management", "chunk_index": 12, "token_count": 871, "text_raw": "bar colors\nIdentify missing years\nCompute all three\nOnly for the April plot, add marker and text\nHorizontal line from y-axis to the bar at y_val\nAdd text above the 2023 bar\nShared x-label\n\n*Figure 7. Yearly differences between water levels of the Titicaca Lake form the C3S-LWL v5.0 dataset in April, February, and January of year *n* and December of year *n-1*, for the period 1995-2023. Bars are color-coded based on ENSO conditions between October *n-1* and April *n*: red for El Niño years (≥ 4 months with Niño 3.4 index > 0.75), blue for La Niña years (≥ 4 months with index < −0.75), and gray for neutral years. Missing bars indicate years where data was unavailable for the respective months.*\n\nPrint missing years for each\n\n```text\nMissing years (no data for Apr(n) or Dec(n-1)): [1995, 1999, 2000, 2011, 2012, 2013, 2014]\nMissing years (no data for Feb(n) or Dec(n-1)): [1995, 1996, 2000, 2007, 2011, 2013, 2014, 2022]\nMissing years (no data for Jan(n) or Dec(n-1)): [1995, 1998, 2000, 2011, 2013, 2014]\n```\n\nIn the work of Gutierrez-Villarreal et al. (2024) (see *Figure 6*), the blue bars representing La Niña years are generally taller than the red bars representing El Niño years. This suggests that water level gains during the wet season tend to be greater during La Niña years. Such a trend is clearly visible thanks to the extensive temporal coverage of the SENAMHI dataset, which spans more than 100 years. In contrast, this pattern is not noticeable on *Figure 7*, which uses the C3S-LWL v5.0 dataset. This is likely due to the shorter time span of the C3S dataset and the presence of data gaps, resulting in missing values for several years.\n\nAdditionally, some inconsistencies between the two figures can be noted. For instance, in 2012, *Figure 7* shows slightly negative differences for both Jan(*n*)-Dec(*n-1*) and Feb(*n*)-Dec(*n-1*), while according to the SENAMHI dataset (see *Figure 6*) these differences are clearly positive. This discrepancy indicates that the C3S-LWL v5.0 dataset may not be sufficiently accurate for these types of analyses, especially prior to 2016. As mentioned earlier, satellite instrument quality and measurement accuracy have improved significantly since 2016, which may explain some of the observed inconsistencies.\n\nDespite these limitations, the C3S-LWL v5.0 dataset successfully captures the small negative difference in Jan(*n*)-Dec(*n-1*) for the year 2023, as well as the anomalously weak water level increase during the 2022-2023 wet season (see *Figure 7*). However, according to the C3S-LWL v5.0 dataset, the rise was 0.17 m, while the SENAMHI dataset recorded a lower value of 0.09 m. Although not identical, this similarity suggests that the C3S-LWL v5.0 dataset has some utility in identifying recent anomalies, even if it is not yet robust enough for long-term ENSO-related hydrological analysis in Lake Titicaca.\n\n(section-4)=", "text_with_prefix": "EQC Quality Assessment: \"Assessing Lake Titicaca’s water levels in support of water management\"\nDataset: satellite-lake-water-level [CDS]\nAspect: completeness_q02 | Category: Satellite_ECVs\nSection: Assessing Lake Titicaca’s water levels in support of water management > Analysis and results > 3. IWRM applications > Relation to ENSO\n---\nbar colors\nIdentify missing years\nCompute all three\nOnly for the April plot, add marker and text\nHorizontal line from y-axis to the bar at y_val\nAdd text above the 2023 bar\nShared x-label\n\n*Figure 7. Yearly differences between water levels of the Titicaca Lake form the C3S-LWL v5.0 dataset in April, February, and January of year *n* and December of year *n-1*, for the period 1995-2023. Bars are color-coded based on ENSO conditions between October *n-1* and April *n*: red for El Niño years (≥ 4 months with Niño 3.4 index > 0.75), blue for La Niña years (≥ 4 months with index < −0.75), and gray for neutral years. Missing bars indicate years where data was unavailable for the respective months.*\n\nPrint missing years for each\n\n```text\nMissing years (no data for Apr(n) or Dec(n-1)): [1995, 1999, 2000, 2011, 2012, 2013, 2014]\nMissing years (no data for Feb(n) or Dec(n-1)): [1995, 1996, 2000, 2007, 2011, 2013, 2014, 2022]\nMissing years (no data for Jan(n) or Dec(n-1)): [1995, 1998, 2000, 2011, 2013, 2014]\n```\n\nIn the work of Gutierrez-Villarreal et al. (2024) (see *Figure 6*), the blue bars representing La Niña years are generally taller than the red bars representing El Niño years. This suggests that water level gains during the wet season tend to be greater during La Niña years. Such a trend is clearly visible thanks to the extensive temporal coverage of the SENAMHI dataset, which spans more than 100 years. In contrast, this pattern is not noticeable on *Figure 7*, which uses the C3S-LWL v5.0 dataset. This is likely due to the shorter time span of the C3S dataset and the presence of data gaps, resulting in missing values for several years.\n\nAdditionally, some inconsistencies between the two figures can be noted. For instance, in 2012, *Figure 7* shows slightly negative differences for both Jan(*n*)-Dec(*n-1*) and Feb(*n*)-Dec(*n-1*), while according to the SENAMHI dataset (see *Figure 6*) these differences are clearly positive. This discrepancy indicates that the C3S-LWL v5.0 dataset may not be sufficiently accurate for these types of analyses, especially prior to 2016. As mentioned earlier, satellite instrument quality and measurement accuracy have improved significantly since 2016, which may explain some of the observed inconsistencies.\n\nDespite these limitations, the C3S-LWL v5.0 dataset successfully captures the small negative difference in Jan(*n*)-Dec(*n-1*) for the year 2023, as well as the anomalously weak water level increase during the 2022-2023 wet season (see *Figure 7*). However, according to the C3S-LWL v5.0 dataset, the rise was 0.17 m, while the SENAMHI dataset recorded a lower value of 0.09 m. Although not identical, this similarity suggests that the C3S-LWL v5.0 dataset has some utility in identifying recent anomalies, even if it is not yet robust enough for long-term ENSO-related hydrological analysis in Lake Titicaca.\n\n(section-4)="} {"chunk_id": "satellite_satellite-lake-water-level_completeness_q02__8cdb193fc567", "report_id": "satellite_satellite-lake-water-level_completeness_q02", "dataset_id": "satellite-lake-water-level", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q02", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Assessing Lake Titicaca’s water levels in support of water management > Analysis and results > 4. Applicability for other lakes", "title": "Assessing Lake Titicaca’s water levels in support of water management", "chunk_index": 13, "token_count": 1035, "text_raw": "The values from the table presented in *Annex A. LWL Performance indicators* of the [PQAR](https://confluence.ecmwf.int/pages/viewpage.action?pageId=428248112) were used for the analysis in this section.\n\nLoad PQAR table (ANNEX A)\n\n```text\nLake name Dispersion 1992-2023 (cm) \\\n0 Albert 8.0 \n1 Bagre 20.0 \n2 Bankim 39.0 \n3 Bogoria 13.0 \n4 Fitri 10.0 \n.. ... ... \n246 Sevan 6.0 \n247 Srisailam 68.0 \n248 Tharthar 3.0 \n249 Toktogul 14.0 \n250 Van 6.0\n\nHigh Frequency variation 1992-2023 (cm) \\\n0 3.32 \n1 8.65 \n2 11.03 \n3 0.26 \n4 0.32 \n.. ... \n246 3.40 \n247 12.74 \n248 18.20 \n249 0.00 \n250 3.39\n\nMedian Timestep 1992-2023 (days) Max Timestep 1992-2023 (days) \\\n0 26.47 76.81 \n1 9.98 99.01 \n2 9.41 50.84 \n3 27.00 55.78 \n4 27.00 70.00 \n.. ... ... \n246 25.55 124.39 \n247 27.00 27.00 \n248 9.92 63.87 \n249 34.44 1006.73 \n250 12.46 73.00\n\nTimeseries duration 1992-2023 Dispersion 2014-2023 (cm) \\\n0 28.5 6.0 \n1 15.3 18.0 \n2 15.5 9.0 \n3 7.8 13.0 \n4 10.6 10.0 \n.. ... ... \n246 28.5 5.0 \n247 4.8 68.0 \n248 31.2 2.0 \n249 28.5 8.0 \n250 28.6 5.0\n\nHigh Frequency variation 2014-2023 (cm) \\\n0 3.24 \n1 8.71 \n2 12.29 \n3 0.26 \n4 0.33 \n.. ... \n246 3.25 \n247 12.74 \n248 24.16 \n249 2.86 \n250 3.03\n\nMedian Timestep 2014-2023 (days) Max Timestep 2014-2023 (days) \n0 13.00 76.81 \n1 9.98 89.25 \n2 9.31 50.84 \n3 27.00 55.78 \n4 27.00 70.00 \n.. ... ... \n246 6.54 93.70 \n247 27.00 27.00 \n248 9.92 29.75 \n249 27.00 951.49 \n250 10.00 39.43\n\n[251 rows x 10 columns]\n```\n\nDefine the columns of interest\nGet the row for Lake Titicaca\nCompute percentile ranks for Titicaca\nCreate a DataFrame for display\n\n```text\nTiticaca Percentile\nMedian Timestep 1992-2023 (days) 95.22\nMax Timestep 1992-2023 (days) 76.89\nTimeseries duration 1992-2023 66.53\nMedian Timestep 2014-2023 (days) 45.42\nMax Timestep 2014-2023 (days) 73.31\n```", "text_with_prefix": "EQC Quality Assessment: \"Assessing Lake Titicaca’s water levels in support of water management\"\nDataset: satellite-lake-water-level [CDS]\nAspect: completeness_q02 | Category: Satellite_ECVs\nSection: Assessing Lake Titicaca’s water levels in support of water management > Analysis and results > 4. Applicability for other lakes\n---\nThe values from the table presented in *Annex A. LWL Performance indicators* of the [PQAR](https://confluence.ecmwf.int/pages/viewpage.action?pageId=428248112) were used for the analysis in this section.\n\nLoad PQAR table (ANNEX A)\n\n```text\nLake name Dispersion 1992-2023 (cm) \\\n0 Albert 8.0 \n1 Bagre 20.0 \n2 Bankim 39.0 \n3 Bogoria 13.0 \n4 Fitri 10.0 \n.. ... ... \n246 Sevan 6.0 \n247 Srisailam 68.0 \n248 Tharthar 3.0 \n249 Toktogul 14.0 \n250 Van 6.0\n\nHigh Frequency variation 1992-2023 (cm) \\\n0 3.32 \n1 8.65 \n2 11.03 \n3 0.26 \n4 0.32 \n.. ... \n246 3.40 \n247 12.74 \n248 18.20 \n249 0.00 \n250 3.39\n\nMedian Timestep 1992-2023 (days) Max Timestep 1992-2023 (days) \\\n0 26.47 76.81 \n1 9.98 99.01 \n2 9.41 50.84 \n3 27.00 55.78 \n4 27.00 70.00 \n.. ... ... \n246 25.55 124.39 \n247 27.00 27.00 \n248 9.92 63.87 \n249 34.44 1006.73 \n250 12.46 73.00\n\nTimeseries duration 1992-2023 Dispersion 2014-2023 (cm) \\\n0 28.5 6.0 \n1 15.3 18.0 \n2 15.5 9.0 \n3 7.8 13.0 \n4 10.6 10.0 \n.. ... ... \n246 28.5 5.0 \n247 4.8 68.0 \n248 31.2 2.0 \n249 28.5 8.0 \n250 28.6 5.0\n\nHigh Frequency variation 2014-2023 (cm) \\\n0 3.24 \n1 8.71 \n2 12.29 \n3 0.26 \n4 0.33 \n.. ... \n246 3.25 \n247 12.74 \n248 24.16 \n249 2.86 \n250 3.03\n\nMedian Timestep 2014-2023 (days) Max Timestep 2014-2023 (days) \n0 13.00 76.81 \n1 9.98 89.25 \n2 9.31 50.84 \n3 27.00 55.78 \n4 27.00 70.00 \n.. ... ... \n246 6.54 93.70 \n247 27.00 27.00 \n248 9.92 29.75 \n249 27.00 951.49 \n250 10.00 39.43\n\n[251 rows x 10 columns]\n```\n\nDefine the columns of interest\nGet the row for Lake Titicaca\nCompute percentile ranks for Titicaca\nCreate a DataFrame for display\n\n```text\nTiticaca Percentile\nMedian Timestep 1992-2023 (days) 95.22\nMax Timestep 1992-2023 (days) 76.89\nTimeseries duration 1992-2023 66.53\nMedian Timestep 2014-2023 (days) 45.42\nMax Timestep 2014-2023 (days) 73.31\n```"} {"chunk_id": "satellite_satellite-lake-water-level_completeness_q02__cd6dc6e5933f", "report_id": "satellite_satellite-lake-water-level_completeness_q02", "dataset_id": "satellite-lake-water-level", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q02", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Assessing Lake Titicaca’s water levels in support of water management > Analysis and results > 4. Applicability for other lakes", "title": "Assessing Lake Titicaca’s water levels in support of water management", "chunk_index": 14, "token_count": 445, "text_raw": "2-2023 66.53\nMedian Timestep 2014-2023 (days) 45.42\nMax Timestep 2014-2023 (days) 73.31\n```\n\nSince the dataset has different median timestep, maximum timestep, and timeseries duration for each lake, it is recommended to consult Annex A: LWL Performance Indicators in the [PQAR](https://confluence.ecmwf.int/pages/viewpage.action?pageId=428248112) for further interpretation when it comes to suitability for IWRM applications.\n\nFor Lake Titicaca, the median timestep for the full period (1992–2023) is among the highest, falling in the 95th percentile. This indicates that observations are relatively sparse, many other lakes have more frequent data. In the more recent period (2014–2023), Titicaca's median timestep falls near the 45th percentile, suggesting that data coverage has improved and is closer to the dataset median.\n\nThe maximum timestep (which reflects the size of the largest gap in the timeseries) is in the 77th percentile for the full period, indicating that Titicaca has experienced longer gaps than most lakes. Additionally, the timeseries duration places Titicaca in the 67th percentile, meaning it has longer data coverage than average, but still not among the very longest.\n\nThis implies that if Lake Titicaca is considered suitable for applications like extreme event detection and seasonality analysis, then most other lakes in the dataset—many of which have higher temporal resolution and fewer gaps—should be equally or more suitable. However, users are encouraged to consult Annex A or rerun the code from previous sections to assess whether the temporal completeness of each individual lake meets their specific requirements.", "text_with_prefix": "EQC Quality Assessment: \"Assessing Lake Titicaca’s water levels in support of water management\"\nDataset: satellite-lake-water-level [CDS]\nAspect: completeness_q02 | Category: Satellite_ECVs\nSection: Assessing Lake Titicaca’s water levels in support of water management > Analysis and results > 4. Applicability for other lakes\n---\n2-2023 66.53\nMedian Timestep 2014-2023 (days) 45.42\nMax Timestep 2014-2023 (days) 73.31\n```\n\nSince the dataset has different median timestep, maximum timestep, and timeseries duration for each lake, it is recommended to consult Annex A: LWL Performance Indicators in the [PQAR](https://confluence.ecmwf.int/pages/viewpage.action?pageId=428248112) for further interpretation when it comes to suitability for IWRM applications.\n\nFor Lake Titicaca, the median timestep for the full period (1992–2023) is among the highest, falling in the 95th percentile. This indicates that observations are relatively sparse, many other lakes have more frequent data. In the more recent period (2014–2023), Titicaca's median timestep falls near the 45th percentile, suggesting that data coverage has improved and is closer to the dataset median.\n\nThe maximum timestep (which reflects the size of the largest gap in the timeseries) is in the 77th percentile for the full period, indicating that Titicaca has experienced longer gaps than most lakes. Additionally, the timeseries duration places Titicaca in the 67th percentile, meaning it has longer data coverage than average, but still not among the very longest.\n\nThis implies that if Lake Titicaca is considered suitable for applications like extreme event detection and seasonality analysis, then most other lakes in the dataset—many of which have higher temporal resolution and fewer gaps—should be equally or more suitable. However, users are encouraged to consult Annex A or rerun the code from previous sections to assess whether the temporal completeness of each individual lake meets their specific requirements."} {"chunk_id": "satellite_satellite-lake-water-level_completeness_q02__73544fb449b1", "report_id": "satellite_satellite-lake-water-level_completeness_q02", "dataset_id": "satellite-lake-water-level", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q02", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Assessing Lake Titicaca’s water levels in support of water management > ℹ️ If you want to know more", "title": "Assessing Lake Titicaca’s water levels in support of water management", "chunk_index": 15, "token_count": 284, "text_raw": "* Autonomous Binational Authority of the Lake Titicaca, Desaguadero River, Lake Poopó, and Salar de Coipasa Hydrological System (ALT): [Publications IWRM-TDPS](https://alt-perubolivia.org/?page_id=2178)\n* National Meteorological and Hydrological Service of Bolivia (SENAMHI Bolivia): [Hydrological Lake Titicaca Level Reports](https://senamhi.gob.bo/index.php/thidrologico)\n* National Meteorological and Hydrological Service of Peru (SENAMHI Peru): [Download Meteorological Data](https://www.senamhi.gob.pe/site/descarga-datos/) \n* Messager, M., Lehner, B., Grill, G. et al. Estimating the volume and age of water stored in global lakes using a geo-statistical approach. Nat Commun 7, 13603 (2016). [](https://doi.org/10.1038/ncomms13603)", "text_with_prefix": "EQC Quality Assessment: \"Assessing Lake Titicaca’s water levels in support of water management\"\nDataset: satellite-lake-water-level [CDS]\nAspect: completeness_q02 | Category: Satellite_ECVs\nSection: Assessing Lake Titicaca’s water levels in support of water management > ℹ️ If you want to know more\n---\n* Autonomous Binational Authority of the Lake Titicaca, Desaguadero River, Lake Poopó, and Salar de Coipasa Hydrological System (ALT): [Publications IWRM-TDPS](https://alt-perubolivia.org/?page_id=2178)\n* National Meteorological and Hydrological Service of Bolivia (SENAMHI Bolivia): [Hydrological Lake Titicaca Level Reports](https://senamhi.gob.bo/index.php/thidrologico)\n* National Meteorological and Hydrological Service of Peru (SENAMHI Peru): [Download Meteorological Data](https://www.senamhi.gob.pe/site/descarga-datos/) \n* Messager, M., Lehner, B., Grill, G. et al. Estimating the volume and age of water stored in global lakes using a geo-statistical approach. Nat Commun 7, 13603 (2016). [](https://doi.org/10.1038/ncomms13603)"} {"chunk_id": "satellite_satellite-lake-water-level_completeness_q02__f677a9be5a3f", "report_id": "satellite_satellite-lake-water-level_completeness_q02", "dataset_id": "satellite-lake-water-level", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q02", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Assessing Lake Titicaca’s water levels in support of water management > ℹ️ If you want to know more > Key resources", "title": "Assessing Lake Titicaca’s water levels in support of water management", "chunk_index": 16, "token_count": 288, "text_raw": "Code libraries used:\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)\n\nDataset documentation:\n\n* [LWL v5.0 and LWL-S v1.0: Product User Guide and Specification (PUGS)](https://confluence.ecmwf.int/pages/viewpage.action?pageId=425328964)\n* [LWL v5.0 and LWL-S v1.0: Algorithm Theoretical Basis Document (ATBD)](https://confluence.ecmwf.int/pages/viewpage.action?pageId=425328942)\n* [LWL v5.0 and LWL-S v1.0: Product Quality Assessment Report (PQAR)](https://confluence.ecmwf.int/pages/viewpage.action?pageId=428248112)", "text_with_prefix": "EQC Quality Assessment: \"Assessing Lake Titicaca’s water levels in support of water management\"\nDataset: satellite-lake-water-level [CDS]\nAspect: completeness_q02 | Category: Satellite_ECVs\nSection: Assessing Lake Titicaca’s water levels in support of water management > ℹ️ If you want to know more > Key resources\n---\nCode libraries used:\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)\n\nDataset documentation:\n\n* [LWL v5.0 and LWL-S v1.0: Product User Guide and Specification (PUGS)](https://confluence.ecmwf.int/pages/viewpage.action?pageId=425328964)\n* [LWL v5.0 and LWL-S v1.0: Algorithm Theoretical Basis Document (ATBD)](https://confluence.ecmwf.int/pages/viewpage.action?pageId=425328942)\n* [LWL v5.0 and LWL-S v1.0: Product Quality Assessment Report (PQAR)](https://confluence.ecmwf.int/pages/viewpage.action?pageId=428248112)"} {"chunk_id": "satellite_satellite-lake-water-level_completeness_q02__d661dbdd0b75", "report_id": "satellite_satellite-lake-water-level_completeness_q02", "dataset_id": "satellite-lake-water-level", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q02", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Assessing Lake Titicaca’s water levels in support of water management > ℹ️ If you want to know more > References", "title": "Assessing Lake Titicaca’s water levels in support of water management", "chunk_index": 17, "token_count": 897, "text_raw": "[[1]](https://doi.org/10.3390/w8040144) Canedo, C., Pillco Zolá, R., & Berndtsson, R. (2016). Role of Hydrological Studies for the Development of the TDPS System. Water, 8(4), 144.\n\n[[2]](https://doi.org/10.1016/j.ejrh.2021.100927) Lima-Quispe, N., Escobar, M., Wickel, A. J., von Kaenel, M., & Purkey, D. (2021). Untangling the effects of climate variability and irrigation management on water levels in Lakes Titicaca and Poopó. Journal of Hydrology: Regional Studies, 37, 100927.\n\n[[3]](https://doi.org/10.1016/j.wace.2024.100710) Gutierrez-Villarreal, R. A., Espinoza, J.-C., Lavado-Casimiro, W., Junquas, C., Molina-Carpio, J., Condom, T., & Marengo, J. A. (2024). The 2022–23 drought in the South American Altiplano: ENSO effects on moisture flux in the western Amazon during the pre-wet season. Weather and Climate Extremes, 43, 100710.\n\n[[4]](https://senamhi.gob.bo/meteorologia/boletines/hidrologico/LagoTiticaca/2025/Agosto/LAGO%20TITICACA%20-%20Informe%20Semanal%20%28Actualizado%2001_08_2025%29.pdf) SENAMHI Bolivia (2025) BOLETÍN SEMANAL LAGO TITICACA ACTUALIZADO AL 01 DE AGOSTO DE 2025. Ministry of Environment and Water of Bolivia.\n\n[[5]](https://doi.org/10.5194/hess-29-655-2025) Lima-Quispe, N., Ruelland, D., Rabatel, A., Lavado-Casimiro, W., & Condom, T. (2025). Modeling Lake Titicaca's water balance: The dominant roles of precipitation and evaporation. Hydrology and Earth System Sciences, 29, 655–675.\n\n[[6]](https://reliefweb.int/report/peru/act-peru-12003-heavy-rainfall) ACT Alliance (2003) ACT Peru 1/2003: Heavy rainfall. ReliefWeb, United Nations Office for the Coordination of Humanitarian Affairs (OCHA).\n\n[[7]](https://reliefweb.int/report/peru/flash-floods-leave-peruvian-farmers-devastated) IFRC (2003) Flash floods leave Peruvian farmers devastated. ReliefWeb, United Nations Office for the Coordination of Humanitarian Affairs (OCHA).\n\n[[8]](https://www.inei.gob.pe/media/MenuRecursivo/boletines/5340.pdf) Instituto Nacional de Estadística e Informática del Perú INE (2004) Estadísticas Ambientales.\n\n[[9]](https://www.copernicus.eu/fr/node/40141) Copernicus (2023) Drought in Lake Titicaca, South America.\n\n[[10]](https://doi.org/10.1029/2002jd002670) Rayner N. A., Parker D. E., Horton E. B., Folland C. K., Alexander L. V., Rowell D. P., Kent E. C., & Kaplan A. (2003) Global analyses of sea surface temperature, sea ice, and night marine air temperature since the late nineteenth century, J. Geophys. Res., 108 (D14), 4407.", "text_with_prefix": "EQC Quality Assessment: \"Assessing Lake Titicaca’s water levels in support of water management\"\nDataset: satellite-lake-water-level [CDS]\nAspect: completeness_q02 | Category: Satellite_ECVs\nSection: Assessing Lake Titicaca’s water levels in support of water management > ℹ️ If you want to know more > References\n---\n[[1]](https://doi.org/10.3390/w8040144) Canedo, C., Pillco Zolá, R., & Berndtsson, R. (2016). Role of Hydrological Studies for the Development of the TDPS System. Water, 8(4), 144.\n\n[[2]](https://doi.org/10.1016/j.ejrh.2021.100927) Lima-Quispe, N., Escobar, M., Wickel, A. J., von Kaenel, M., & Purkey, D. (2021). Untangling the effects of climate variability and irrigation management on water levels in Lakes Titicaca and Poopó. Journal of Hydrology: Regional Studies, 37, 100927.\n\n[[3]](https://doi.org/10.1016/j.wace.2024.100710) Gutierrez-Villarreal, R. A., Espinoza, J.-C., Lavado-Casimiro, W., Junquas, C., Molina-Carpio, J., Condom, T., & Marengo, J. A. (2024). The 2022–23 drought in the South American Altiplano: ENSO effects on moisture flux in the western Amazon during the pre-wet season. Weather and Climate Extremes, 43, 100710.\n\n[[4]](https://senamhi.gob.bo/meteorologia/boletines/hidrologico/LagoTiticaca/2025/Agosto/LAGO%20TITICACA%20-%20Informe%20Semanal%20%28Actualizado%2001_08_2025%29.pdf) SENAMHI Bolivia (2025) BOLETÍN SEMANAL LAGO TITICACA ACTUALIZADO AL 01 DE AGOSTO DE 2025. Ministry of Environment and Water of Bolivia.\n\n[[5]](https://doi.org/10.5194/hess-29-655-2025) Lima-Quispe, N., Ruelland, D., Rabatel, A., Lavado-Casimiro, W., & Condom, T. (2025). Modeling Lake Titicaca's water balance: The dominant roles of precipitation and evaporation. Hydrology and Earth System Sciences, 29, 655–675.\n\n[[6]](https://reliefweb.int/report/peru/act-peru-12003-heavy-rainfall) ACT Alliance (2003) ACT Peru 1/2003: Heavy rainfall. ReliefWeb, United Nations Office for the Coordination of Humanitarian Affairs (OCHA).\n\n[[7]](https://reliefweb.int/report/peru/flash-floods-leave-peruvian-farmers-devastated) IFRC (2003) Flash floods leave Peruvian farmers devastated. ReliefWeb, United Nations Office for the Coordination of Humanitarian Affairs (OCHA).\n\n[[8]](https://www.inei.gob.pe/media/MenuRecursivo/boletines/5340.pdf) Instituto Nacional de Estadística e Informática del Perú INE (2004) Estadísticas Ambientales.\n\n[[9]](https://www.copernicus.eu/fr/node/40141) Copernicus (2023) Drought in Lake Titicaca, South America.\n\n[[10]](https://doi.org/10.1029/2002jd002670) Rayner N. A., Parker D. E., Horton E. B., Folland C. K., Alexander L. V., Rowell D. P., Kent E. C., & Kaplan A. (2003) Global analyses of sea surface temperature, sea ice, and night marine air temperature since the late nineteenth century, J. Geophys. Res., 108 (D14), 4407."} {"chunk_id": "satellite_satellite-lake-water-temperature_validation_q02__323eed544835", "report_id": "satellite_satellite-lake-water-temperature_validation_q02", "dataset_id": "satellite-lake-water-temperature", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q02", "aspect_base": "validation", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite-Derived Monitoring of Summer LSWT in Lake Superior", "title": "Satellite-Derived Monitoring of Summer LSWT in Lake Superior", "chunk_index": 0, "token_count": 97, "text_raw": "Production date: 15-05-2026\n\nProduced by: Victor Couplet (VUB) & Camila Trigoso (VUB)", "text_with_prefix": "EQC Quality Assessment: \"Satellite-Derived Monitoring of Summer LSWT in Lake Superior\"\nDataset: satellite-lake-water-temperature [CDS]\nAspect: validation_q02 | Category: Satellite_ECVs\nSection: Satellite-Derived Monitoring of Summer LSWT in Lake Superior\n---\nProduction date: 15-05-2026\n\nProduced by: Victor Couplet (VUB) & Camila Trigoso (VUB)"} {"chunk_id": "satellite_satellite-lake-water-temperature_validation_q02__d6ed8bf3894d", "report_id": "satellite_satellite-lake-water-temperature_validation_q02", "dataset_id": "satellite-lake-water-temperature", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q02", "aspect_base": "validation", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite-Derived Monitoring of Summer LSWT in Lake Superior > Quality assessment question(s)", "title": "Satellite-Derived Monitoring of Summer LSWT in Lake Superior", "chunk_index": 1, "token_count": 660, "text_raw": "* **Can the C3S satellite lake surface water temperature dataset accurately reproduce summer LSWT observations in Lake Superior, and is it suitable for assessing long-term summer temperature variability and trends?**\n\nCases of an increasing trend in lake surface water temperature (LSWT) have been observed in numerous lakes across various regions, including the United States, Europe, and other parts of the world [[1]](https://doi.org/10.1038/s41558-024-02122-y) [[2]](https://climate.copernicus.eu/lake-surface-temperatures). For example, Austin and Colman (2008) found that summer LSWT in Lake Superior increased at a rate of (11±6)×10-2 °C/year from 1979 to 2006. They attributed this trend to the earlier retreat of winter ice, which causes the positive overturning period to begin sooner, allowing the lake more time to warm. Their analysis was based on measurements from in-situ buoys [[3]](https://doi.org/10.1029/2006GL029021). Later, the Great Lakes Integrated Sciences and Assessments (GLISA) analyzed the annual mean LSWT of Lake Superior between 1995 and 2021. The increasing long-term trend observed by Austin and Colman was not apparent in the period analyzed by GLISA. However, they noted a shift to higher temperatures after 1998 [[4]](https://glisa.umich.edu/lake-superior-retrospective/). The LSWT data analyzed by GLISA was derived from satellite observations and obtained through the NOAA Great Lakes CoastWatch [[5]](https://coastwatch.glerl.noaa.gov/statistics/average-surface-water-temperature-glsea/). Cannon et al. (2024) investigated summer LSWT of Lake Superior using simulation data over a longer time interval (1980-2021), and found a warming trend of (5.2±3.1)×10-2 °C/year [[6]](https://doi.org/10.1175/JCLI-D-23-0092.1).\n\nThe objective of this assessment is to evaluate whether the [satellite-lake-water-temperature](https://cds.climate.copernicus.eu/datasets/satellite-lake-water-temperature?tab=overview) dataset from the Climate Data Store (CDS) is suitable for climate change monitoring, using Lake Superior as a case study. We calculated the trend summer LWST for a period of 28 years, from 1995 to 2023, and validated the data and trend computation against in-situ buoy measurements from the NOAA National Buoy Data Center (NBDC) [[7]](https://www.ndbc.noaa.gov/).", "text_with_prefix": "EQC Quality Assessment: \"Satellite-Derived Monitoring of Summer LSWT in Lake Superior\"\nDataset: satellite-lake-water-temperature [CDS]\nAspect: validation_q02 | Category: Satellite_ECVs\nSection: Satellite-Derived Monitoring of Summer LSWT in Lake Superior > Quality assessment question(s)\n---\n* **Can the C3S satellite lake surface water temperature dataset accurately reproduce summer LSWT observations in Lake Superior, and is it suitable for assessing long-term summer temperature variability and trends?**\n\nCases of an increasing trend in lake surface water temperature (LSWT) have been observed in numerous lakes across various regions, including the United States, Europe, and other parts of the world [[1]](https://doi.org/10.1038/s41558-024-02122-y) [[2]](https://climate.copernicus.eu/lake-surface-temperatures). For example, Austin and Colman (2008) found that summer LSWT in Lake Superior increased at a rate of (11±6)×10-2 °C/year from 1979 to 2006. They attributed this trend to the earlier retreat of winter ice, which causes the positive overturning period to begin sooner, allowing the lake more time to warm. Their analysis was based on measurements from in-situ buoys [[3]](https://doi.org/10.1029/2006GL029021). Later, the Great Lakes Integrated Sciences and Assessments (GLISA) analyzed the annual mean LSWT of Lake Superior between 1995 and 2021. The increasing long-term trend observed by Austin and Colman was not apparent in the period analyzed by GLISA. However, they noted a shift to higher temperatures after 1998 [[4]](https://glisa.umich.edu/lake-superior-retrospective/). The LSWT data analyzed by GLISA was derived from satellite observations and obtained through the NOAA Great Lakes CoastWatch [[5]](https://coastwatch.glerl.noaa.gov/statistics/average-surface-water-temperature-glsea/). Cannon et al. (2024) investigated summer LSWT of Lake Superior using simulation data over a longer time interval (1980-2021), and found a warming trend of (5.2±3.1)×10-2 °C/year [[6]](https://doi.org/10.1175/JCLI-D-23-0092.1).\n\nThe objective of this assessment is to evaluate whether the [satellite-lake-water-temperature](https://cds.climate.copernicus.eu/datasets/satellite-lake-water-temperature?tab=overview) dataset from the Climate Data Store (CDS) is suitable for climate change monitoring, using Lake Superior as a case study. We calculated the trend summer LWST for a period of 28 years, from 1995 to 2023, and validated the data and trend computation against in-situ buoy measurements from the NOAA National Buoy Data Center (NBDC) [[7]](https://www.ndbc.noaa.gov/)."} {"chunk_id": "satellite_satellite-lake-water-temperature_validation_q02__f542eb863ba0", "report_id": "satellite_satellite-lake-water-temperature_validation_q02", "dataset_id": "satellite-lake-water-temperature", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q02", "aspect_base": "validation", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite-Derived Monitoring of Summer LSWT in Lake Superior > Quality assessment statement", "title": "Satellite-Derived Monitoring of Summer LSWT in Lake Superior", "chunk_index": 2, "token_count": 322, "text_raw": "These are the key outcomes of this assessment\n\n* Lake Superior’s summer LSWT trend over 1995–2023, derived from the C3S satellite lake water temperature dataset, is small and not statistically significant. This indicates that no robust long-term warming or cooling signal can be detected over this period.\n* The C3S dataset shows good agreement with NOAA in situ buoy observations at selected locations. Differences in summer mean temperatures are partly attributable to the higher temporal sampling of in situ measurements and to differences in measurement depth, with satellites observing skin temperature and buoys measuring subsurface temperature (≈1 m). Additional discrepancies may result from cloud contamination in the satellite record, as well as from limitations in the in situ data, which may include temporal gaps, quality issues, and measurement inconsistencies.\n* The dataset is well suited for monitoring interannual variability in lake surface temperature; however, trend analyses require caution. The lack of statistical significance in the estimated trends, combined with strong interannual variability (e.g., elevated temperatures during the 1998 El Niño event), limits the ability to detect long-term changes over this period. Incorporating longer time series, including pre-1995 observations where available, would improve the robustness of trend assessments.\n```", "text_with_prefix": "EQC Quality Assessment: \"Satellite-Derived Monitoring of Summer LSWT in Lake Superior\"\nDataset: satellite-lake-water-temperature [CDS]\nAspect: validation_q02 | Category: Satellite_ECVs\nSection: Satellite-Derived Monitoring of Summer LSWT in Lake Superior > Quality assessment statement\n---\nThese are the key outcomes of this assessment\n\n* Lake Superior’s summer LSWT trend over 1995–2023, derived from the C3S satellite lake water temperature dataset, is small and not statistically significant. This indicates that no robust long-term warming or cooling signal can be detected over this period.\n* The C3S dataset shows good agreement with NOAA in situ buoy observations at selected locations. Differences in summer mean temperatures are partly attributable to the higher temporal sampling of in situ measurements and to differences in measurement depth, with satellites observing skin temperature and buoys measuring subsurface temperature (≈1 m). Additional discrepancies may result from cloud contamination in the satellite record, as well as from limitations in the in situ data, which may include temporal gaps, quality issues, and measurement inconsistencies.\n* The dataset is well suited for monitoring interannual variability in lake surface temperature; however, trend analyses require caution. The lack of statistical significance in the estimated trends, combined with strong interannual variability (e.g., elevated temperatures during the 1998 El Niño event), limits the ability to detect long-term changes over this period. Incorporating longer time series, including pre-1995 observations where available, would improve the robustness of trend assessments.\n```"} {"chunk_id": "satellite_satellite-lake-water-temperature_validation_q02__81ef3387b77c", "report_id": "satellite_satellite-lake-water-temperature_validation_q02", "dataset_id": "satellite-lake-water-temperature", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q02", "aspect_base": "validation", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite-Derived Monitoring of Summer LSWT in Lake Superior > Methodology", "title": "Satellite-Derived Monitoring of Summer LSWT in Lake Superior", "chunk_index": 3, "token_count": 357, "text_raw": "The analysis and results are organized in the following steps, which are detailed in the sections below:\n\n**[](section-1)**\n\n* Satellite lake surface water temperature (LSWT) data are downloaded for the summer months (July–September) over the period 1995–2023.\n\n**[](section-2)**\n\n* The dataset is filtered based on quality flags and lake identification to retain only high-quality observations for Lake Superior.\n\n**[](section-3)**\n\n* A spatially weighted mean LSWT is computed, followed by the calculation of annual summer mean LSWT values.\n* The temporal trend is estimated using the Theil–Sen estimator, and its statistical significance is assessed with the Mann–Kendall test.\n\n**[](section-4)**\n\n* Satellite LSWT data are validated against in-situ measurements from three NOAA-NBDC buoys located in Lake Superior.\n* Performance metrics, including mean bias, mean absolute error (MAE), and root mean square error (RMSE), are calculated.\n* Buoy observations are resampled to match the less frequent satellite overpass times. Metrics are then recalculated on these time-matched datasets to determine whether differences in summer means arise from sampling frequency or from measurement characteristics (e.g., sensing depth, retrieval uncertainties).\n* Finally, LSWT trends derived from buoy data over multiple periods are computed and compared with satellite-based trends and published literature results.", "text_with_prefix": "EQC Quality Assessment: \"Satellite-Derived Monitoring of Summer LSWT in Lake Superior\"\nDataset: satellite-lake-water-temperature [CDS]\nAspect: validation_q02 | Category: Satellite_ECVs\nSection: Satellite-Derived Monitoring of Summer LSWT in Lake Superior > Methodology\n---\nThe analysis and results are organized in the following steps, which are detailed in the sections below:\n\n**[](section-1)**\n\n* Satellite lake surface water temperature (LSWT) data are downloaded for the summer months (July–September) over the period 1995–2023.\n\n**[](section-2)**\n\n* The dataset is filtered based on quality flags and lake identification to retain only high-quality observations for Lake Superior.\n\n**[](section-3)**\n\n* A spatially weighted mean LSWT is computed, followed by the calculation of annual summer mean LSWT values.\n* The temporal trend is estimated using the Theil–Sen estimator, and its statistical significance is assessed with the Mann–Kendall test.\n\n**[](section-4)**\n\n* Satellite LSWT data are validated against in-situ measurements from three NOAA-NBDC buoys located in Lake Superior.\n* Performance metrics, including mean bias, mean absolute error (MAE), and root mean square error (RMSE), are calculated.\n* Buoy observations are resampled to match the less frequent satellite overpass times. Metrics are then recalculated on these time-matched datasets to determine whether differences in summer means arise from sampling frequency or from measurement characteristics (e.g., sensing depth, retrieval uncertainties).\n* Finally, LSWT trends derived from buoy data over multiple periods are computed and compared with satellite-based trends and published literature results."} {"chunk_id": "satellite_satellite-lake-water-temperature_validation_q02__73d2d23917f2", "report_id": "satellite_satellite-lake-water-temperature_validation_q02", "dataset_id": "satellite-lake-water-temperature", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q02", "aspect_base": "validation", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite-Derived Monitoring of Summer LSWT in Lake Superior > Analysis and results > 1. Request and download data > Download data", "title": "Satellite-Derived Monitoring of Summer LSWT in Lake Superior", "chunk_index": 4, "token_count": 117, "text_raw": "```text\n100%|██████████| 29/29 [00:06<00:00, 4.25it/s]\n```\n\n(section-2)=", "text_with_prefix": "EQC Quality Assessment: \"Satellite-Derived Monitoring of Summer LSWT in Lake Superior\"\nDataset: satellite-lake-water-temperature [CDS]\nAspect: validation_q02 | Category: Satellite_ECVs\nSection: Satellite-Derived Monitoring of Summer LSWT in Lake Superior > Analysis and results > 1. Request and download data > Download data\n---\n```text\n100%|██████████| 29/29 [00:06<00:00, 4.25it/s]\n```\n\n(section-2)="} {"chunk_id": "satellite_satellite-lake-water-temperature_validation_q02__cc4e3b9982ef", "report_id": "satellite_satellite-lake-water-temperature_validation_q02", "dataset_id": "satellite-lake-water-temperature", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q02", "aspect_base": "validation", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite-Derived Monitoring of Summer LSWT in Lake Superior > Analysis and results > 2. Data preprocessing", "title": "Satellite-Derived Monitoring of Summer LSWT in Lake Superior", "chunk_index": 5, "token_count": 146, "text_raw": "We download satellite lake surface temperature data and visualize the water masses over the region spanning 46.30–49.00°N latitude and 92.10–84.80°W longitude. In the dataset, Lake Superior is identified by ID number 2. The locations of the three NOAA buoys used for data comparison are also plotted.", "text_with_prefix": "EQC Quality Assessment: \"Satellite-Derived Monitoring of Summer LSWT in Lake Superior\"\nDataset: satellite-lake-water-temperature [CDS]\nAspect: validation_q02 | Category: Satellite_ECVs\nSection: Satellite-Derived Monitoring of Summer LSWT in Lake Superior > Analysis and results > 2. Data preprocessing\n---\nWe download satellite lake surface temperature data and visualize the water masses over the region spanning 46.30–49.00°N latitude and 92.10–84.80°W longitude. In the dataset, Lake Superior is identified by ID number 2. The locations of the three NOAA buoys used for data comparison are also plotted."} {"chunk_id": "satellite_satellite-lake-water-temperature_validation_q02__bcadf821fb40", "report_id": "satellite_satellite-lake-water-temperature_validation_q02", "dataset_id": "satellite-lake-water-temperature", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q02", "aspect_base": "validation", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite-Derived Monitoring of Summer LSWT in Lake Superior > Analysis and results > 2. Data preprocessing > Plot lakeid", "title": "Satellite-Derived Monitoring of Summer LSWT in Lake Superior", "chunk_index": 6, "token_count": 175, "text_raw": "Use one time slice for plotting\nUnique lake IDs\nMap real lake IDs to categorical indices: 2 -> 0, 300000709 -> 1, etc.\n------------------------------------------------------------------\nBuoy coordinates (lon, lat)\n------------------------------------------------------------------\n------------------------------------------------------------------\n------------------------------------------------------------------\nLongitude / latitude ticks\n------------------------------------------------------------------\nLegend for lake IDs\n\n*Figure 1 : Lake Superior identification in the dataset. Locations of the three NOAA-NBDC buoys used for data comparison are shown as orange crosses.*", "text_with_prefix": "EQC Quality Assessment: \"Satellite-Derived Monitoring of Summer LSWT in Lake Superior\"\nDataset: satellite-lake-water-temperature [CDS]\nAspect: validation_q02 | Category: Satellite_ECVs\nSection: Satellite-Derived Monitoring of Summer LSWT in Lake Superior > Analysis and results > 2. Data preprocessing > Plot lakeid\n---\nUse one time slice for plotting\nUnique lake IDs\nMap real lake IDs to categorical indices: 2 -> 0, 300000709 -> 1, etc.\n------------------------------------------------------------------\nBuoy coordinates (lon, lat)\n------------------------------------------------------------------\n------------------------------------------------------------------\n------------------------------------------------------------------\nLongitude / latitude ticks\n------------------------------------------------------------------\nLegend for lake IDs\n\n*Figure 1 : Lake Superior identification in the dataset. Locations of the three NOAA-NBDC buoys used for data comparison are shown as orange crosses.*"} {"chunk_id": "satellite_satellite-lake-water-temperature_validation_q02__83722cd86950", "report_id": "satellite_satellite-lake-water-temperature_validation_q02", "dataset_id": "satellite-lake-water-temperature", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q02", "aspect_base": "validation", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite-Derived Monitoring of Summer LSWT in Lake Superior > Analysis and results > 2. Data preprocessing > Data filtering", "title": "Satellite-Derived Monitoring of Summer LSWT in Lake Superior", "chunk_index": 7, "token_count": 113, "text_raw": "Only observations with quality levels 4 (good) and 5 (best) are retained for analysis.\n\nReindex using lakeids and min_quality_level\n\n(section-3)=", "text_with_prefix": "EQC Quality Assessment: \"Satellite-Derived Monitoring of Summer LSWT in Lake Superior\"\nDataset: satellite-lake-water-temperature [CDS]\nAspect: validation_q02 | Category: Satellite_ECVs\nSection: Satellite-Derived Monitoring of Summer LSWT in Lake Superior > Analysis and results > 2. Data preprocessing > Data filtering\n---\nOnly observations with quality levels 4 (good) and 5 (best) are retained for analysis.\n\nReindex using lakeids and min_quality_level\n\n(section-3)="} {"chunk_id": "satellite_satellite-lake-water-temperature_validation_q02__754be0d7a5bc", "report_id": "satellite_satellite-lake-water-temperature_validation_q02", "dataset_id": "satellite-lake-water-temperature", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q02", "aspect_base": "validation", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite-Derived Monitoring of Summer LSWT in Lake Superior > Analysis and results > 3. Summer mean analysis", "title": "Satellite-Derived Monitoring of Summer LSWT in Lake Superior", "chunk_index": 8, "token_count": 141, "text_raw": "For each time step, lake surface water temperatures are averaged spatially across Lake Superior, and these values are then averaged over the summer months (July–September). The 1995–2023 trend is estimated using the Theil–Sen estimator, and its statistical significance is assessed with the Mann–Kendall test.", "text_with_prefix": "EQC Quality Assessment: \"Satellite-Derived Monitoring of Summer LSWT in Lake Superior\"\nDataset: satellite-lake-water-temperature [CDS]\nAspect: validation_q02 | Category: Satellite_ECVs\nSection: Satellite-Derived Monitoring of Summer LSWT in Lake Superior > Analysis and results > 3. Summer mean analysis\n---\nFor each time step, lake surface water temperatures are averaged spatially across Lake Superior, and these values are then averaged over the summer months (July–September). The 1995–2023 trend is estimated using the Theil–Sen estimator, and its statistical significance is assessed with the Mann–Kendall test."} {"chunk_id": "satellite_satellite-lake-water-temperature_validation_q02__e38bd76745d5", "report_id": "satellite_satellite-lake-water-temperature_validation_q02", "dataset_id": "satellite-lake-water-temperature", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q02", "aspect_base": "validation", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite-Derived Monitoring of Summer LSWT in Lake Superior > Analysis and results > 3. Summer mean analysis > Trend estimation", "title": "Satellite-Derived Monitoring of Summer LSWT in Lake Superior", "chunk_index": 9, "token_count": 314, "text_raw": "Run Theil-Sen + Mann-Kendall\nCompute Theil-Sen line for plotting\n\n*Figure 2 : Mean summer LSWT for Lake Superior between 1995 and 2023, and trend estimation.*\n\nThe trend line in *Figure 2* shows a slight negative slope of −0.0032 °C yr⁻¹. However, the p-value of 0.81 is well above the common significance threshold of 0.05, indicating that the trend is not statistically significant. This lack of significance may partly reflect the relatively short 28-year period analyzed, as longer time series are typically required to detect meaningful climate trends in large lakes.\n\nAt first glance, the absence of a significant warming trend may seem surprising. Previous work [[3]](https://doi.org/10.1029/2006GL029021) reported a significant increase in summer temperatures from 1979 to 2006, and it seems unlikely that summer lake surface temperatures have stabilised since then, given ongoing global warming.\n\nTo verify these results, we will compare the satellite data with in-situ buoy measurements.\n\n(section-4)=", "text_with_prefix": "EQC Quality Assessment: \"Satellite-Derived Monitoring of Summer LSWT in Lake Superior\"\nDataset: satellite-lake-water-temperature [CDS]\nAspect: validation_q02 | Category: Satellite_ECVs\nSection: Satellite-Derived Monitoring of Summer LSWT in Lake Superior > Analysis and results > 3. Summer mean analysis > Trend estimation\n---\nRun Theil-Sen + Mann-Kendall\nCompute Theil-Sen line for plotting\n\n*Figure 2 : Mean summer LSWT for Lake Superior between 1995 and 2023, and trend estimation.*\n\nThe trend line in *Figure 2* shows a slight negative slope of −0.0032 °C yr⁻¹. However, the p-value of 0.81 is well above the common significance threshold of 0.05, indicating that the trend is not statistically significant. This lack of significance may partly reflect the relatively short 28-year period analyzed, as longer time series are typically required to detect meaningful climate trends in large lakes.\n\nAt first glance, the absence of a significant warming trend may seem surprising. Previous work [[3]](https://doi.org/10.1029/2006GL029021) reported a significant increase in summer temperatures from 1979 to 2006, and it seems unlikely that summer lake surface temperatures have stabilised since then, given ongoing global warming.\n\nTo verify these results, we will compare the satellite data with in-situ buoy measurements.\n\n(section-4)="} {"chunk_id": "satellite_satellite-lake-water-temperature_validation_q02__0980e5106cc8", "report_id": "satellite_satellite-lake-water-temperature_validation_q02", "dataset_id": "satellite-lake-water-temperature", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q02", "aspect_base": "validation", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite-Derived Monitoring of Summer LSWT in Lake Superior > Analysis and results > 4. Comparison with NOAA buoy data", "title": "Satellite-Derived Monitoring of Summer LSWT in Lake Superior", "chunk_index": 10, "token_count": 1079, "text_raw": "The satellite data are compared with in-situ measurements from three NOAA NDBC buoys in Lake Superior [[7]](https://www.ndbc.noaa.gov/). These are the same buoys used in [[3]](https://doi.org/10.1029/2006GL029021), which reported a significant increase in summer temperatures from 1979 to 2006. In-situ data are now available through 2024. The goal of this section is thus to compare satellite and buoy data over 1995–2023 and evaluate whether similar trends are observed in the buoy measurements.\n\nThe buoy data are downloaded from the NOAA National Data Buoy Center website [[7]](https://www.ndbc.noaa.gov/), which provides data by station and year. We developed a Python script that automatically downloads the data for buoys 45001, 45004, and 45006 for all available years between 1979 and 2024, and combines them into a single, unified dataset.\n\nFind header: first non-empty line whose first token contains letters\nExtract data lines (non-empty, non-comment lines after header)\nprint(f\"Downloading {url} ...\", end=\" \")\nEnsure numeric columns for date/time\nDetermine proper year\nIf any 4-digit year present, treat as full year\n2-digit year: <50 -> 2000s, >=50 -> 1900s\nEnsure minutes exist\nBuild datetime column\nProcess WTMP\nBuild output dataframe and filter invalid/extreme values\nprint(f\" → loaded {len(out)} rows\")\nConcatenate all data\nSet MultiIndex (station, datetime)\n\n```text\nOverall progress: 34%|███▍ | 47/138 [00:34<01:21, 1.12it/s, station=45004, year=1980]\n```\n\n```text\n→ download error: 404 Client Error: Not Found for url: https://www.ndbc.noaa.gov/data/historical/stdmet/45004h1979.txt.gz\n```\n\n```text\nOverall progress: 67%|██████▋ | 93/138 [01:09<00:46, 1.03s/it, station=45006, year=1980]\n```\n\n```text\n→ download error: 404 Client Error: Not Found for url: https://www.ndbc.noaa.gov/data/historical/stdmet/45006h1979.txt.gz\n```\n\n```text\nOverall progress: 68%|██████▊ | 94/138 [01:09<00:34, 1.26it/s, station=45006, year=1981]\n```\n\n```text\n→ download error: 404 Client Error: Not Found for url: https://www.ndbc.noaa.gov/data/historical/stdmet/45006h1980.txt.gz\n```\n\n```text\nOverall progress: 100%|██████████| 138/138 [01:42<00:00, 1.34it/s, station=45006, year=2024]\n```\n\n```text\nTotal rows loaded: 958592\n```\n\nAlthough NOAA National Data Buoy Center (NDBC) buoy observations provide an important independent reference for evaluating satellite-derived LSWT, they should not be treated as error-free ground truth. NDBC applies both real-time automated quality control and delayed/manual quality control to its observations. According to the NDBC Handbook of Automated Data Quality Control Checks and Procedures [[8]](https://www.ndbc.noaa.gov/publications/NDBCHandbookofAutomatedDataQualityControl2023.pdf), automated checks are used to detect issues such as transmission errors, missing data, total sensor failure, range-limit exceedances, and unrealistic temporal changes. Observations may be flagged as good, suspect, failed, or missing, and some failed observations are not released in real time. Manual quality control can also be applied after data are stored, including manual failure of data from sensors judged to be unreliable.\n\nTherefore, discrepancies between satellite and buoy LSWT should not automatically be interpreted as satellite retrieval errors. They may also reflect limitations in the in situ record, including temporal gaps, sensor problems, incomplete quality control, or unrealistic variability.\n\nThe buoy data are averaged over the summer months, and each buoy is compared to the summer mean temperature of the nearest satellite pixel corresponding to its location (see *Figure 3*).\n\nCompare buoy and satellite summer mean temperatures\nExtract nearest-neighbor satellite temperature at each buoy position,\nconvert from Kelvin to Celsius, and compute summer annual means\nNearest satellite pixel to buoy location\nConvert from Kelvin to Celsius\nRestrict to summer months\nCompute annual summer mean\nConcatenate all stations into a single reusable table\nPlot comparison", "text_with_prefix": "EQC Quality Assessment: \"Satellite-Derived Monitoring of Summer LSWT in Lake Superior\"\nDataset: satellite-lake-water-temperature [CDS]\nAspect: validation_q02 | Category: Satellite_ECVs\nSection: Satellite-Derived Monitoring of Summer LSWT in Lake Superior > Analysis and results > 4. Comparison with NOAA buoy data\n---\nThe satellite data are compared with in-situ measurements from three NOAA NDBC buoys in Lake Superior [[7]](https://www.ndbc.noaa.gov/). These are the same buoys used in [[3]](https://doi.org/10.1029/2006GL029021), which reported a significant increase in summer temperatures from 1979 to 2006. In-situ data are now available through 2024. The goal of this section is thus to compare satellite and buoy data over 1995–2023 and evaluate whether similar trends are observed in the buoy measurements.\n\nThe buoy data are downloaded from the NOAA National Data Buoy Center website [[7]](https://www.ndbc.noaa.gov/), which provides data by station and year. We developed a Python script that automatically downloads the data for buoys 45001, 45004, and 45006 for all available years between 1979 and 2024, and combines them into a single, unified dataset.\n\nFind header: first non-empty line whose first token contains letters\nExtract data lines (non-empty, non-comment lines after header)\nprint(f\"Downloading {url} ...\", end=\" \")\nEnsure numeric columns for date/time\nDetermine proper year\nIf any 4-digit year present, treat as full year\n2-digit year: <50 -> 2000s, >=50 -> 1900s\nEnsure minutes exist\nBuild datetime column\nProcess WTMP\nBuild output dataframe and filter invalid/extreme values\nprint(f\" → loaded {len(out)} rows\")\nConcatenate all data\nSet MultiIndex (station, datetime)\n\n```text\nOverall progress: 34%|███▍ | 47/138 [00:34<01:21, 1.12it/s, station=45004, year=1980]\n```\n\n```text\n→ download error: 404 Client Error: Not Found for url: https://www.ndbc.noaa.gov/data/historical/stdmet/45004h1979.txt.gz\n```\n\n```text\nOverall progress: 67%|██████▋ | 93/138 [01:09<00:46, 1.03s/it, station=45006, year=1980]\n```\n\n```text\n→ download error: 404 Client Error: Not Found for url: https://www.ndbc.noaa.gov/data/historical/stdmet/45006h1979.txt.gz\n```\n\n```text\nOverall progress: 68%|██████▊ | 94/138 [01:09<00:34, 1.26it/s, station=45006, year=1981]\n```\n\n```text\n→ download error: 404 Client Error: Not Found for url: https://www.ndbc.noaa.gov/data/historical/stdmet/45006h1980.txt.gz\n```\n\n```text\nOverall progress: 100%|██████████| 138/138 [01:42<00:00, 1.34it/s, station=45006, year=2024]\n```\n\n```text\nTotal rows loaded: 958592\n```\n\nAlthough NOAA National Data Buoy Center (NDBC) buoy observations provide an important independent reference for evaluating satellite-derived LSWT, they should not be treated as error-free ground truth. NDBC applies both real-time automated quality control and delayed/manual quality control to its observations. According to the NDBC Handbook of Automated Data Quality Control Checks and Procedures [[8]](https://www.ndbc.noaa.gov/publications/NDBCHandbookofAutomatedDataQualityControl2023.pdf), automated checks are used to detect issues such as transmission errors, missing data, total sensor failure, range-limit exceedances, and unrealistic temporal changes. Observations may be flagged as good, suspect, failed, or missing, and some failed observations are not released in real time. Manual quality control can also be applied after data are stored, including manual failure of data from sensors judged to be unreliable.\n\nTherefore, discrepancies between satellite and buoy LSWT should not automatically be interpreted as satellite retrieval errors. They may also reflect limitations in the in situ record, including temporal gaps, sensor problems, incomplete quality control, or unrealistic variability.\n\nThe buoy data are averaged over the summer months, and each buoy is compared to the summer mean temperature of the nearest satellite pixel corresponding to its location (see *Figure 3*).\n\nCompare buoy and satellite summer mean temperatures\nExtract nearest-neighbor satellite temperature at each buoy position,\nconvert from Kelvin to Celsius, and compute summer annual means\nNearest satellite pixel to buoy location\nConvert from Kelvin to Celsius\nRestrict to summer months\nCompute annual summer mean\nConcatenate all stations into a single reusable table\nPlot comparison"} {"chunk_id": "satellite_satellite-lake-water-temperature_validation_q02__552b3a9ad9aa", "report_id": "satellite_satellite-lake-water-temperature_validation_q02", "dataset_id": "satellite-lake-water-temperature", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q02", "aspect_base": "validation", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite-Derived Monitoring of Summer LSWT in Lake Superior > Analysis and results > 4. Comparison with NOAA buoy data", "title": "Satellite-Derived Monitoring of Summer LSWT in Lake Superior", "chunk_index": 11, "token_count": 1047, "text_raw": ",\nconvert from Kelvin to Celsius, and compute summer annual means\nNearest satellite pixel to buoy location\nConvert from Kelvin to Celsius\nRestrict to summer months\nCompute annual summer mean\nConcatenate all stations into a single reusable table\nPlot comparison\n\n*Figure 3: Mean summer LSWT for three Lake Superior buoys, compared to the mean summer LSWT from satellite data at the nearest pixel locations.*\n\nWe observe that the satellite data are generally close to the buoy measurements; for instance, the 1998 peak in summer LSWT is well captured. To further quantify the differences between satellite and buoy data, we compute the mean bias, mean absolute error (MAE), and root mean square error (RMSE).\n\n---- Helper to compute metrics for any subset ----\n---- Per-station metrics ----\n---- Overall metrics (all stations pooled) ----\n\n```text\nstation n_samples bias MAE RMSE\n0 45001 29 0.32 1.23 1.88\n1 45004 28 -0.12 1.20 1.58\n2 45006 28 0.21 0.95 1.18\n3 ALL 85 0.14 1.13 1.58\n```\n\nOverall, the satellite data exhibit a small warm bias of 0.14 °C. However, the MAE and RMSE are relatively large, at 1.13 °C and 1.58 °C, respectively. A likely source of error is the difference in sampling frequency: the satellite dataset provides roughly one measurement per day, whereas the buoys record about 30 measurements per day. In addition, some days are missing in the satellite data, possibly due to cloud cover. To explore this further, we focus on summer 2016 for buoy 45001 (*see Figure 4*).\n\nWe observe that only two satellite observations are available between late July and mid-August, a particularly warm period, whereas the buoy provides continuous coverage throughout the three summer months (*Figure 4, top pannel*). This limited satellite data availability likely contributes to the satellite-derived mean summer LSWT being approximately 2 °C lower than the buoy-based mean. Indeed, when restricting the buoy data to the 31 measurements closest in time to the 31 satellite observations during summer 2016, the resulting mean LSWT is very similar to the satellite-derived value (*Figure 4, bottom pannel*).\n\nFlatten buoy data and add year/month\nSatellite at buoy location\nBuoy summer series\nMatch satellite and buoy points year by year\n\n--------------------------------------------------\n--------------------------------------------------\nStation coordinates\n--------------------------------------------------\n1. All points (summer) for the station/year\n--------------------------------------------------\nBuoy: all summer points\nSatellite: all summer points at gridcell\nMeans for all points\n--------------------------------------------------\n2. Matched points from df_matched\n--------------------------------------------------\n--------------------------------------------------\n3. Shared y-axis limits\n--------------------------------------------------\n--------------------------------------------------\n--------------------------------------------------\n---- Date formatting ----\n-------------------------\nTop subplot: all summer points\n-------------------------\n-------------------------\nBottom subplot: matched points only\n-------------------------\n---- Shared x-axis formatting ----\n\n*Figure 4 : Comparison of buoy and satellite measurements used to compute the 2016 mean summer LSWT at the location of buoy 45001.*\n\nThis time-matching procedure is repeated for each buoy and year. Specifically, each satellite observation is paired with the nearest buoy measurement in time, provided that the time difference is less than the specified tolerance (default: 1 hour).\n\nThe mean summer LSWT computed from the successfully time-matched satellite–buoy pairs are compared in *Figure 5*. To further quantify differences between the time-matched satellite and buoy data, we compute the mean bias, mean absolute error (MAE), and root mean square error (RMSE).\n\nCompute summer mean per year for each station\n\nPlot buoy data with cross marker\nPlot satellite data with circle marker\n\n*Figure 5: Mean summer LSWT from satellite data compared to buoy data using only successfully time-matched satellite–buoy pairs.*\n\n```text\nstation n_samples bias MAE RMSE\n0 45001 29 0.30 0.43 0.50\n1 45004 28 0.10 0.33 0.39\n2 45006 28 0.26 0.41 0.53\n3 ALL 85 0.22 0.39 0.48\n```", "text_with_prefix": "EQC Quality Assessment: \"Satellite-Derived Monitoring of Summer LSWT in Lake Superior\"\nDataset: satellite-lake-water-temperature [CDS]\nAspect: validation_q02 | Category: Satellite_ECVs\nSection: Satellite-Derived Monitoring of Summer LSWT in Lake Superior > Analysis and results > 4. Comparison with NOAA buoy data\n---\n,\nconvert from Kelvin to Celsius, and compute summer annual means\nNearest satellite pixel to buoy location\nConvert from Kelvin to Celsius\nRestrict to summer months\nCompute annual summer mean\nConcatenate all stations into a single reusable table\nPlot comparison\n\n*Figure 3: Mean summer LSWT for three Lake Superior buoys, compared to the mean summer LSWT from satellite data at the nearest pixel locations.*\n\nWe observe that the satellite data are generally close to the buoy measurements; for instance, the 1998 peak in summer LSWT is well captured. To further quantify the differences between satellite and buoy data, we compute the mean bias, mean absolute error (MAE), and root mean square error (RMSE).\n\n---- Helper to compute metrics for any subset ----\n---- Per-station metrics ----\n---- Overall metrics (all stations pooled) ----\n\n```text\nstation n_samples bias MAE RMSE\n0 45001 29 0.32 1.23 1.88\n1 45004 28 -0.12 1.20 1.58\n2 45006 28 0.21 0.95 1.18\n3 ALL 85 0.14 1.13 1.58\n```\n\nOverall, the satellite data exhibit a small warm bias of 0.14 °C. However, the MAE and RMSE are relatively large, at 1.13 °C and 1.58 °C, respectively. A likely source of error is the difference in sampling frequency: the satellite dataset provides roughly one measurement per day, whereas the buoys record about 30 measurements per day. In addition, some days are missing in the satellite data, possibly due to cloud cover. To explore this further, we focus on summer 2016 for buoy 45001 (*see Figure 4*).\n\nWe observe that only two satellite observations are available between late July and mid-August, a particularly warm period, whereas the buoy provides continuous coverage throughout the three summer months (*Figure 4, top pannel*). This limited satellite data availability likely contributes to the satellite-derived mean summer LSWT being approximately 2 °C lower than the buoy-based mean. Indeed, when restricting the buoy data to the 31 measurements closest in time to the 31 satellite observations during summer 2016, the resulting mean LSWT is very similar to the satellite-derived value (*Figure 4, bottom pannel*).\n\nFlatten buoy data and add year/month\nSatellite at buoy location\nBuoy summer series\nMatch satellite and buoy points year by year\n\n--------------------------------------------------\n--------------------------------------------------\nStation coordinates\n--------------------------------------------------\n1. All points (summer) for the station/year\n--------------------------------------------------\nBuoy: all summer points\nSatellite: all summer points at gridcell\nMeans for all points\n--------------------------------------------------\n2. Matched points from df_matched\n--------------------------------------------------\n--------------------------------------------------\n3. Shared y-axis limits\n--------------------------------------------------\n--------------------------------------------------\n--------------------------------------------------\n---- Date formatting ----\n-------------------------\nTop subplot: all summer points\n-------------------------\n-------------------------\nBottom subplot: matched points only\n-------------------------\n---- Shared x-axis formatting ----\n\n*Figure 4 : Comparison of buoy and satellite measurements used to compute the 2016 mean summer LSWT at the location of buoy 45001.*\n\nThis time-matching procedure is repeated for each buoy and year. Specifically, each satellite observation is paired with the nearest buoy measurement in time, provided that the time difference is less than the specified tolerance (default: 1 hour).\n\nThe mean summer LSWT computed from the successfully time-matched satellite–buoy pairs are compared in *Figure 5*. To further quantify differences between the time-matched satellite and buoy data, we compute the mean bias, mean absolute error (MAE), and root mean square error (RMSE).\n\nCompute summer mean per year for each station\n\nPlot buoy data with cross marker\nPlot satellite data with circle marker\n\n*Figure 5: Mean summer LSWT from satellite data compared to buoy data using only successfully time-matched satellite–buoy pairs.*\n\n```text\nstation n_samples bias MAE RMSE\n0 45001 29 0.30 0.43 0.50\n1 45004 28 0.10 0.33 0.39\n2 45006 28 0.26 0.41 0.53\n3 ALL 85 0.22 0.39 0.48\n```"} {"chunk_id": "satellite_satellite-lake-water-temperature_validation_q02__f7755a658525", "report_id": "satellite_satellite-lake-water-temperature_validation_q02", "dataset_id": "satellite-lake-water-temperature", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q02", "aspect_base": "validation", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite-Derived Monitoring of Summer LSWT in Lake Superior > Analysis and results > 4. Comparison with NOAA buoy data", "title": "Satellite-Derived Monitoring of Summer LSWT in Lake Superior", "chunk_index": 12, "token_count": 341, "text_raw": "2 45006 28 0.26 0.41 0.53\n3 ALL 85 0.22 0.39 0.48\n```\n\nOverall, the satellite data exhibit a small warm bias of 0.22 °C relative to the time-matched buoy observations, slightly larger than the bias obtained when using all available data to compute summer means (0.14 °C). In contrast, the MAE and RMSE are substantially reduced, at 0.39 °C and 0.48 °C, respectively (compared to 1.13 °C and 1.58 °C). This indicates that a large fraction of the difference between the LSWT means shown in *Figure 3* (computed using all data) is attributable to the lower sampling frequency and limited temporal availability of the satellite observations, while individual satellite measurements are comparatively precise. The remaining discrepancies with buoy data are likely due to fundamental measurement differences: satellites measure the lake skin temperature, whereas buoys measure subsurface temperature (typically at ~1 m depth). Depending on atmospheric and lake conditions (e.g., wind and stratification), these measurements can differ. In addition, satellite retrievals may be affected by cloud contamination.", "text_with_prefix": "EQC Quality Assessment: \"Satellite-Derived Monitoring of Summer LSWT in Lake Superior\"\nDataset: satellite-lake-water-temperature [CDS]\nAspect: validation_q02 | Category: Satellite_ECVs\nSection: Satellite-Derived Monitoring of Summer LSWT in Lake Superior > Analysis and results > 4. Comparison with NOAA buoy data\n---\n2 45006 28 0.26 0.41 0.53\n3 ALL 85 0.22 0.39 0.48\n```\n\nOverall, the satellite data exhibit a small warm bias of 0.22 °C relative to the time-matched buoy observations, slightly larger than the bias obtained when using all available data to compute summer means (0.14 °C). In contrast, the MAE and RMSE are substantially reduced, at 0.39 °C and 0.48 °C, respectively (compared to 1.13 °C and 1.58 °C). This indicates that a large fraction of the difference between the LSWT means shown in *Figure 3* (computed using all data) is attributable to the lower sampling frequency and limited temporal availability of the satellite observations, while individual satellite measurements are comparatively precise. The remaining discrepancies with buoy data are likely due to fundamental measurement differences: satellites measure the lake skin temperature, whereas buoys measure subsurface temperature (typically at ~1 m depth). Depending on atmospheric and lake conditions (e.g., wind and stratification), these measurements can differ. In addition, satellite retrievals may be affected by cloud contamination."} {"chunk_id": "satellite_satellite-lake-water-temperature_validation_q02__d80074f15708", "report_id": "satellite_satellite-lake-water-temperature_validation_q02", "dataset_id": "satellite-lake-water-temperature", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q02", "aspect_base": "validation", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite-Derived Monitoring of Summer LSWT in Lake Superior > Analysis and results > 4. Comparison with NOAA buoy data > Estimation of linear trends", "title": "Satellite-Derived Monitoring of Summer LSWT in Lake Superior", "chunk_index": 13, "token_count": 722, "text_raw": "Using all the available buoy data, we estimate linear trends for three periods: 1979–2006, 1995–2023, and 1979–2023 (Figure 6). These are compared with trends derived from satellite data and with previously published results.\n\nFor 1979–2006, the three buoys show warming trends of 0.108 °C/year, 0.098 °C/year, and 0.100 °C/year. Unsurprisingly, these values are consistent with the 0.11 ± 0.06 °C/year estimate reported in [[3]](https://doi.org/10.1029/2006GL029021), which used the same buoy dataset. The associated p-values are relatively small (0.10, 0.30, and 0.07) but fall short of the conventional 0.05 significance threshold, likely due to the limited number of observations.\n\nFor 1995-2023, the three buoys show diverging trends of -0.013 °C/year, 0.035 °C/year, and 0.014 °C/year, with large associated p-values ($\\geq$0.8), indicating that none of these trends are significant. Satellite-derived trends at the corresponding buoy locations (0.003 °C/year, −0.032 °C/year, and −0.004 °C/year) display the same qualitative behavior: diverging and non-significant. These results are consistent with those in *Figure 2* based on spatially averaged satellite data, and with previous analyses [[4]](https://glisa.umich.edu/lake-superior-retrospective/), confirming the absence of a significant summer LSWT warming trend over this period.\n\nFor 1979–2023, the three buoys show warming trends of 0.048 °C/year, 0.061 °C/year, and 0.044 °C/year, with p-values just above the conventional 0.05 significance threshold. These results are consistent with [[6]](https://doi.org/10.1175/JCLI-D-23-0092.1), which reported a warming trend of 0.052±0.031°C/year for 1980-2021 based on simulation data.\n\n--------------------------------------------------\nAdjustable line transparency\n--------------------------------------------------\nBuoy trend periods\nSatellite trend period\n--------------------------------------------------\nHelper function to plot series + trend\n--------------------------------------------------\nAdd trends if periods provided\n--------------------------------------------------\n--------------------------------------------------\n---- Buoy summer means ----\n---- Satellite summer means ----\n---- Axes settings ----\n\n*Figure 6: Time series of mean summer LSWT for buoy and satellite observations at each buoy location. Semi-transparent lines show summer means, and overlaid trend lines (dotted, dashed, or solid) indicate linear trends for different periods, with slope and p-values reported. Trends were estimated using the Theil-Sen method in combination with the Mann-Kendall test for statistical significance.*", "text_with_prefix": "EQC Quality Assessment: \"Satellite-Derived Monitoring of Summer LSWT in Lake Superior\"\nDataset: satellite-lake-water-temperature [CDS]\nAspect: validation_q02 | Category: Satellite_ECVs\nSection: Satellite-Derived Monitoring of Summer LSWT in Lake Superior > Analysis and results > 4. Comparison with NOAA buoy data > Estimation of linear trends\n---\nUsing all the available buoy data, we estimate linear trends for three periods: 1979–2006, 1995–2023, and 1979–2023 (Figure 6). These are compared with trends derived from satellite data and with previously published results.\n\nFor 1979–2006, the three buoys show warming trends of 0.108 °C/year, 0.098 °C/year, and 0.100 °C/year. Unsurprisingly, these values are consistent with the 0.11 ± 0.06 °C/year estimate reported in [[3]](https://doi.org/10.1029/2006GL029021), which used the same buoy dataset. The associated p-values are relatively small (0.10, 0.30, and 0.07) but fall short of the conventional 0.05 significance threshold, likely due to the limited number of observations.\n\nFor 1995-2023, the three buoys show diverging trends of -0.013 °C/year, 0.035 °C/year, and 0.014 °C/year, with large associated p-values ($\\geq$0.8), indicating that none of these trends are significant. Satellite-derived trends at the corresponding buoy locations (0.003 °C/year, −0.032 °C/year, and −0.004 °C/year) display the same qualitative behavior: diverging and non-significant. These results are consistent with those in *Figure 2* based on spatially averaged satellite data, and with previous analyses [[4]](https://glisa.umich.edu/lake-superior-retrospective/), confirming the absence of a significant summer LSWT warming trend over this period.\n\nFor 1979–2023, the three buoys show warming trends of 0.048 °C/year, 0.061 °C/year, and 0.044 °C/year, with p-values just above the conventional 0.05 significance threshold. These results are consistent with [[6]](https://doi.org/10.1175/JCLI-D-23-0092.1), which reported a warming trend of 0.052±0.031°C/year for 1980-2021 based on simulation data.\n\n--------------------------------------------------\nAdjustable line transparency\n--------------------------------------------------\nBuoy trend periods\nSatellite trend period\n--------------------------------------------------\nHelper function to plot series + trend\n--------------------------------------------------\nAdd trends if periods provided\n--------------------------------------------------\n--------------------------------------------------\n---- Buoy summer means ----\n---- Satellite summer means ----\n---- Axes settings ----\n\n*Figure 6: Time series of mean summer LSWT for buoy and satellite observations at each buoy location. Semi-transparent lines show summer means, and overlaid trend lines (dotted, dashed, or solid) indicate linear trends for different periods, with slope and p-values reported. Trends were estimated using the Theil-Sen method in combination with the Mann-Kendall test for statistical significance.*"} {"chunk_id": "satellite_satellite-lake-water-temperature_validation_q02__075931ad6b1c", "report_id": "satellite_satellite-lake-water-temperature_validation_q02", "dataset_id": "satellite-lake-water-temperature", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q02", "aspect_base": "validation", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite-Derived Monitoring of Summer LSWT in Lake Superior > Analysis and results > 4. Comparison with NOAA buoy data > Discussion", "title": "Satellite-Derived Monitoring of Summer LSWT in Lake Superior", "chunk_index": 14, "token_count": 415, "text_raw": "Lake Superior’s summer LSWT trend over 1995–2023, derived from the Copernicus [satellite-lake-water-temperature](https://cds.climate.copernicus.eu/datasets/satellite-lake-water-temperature?tab=overview) dataset, is slightly negative but not statistically significant. Although this initially seemed surprising given sustained global warming, the result was confirmed using in-situ buoy data.\n\nOverall, the satellite data show good agreement with the quality-controlled and time-matched buoy measurements, although the buoy records themselves may contain gaps or suspect observations and should not be interpreted as error-free ground truth. Differences in computed summer means primarily arise from the lower temporal sampling and occasional missing satellite observations, but the systematic bias is small. This indicates that the satellite-lake-water-temperature dataset is suitable for analyzing summer LSWT trends in Lake Superior. However, no significant warming trend is observed over the 1995–2023 period.\n\nOne possible explanation is the unusually high summer temperatures in 1998, associated with a strong El Niño event early in the satellite record. This may have contributed to the relatively large warming trend reported for 1979–2006 [[3]](https://doi.org/10.1029/2006GL029021) and simultaneously obscured longer-term warming signals in the 1995–2023 period. Incorporating pre-1995 in-situ observations, where available, would improve assessments of long-term trends. Indeed, when considering the full buoy record, a near-significant warming trend of ~0.05 °C/year is observed over 1979–2023.", "text_with_prefix": "EQC Quality Assessment: \"Satellite-Derived Monitoring of Summer LSWT in Lake Superior\"\nDataset: satellite-lake-water-temperature [CDS]\nAspect: validation_q02 | Category: Satellite_ECVs\nSection: Satellite-Derived Monitoring of Summer LSWT in Lake Superior > Analysis and results > 4. Comparison with NOAA buoy data > Discussion\n---\nLake Superior’s summer LSWT trend over 1995–2023, derived from the Copernicus [satellite-lake-water-temperature](https://cds.climate.copernicus.eu/datasets/satellite-lake-water-temperature?tab=overview) dataset, is slightly negative but not statistically significant. Although this initially seemed surprising given sustained global warming, the result was confirmed using in-situ buoy data.\n\nOverall, the satellite data show good agreement with the quality-controlled and time-matched buoy measurements, although the buoy records themselves may contain gaps or suspect observations and should not be interpreted as error-free ground truth. Differences in computed summer means primarily arise from the lower temporal sampling and occasional missing satellite observations, but the systematic bias is small. This indicates that the satellite-lake-water-temperature dataset is suitable for analyzing summer LSWT trends in Lake Superior. However, no significant warming trend is observed over the 1995–2023 period.\n\nOne possible explanation is the unusually high summer temperatures in 1998, associated with a strong El Niño event early in the satellite record. This may have contributed to the relatively large warming trend reported for 1979–2006 [[3]](https://doi.org/10.1029/2006GL029021) and simultaneously obscured longer-term warming signals in the 1995–2023 period. Incorporating pre-1995 in-situ observations, where available, would improve assessments of long-term trends. Indeed, when considering the full buoy record, a near-significant warming trend of ~0.05 °C/year is observed over 1979–2023."} {"chunk_id": "satellite_satellite-lake-water-temperature_validation_q02__7d22b2a86d46", "report_id": "satellite_satellite-lake-water-temperature_validation_q02", "dataset_id": "satellite-lake-water-temperature", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q02", "aspect_base": "validation", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite-Derived Monitoring of Summer LSWT in Lake Superior > ℹ️ If you want to know more", "title": "Satellite-Derived Monitoring of Summer LSWT in Lake Superior", "chunk_index": 15, "token_count": 173, "text_raw": "* Farina, G., Thiem, H. (2024) [Several Great Lakes experience record-warm water temperatures heading into winter](https://www.climate.gov/news-features/event-tracker/several-great-lakes-experience-record-warm-water-temperatures-heading#). Climate.gov\n* [NOAA CoastWatch: Great Lakes Node](https://coastwatch.glerl.noaa.gov/statistics/average-surface-water-temperature-glsea/)", "text_with_prefix": "EQC Quality Assessment: \"Satellite-Derived Monitoring of Summer LSWT in Lake Superior\"\nDataset: satellite-lake-water-temperature [CDS]\nAspect: validation_q02 | Category: Satellite_ECVs\nSection: Satellite-Derived Monitoring of Summer LSWT in Lake Superior > ℹ️ If you want to know more\n---\n* Farina, G., Thiem, H. (2024) [Several Great Lakes experience record-warm water temperatures heading into winter](https://www.climate.gov/news-features/event-tracker/several-great-lakes-experience-record-warm-water-temperatures-heading#). Climate.gov\n* [NOAA CoastWatch: Great Lakes Node](https://coastwatch.glerl.noaa.gov/statistics/average-surface-water-temperature-glsea/)"} {"chunk_id": "satellite_satellite-lake-water-temperature_validation_q02__be71b6effc1d", "report_id": "satellite_satellite-lake-water-temperature_validation_q02", "dataset_id": "satellite-lake-water-temperature", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q02", "aspect_base": "validation", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite-Derived Monitoring of Summer LSWT in Lake Superior > ℹ️ If you want to know more > Key resources", "title": "Satellite-Derived Monitoring of Summer LSWT in Lake Superior", "chunk_index": 16, "token_count": 185, "text_raw": "Dataset documentation:\n* [LSWT v4.5: Product User Guide and Specification (PUGS)](https://confluence.ecmwf.int/pages/viewpage.action?pageId=348800177)\n\nCode libraries used:\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)", "text_with_prefix": "EQC Quality Assessment: \"Satellite-Derived Monitoring of Summer LSWT in Lake Superior\"\nDataset: satellite-lake-water-temperature [CDS]\nAspect: validation_q02 | Category: Satellite_ECVs\nSection: Satellite-Derived Monitoring of Summer LSWT in Lake Superior > ℹ️ If you want to know more > Key resources\n---\nDataset documentation:\n* [LSWT v4.5: Product User Guide and Specification (PUGS)](https://confluence.ecmwf.int/pages/viewpage.action?pageId=348800177)\n\nCode libraries used:\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)"} {"chunk_id": "satellite_satellite-lake-water-temperature_validation_q02__edb3fc945acf", "report_id": "satellite_satellite-lake-water-temperature_validation_q02", "dataset_id": "satellite-lake-water-temperature", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q02", "aspect_base": "validation", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite-Derived Monitoring of Summer LSWT in Lake Superior > ℹ️ If you want to know more > References", "title": "Satellite-Derived Monitoring of Summer LSWT in Lake Superior", "chunk_index": 17, "token_count": 542, "text_raw": "* [[1]](https://doi.org/10.1038/s41558-024-02122-y) Wang, X., Shi, K., Qin, B. et al. (2024). Disproportionate impact of atmospheric heat events on lake surface water temperature increases. Nat. Clim. Chang.\n* [[2]](https://climate.copernicus.eu/lake-surface-temperatures) Copernicus (2018). Lake surface temperatures. Climate Change Service.\n* [[3]](https://doi.org/10.1029/2006GL029021) Jay A. Austin, Steven M. Colman (2007). Lake Superior summer water temperatures are increasing more rapidly than regional air temperatures: A positive ice-albedo feedback. Geophysical Research Letters Volume 34, Issue 6. \n* [[4]](https://glisa.umich.edu/lake-superior-retrospective/) GLISA (2024). Lake Superior Retrospective.\n* [[5]](https://coastwatch.glerl.noaa.gov/statistics/average-surface-water-temperature-glsea/) NOAA CoastWatch Great Lakes Regional Node (2024). Average Surface Water Temperature (GLSEA).\n* [[6]](https://doi.org/10.1175/JCLI-D-23-0092.1) Cannon, D., Wang, J., Fujisaki-Manome, A., Kessler, J., Ruberg, S., & Constant, S. (2024). Investigating Multidecadal Trends in Ice Cover and Subsurface Temperatures in the Laurentian Great Lakes Using a Coupled Hydrodynamic–Ice Model. Journal of Climate, 37(4), 1249-1276.\n* [[7]](https://www.ndbc.noaa.gov/) NOAA National Buoy Data Centre.\n* [[8]](https://www.ndbc.noaa.gov/publications/NDBCHandbookofAutomatedDataQualityControl2023.pdf) National Data Buoy Center (NDBC) (2023). Handbook of Automated Data Quality Control Checks and Procedures. NDBC Technical Document T80-10. National Oceanic and Atmospheric Administration (NOAA), National Weather Service, Stennis Space Center, Mississippi, USA.", "text_with_prefix": "EQC Quality Assessment: \"Satellite-Derived Monitoring of Summer LSWT in Lake Superior\"\nDataset: satellite-lake-water-temperature [CDS]\nAspect: validation_q02 | Category: Satellite_ECVs\nSection: Satellite-Derived Monitoring of Summer LSWT in Lake Superior > ℹ️ If you want to know more > References\n---\n* [[1]](https://doi.org/10.1038/s41558-024-02122-y) Wang, X., Shi, K., Qin, B. et al. (2024). Disproportionate impact of atmospheric heat events on lake surface water temperature increases. Nat. Clim. Chang.\n* [[2]](https://climate.copernicus.eu/lake-surface-temperatures) Copernicus (2018). Lake surface temperatures. Climate Change Service.\n* [[3]](https://doi.org/10.1029/2006GL029021) Jay A. Austin, Steven M. Colman (2007). Lake Superior summer water temperatures are increasing more rapidly than regional air temperatures: A positive ice-albedo feedback. Geophysical Research Letters Volume 34, Issue 6. \n* [[4]](https://glisa.umich.edu/lake-superior-retrospective/) GLISA (2024). Lake Superior Retrospective.\n* [[5]](https://coastwatch.glerl.noaa.gov/statistics/average-surface-water-temperature-glsea/) NOAA CoastWatch Great Lakes Regional Node (2024). Average Surface Water Temperature (GLSEA).\n* [[6]](https://doi.org/10.1175/JCLI-D-23-0092.1) Cannon, D., Wang, J., Fujisaki-Manome, A., Kessler, J., Ruberg, S., & Constant, S. (2024). Investigating Multidecadal Trends in Ice Cover and Subsurface Temperatures in the Laurentian Great Lakes Using a Coupled Hydrodynamic–Ice Model. Journal of Climate, 37(4), 1249-1276.\n* [[7]](https://www.ndbc.noaa.gov/) NOAA National Buoy Data Centre.\n* [[8]](https://www.ndbc.noaa.gov/publications/NDBCHandbookofAutomatedDataQualityControl2023.pdf) National Data Buoy Center (NDBC) (2023). Handbook of Automated Data Quality Control Checks and Procedures. NDBC Technical Document T80-10. National Oceanic and Atmospheric Administration (NOAA), National Weather Service, Stennis Space Center, Mississippi, USA."} {"chunk_id": "satellite_satellite-soil-moisture_completeness_q02__03c334273f67", "report_id": "satellite_satellite-soil-moisture_completeness_q02", "dataset_id": "satellite-soil-moisture", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q02", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite soil moisture for drought monitoring in Europe (2023 case study) > Quality assessment question(s)", "title": "Satellite soil moisture for drought monitoring in Europe (2023 case study)", "chunk_index": 0, "token_count": 289, "text_raw": "**Is satellite-derived soil moisture data sufficiently complete, consistent, and temporally representative to reliably detect and monitor anomalies in drought conditions at continental and regional scales?**\n\nThis assessment evaluates the suitability of satellite-derived soil moisture data for drought monitoring in Europe, with a specific focus on data completeness in the context of anomaly-based analysis. Completeness is assessed in terms of the availability of valid observations required to construct a 30-year climatological baseline (1991–2020) and to support dekadal (10-day) anomaly calculations. The analysis does not aim to provide a comprehensive evaluation of all quality dimensions, but rather to examine whether the dataset provides sufficient temporal and spatial support for this specific use case. In this context, completeness is considered a critical factor influencing the reliability of standardized soil moisture anomalies (SSMA), particularly given the uneven temporal distribution of observations across the record. The objective of this assessment is to determine to what extent the dataset provides adequate baseline support to reliably detect and monitor drought conditions at continental and regional scales, using the 2023 European drought as a case study.", "text_with_prefix": "EQC Quality Assessment: \"Satellite soil moisture for drought monitoring in Europe (2023 case study)\"\nDataset: satellite-soil-moisture [CDS]\nAspect: completeness_q02 | Category: Satellite_ECVs\nSection: Satellite soil moisture for drought monitoring in Europe (2023 case study) > Quality assessment question(s)\n---\n**Is satellite-derived soil moisture data sufficiently complete, consistent, and temporally representative to reliably detect and monitor anomalies in drought conditions at continental and regional scales?**\n\nThis assessment evaluates the suitability of satellite-derived soil moisture data for drought monitoring in Europe, with a specific focus on data completeness in the context of anomaly-based analysis. Completeness is assessed in terms of the availability of valid observations required to construct a 30-year climatological baseline (1991–2020) and to support dekadal (10-day) anomaly calculations. The analysis does not aim to provide a comprehensive evaluation of all quality dimensions, but rather to examine whether the dataset provides sufficient temporal and spatial support for this specific use case. In this context, completeness is considered a critical factor influencing the reliability of standardized soil moisture anomalies (SSMA), particularly given the uneven temporal distribution of observations across the record. The objective of this assessment is to determine to what extent the dataset provides adequate baseline support to reliably detect and monitor drought conditions at continental and regional scales, using the 2023 European drought as a case study."} {"chunk_id": "satellite_satellite-soil-moisture_completeness_q02__e4d96215c07e", "report_id": "satellite_satellite-soil-moisture_completeness_q02", "dataset_id": "satellite-soil-moisture", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q02", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite soil moisture for drought monitoring in Europe (2023 case study) > Quality assessment statement", "title": "Satellite soil moisture for drought monitoring in Europe (2023 case study)", "chunk_index": 1, "token_count": 191, "text_raw": "These are the key outcomes of this assessment\n\n* The dataset provides adequate spatial and temporal consistency to capture the main large-scale drought patterns across Europe during the 2023 growing season.\n* Data completeness varies significantly across regions and seasons, with lower availability in areas affected by snow cover, complex terrain, and dense vegetation.\n* Applying stricter baseline completeness thresholds improves the reliability of anomaly estimates but reduces spatial coverage, particularly in regions with historically sparse observations.\n* The dataset is suitable for qualitative drought monitoring and regional-scale analysis, but limitations in baseline representativeness and surface sensitivity should be considered for quantitative applications.\n\n```", "text_with_prefix": "EQC Quality Assessment: \"Satellite soil moisture for drought monitoring in Europe (2023 case study)\"\nDataset: satellite-soil-moisture [CDS]\nAspect: completeness_q02 | Category: Satellite_ECVs\nSection: Satellite soil moisture for drought monitoring in Europe (2023 case study) > Quality assessment statement\n---\nThese are the key outcomes of this assessment\n\n* The dataset provides adequate spatial and temporal consistency to capture the main large-scale drought patterns across Europe during the 2023 growing season.\n* Data completeness varies significantly across regions and seasons, with lower availability in areas affected by snow cover, complex terrain, and dense vegetation.\n* Applying stricter baseline completeness thresholds improves the reliability of anomaly estimates but reduces spatial coverage, particularly in regions with historically sparse observations.\n* The dataset is suitable for qualitative drought monitoring and regional-scale analysis, but limitations in baseline representativeness and surface sensitivity should be considered for quantitative applications.\n\n```"} {"chunk_id": "satellite_satellite-soil-moisture_completeness_q02__993fa2e96182", "report_id": "satellite_satellite-soil-moisture_completeness_q02", "dataset_id": "satellite-soil-moisture", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q02", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite soil moisture for drought monitoring in Europe (2023 case study) > Methodology", "title": "Satellite soil moisture for drought monitoring in Europe (2023 case study)", "chunk_index": 2, "token_count": 674, "text_raw": "The C3S COMBINED satellite soil moisture dataset was downloaded for the period 1991–2023 and spatially subsetted to Europe. The analysis focuses on the volumetric surface soil moisture variable aggregated at dekadal (10-day) resolution. A land mask based on European boundaries was applied to exclude ocean pixels from all analyses.\n\nTo assess the suitability of the dataset for anomaly-based drought monitoring, completeness was evaluated in terms of the availability of valid observations required to construct a 30-year climatological baseline (1991–2020). The number of valid years per pixel and dekad was quantified, and different completeness thresholds were applied (no threshold, ≥50%, ≥70%) to evaluate their impact on anomaly calculations.\n\nSSMA for 2023 were computed at the dekadal scale using the baseline mean and standard deviation for each pixel and day-of-year. The analysis focuses on the March–August period to capture the onset, development, and peak of the 2023 European drought.\n\nSpatial maps were generated for selected dekads to compare anomaly patterns under different completeness thresholds, and predefined regions were used to analyse regional behaviour. For each region, the distribution of SSMA across land pixels was summarized using the median and interquartile range (IQR), and data completeness was quantified as the percentage of valid pixels contributing to each dekad.\n\nThe methodology is designed to evaluate how data availability influences the reliability of anomaly-based drought indicators, rather than to provide a full validation against reference datasets.\n\n**[](section-1)**\n * Import packages\n * Define dataset request and parameters\n * Download and transform soil moisture data\n * Spatially subset to Europe\n * Apply land mask to exclude ocean pixels\n\n**[](section-2)**\n * Compute number of valid years per pixel and dekad (1991–2020)\n * Visualize spatial distribution of baseline support\n * Analyse completeness patterns across regions and seasons\n * Identify areas with limited baseline representativeness\n\n**[](section-3)**\n * Evaluate completeness across years and dekads (1991–2023)\n * Analyse seasonal and interannual variability in data availability\n * Relate completeness patterns to known limitations\n\n**[](section-4)**\n * Compute climatological mean and standard deviation per dekad\n * Apply completeness thresholds (no threshold, ≥50%, ≥70%)\n * Calculate standardized soil moisture anomalies (SSMA) for 2023\n * Mask pixels not meeting threshold criteria\n\n**[](section-5)**\n * Generate maps of SSMA for selected dekads (April–July 3rd dekads)\n * Compare spatial patterns across thresholds\n * Relate patterns to drought evolution reported in literature\n\n**[](section-6)**\n * Define regions of interest\n * Compute median SSMA and IQR over land pixels\n * Quantify completeness (% valid pixels) per dekad\n * Compare temporal evolution across thresholds\n * Interpret results in the context of drought development", "text_with_prefix": "EQC Quality Assessment: \"Satellite soil moisture for drought monitoring in Europe (2023 case study)\"\nDataset: satellite-soil-moisture [CDS]\nAspect: completeness_q02 | Category: Satellite_ECVs\nSection: Satellite soil moisture for drought monitoring in Europe (2023 case study) > Methodology\n---\nThe C3S COMBINED satellite soil moisture dataset was downloaded for the period 1991–2023 and spatially subsetted to Europe. The analysis focuses on the volumetric surface soil moisture variable aggregated at dekadal (10-day) resolution. A land mask based on European boundaries was applied to exclude ocean pixels from all analyses.\n\nTo assess the suitability of the dataset for anomaly-based drought monitoring, completeness was evaluated in terms of the availability of valid observations required to construct a 30-year climatological baseline (1991–2020). The number of valid years per pixel and dekad was quantified, and different completeness thresholds were applied (no threshold, ≥50%, ≥70%) to evaluate their impact on anomaly calculations.\n\nSSMA for 2023 were computed at the dekadal scale using the baseline mean and standard deviation for each pixel and day-of-year. The analysis focuses on the March–August period to capture the onset, development, and peak of the 2023 European drought.\n\nSpatial maps were generated for selected dekads to compare anomaly patterns under different completeness thresholds, and predefined regions were used to analyse regional behaviour. For each region, the distribution of SSMA across land pixels was summarized using the median and interquartile range (IQR), and data completeness was quantified as the percentage of valid pixels contributing to each dekad.\n\nThe methodology is designed to evaluate how data availability influences the reliability of anomaly-based drought indicators, rather than to provide a full validation against reference datasets.\n\n**[](section-1)**\n * Import packages\n * Define dataset request and parameters\n * Download and transform soil moisture data\n * Spatially subset to Europe\n * Apply land mask to exclude ocean pixels\n\n**[](section-2)**\n * Compute number of valid years per pixel and dekad (1991–2020)\n * Visualize spatial distribution of baseline support\n * Analyse completeness patterns across regions and seasons\n * Identify areas with limited baseline representativeness\n\n**[](section-3)**\n * Evaluate completeness across years and dekads (1991–2023)\n * Analyse seasonal and interannual variability in data availability\n * Relate completeness patterns to known limitations\n\n**[](section-4)**\n * Compute climatological mean and standard deviation per dekad\n * Apply completeness thresholds (no threshold, ≥50%, ≥70%)\n * Calculate standardized soil moisture anomalies (SSMA) for 2023\n * Mask pixels not meeting threshold criteria\n\n**[](section-5)**\n * Generate maps of SSMA for selected dekads (April–July 3rd dekads)\n * Compare spatial patterns across thresholds\n * Relate patterns to drought evolution reported in literature\n\n**[](section-6)**\n * Define regions of interest\n * Compute median SSMA and IQR over land pixels\n * Quantify completeness (% valid pixels) per dekad\n * Compare temporal evolution across thresholds\n * Interpret results in the context of drought development"} {"chunk_id": "satellite_satellite-soil-moisture_completeness_q02__856b111f3d07", "report_id": "satellite_satellite-soil-moisture_completeness_q02", "dataset_id": "satellite-soil-moisture", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q02", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite soil moisture for drought monitoring in Europe (2023 case study) > Analysis and results > 1. Request and download data > Download data", "title": "Satellite soil moisture for drought monitoring in Europe (2023 case study)", "chunk_index": 3, "token_count": 159, "text_raw": "Administrative boundaries were obtained from the Eurostat GISCO database, ensuring consistency with official European statistical geographies [[1]](https://ec.europa.eu/eurostat/web/gisco/geodata/administrative-units).\n\n```text\n100%|██████████| 31/31 [00:07<00:00, 4.30it/s]\n```", "text_with_prefix": "EQC Quality Assessment: \"Satellite soil moisture for drought monitoring in Europe (2023 case study)\"\nDataset: satellite-soil-moisture [CDS]\nAspect: completeness_q02 | Category: Satellite_ECVs\nSection: Satellite soil moisture for drought monitoring in Europe (2023 case study) > Analysis and results > 1. Request and download data > Download data\n---\nAdministrative boundaries were obtained from the Eurostat GISCO database, ensuring consistency with official European statistical geographies [[1]](https://ec.europa.eu/eurostat/web/gisco/geodata/administrative-units).\n\n```text\n100%|██████████| 31/31 [00:07<00:00, 4.30it/s]\n```"} {"chunk_id": "satellite_satellite-soil-moisture_completeness_q02__5e9aa65a1262", "report_id": "satellite_satellite-soil-moisture_completeness_q02", "dataset_id": "satellite-soil-moisture", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q02", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite soil moisture for drought monitoring in Europe (2023 case study) > Analysis and results > 1. Request and download data > Spatially subset to Europe", "title": "Satellite soil moisture for drought monitoring in Europe (2023 case study)", "chunk_index": 4, "token_count": 145, "text_raw": "Select soil mouisture variable\n\nFilter the shapefile to keep only Europe\nEU + nearby Europe (adjust if needed)\nEnsure soil_moisture has CRS information compatible with europe_shape (e.g., EPSG:4326)\nClip soil moisture data to the Europe shape\n\n(section-2)=", "text_with_prefix": "EQC Quality Assessment: \"Satellite soil moisture for drought monitoring in Europe (2023 case study)\"\nDataset: satellite-soil-moisture [CDS]\nAspect: completeness_q02 | Category: Satellite_ECVs\nSection: Satellite soil moisture for drought monitoring in Europe (2023 case study) > Analysis and results > 1. Request and download data > Spatially subset to Europe\n---\nSelect soil mouisture variable\n\nFilter the shapefile to keep only Europe\nEU + nearby Europe (adjust if needed)\nEnsure soil_moisture has CRS information compatible with europe_shape (e.g., EPSG:4326)\nClip soil moisture data to the Europe shape\n\n(section-2)="} {"chunk_id": "satellite_satellite-soil-moisture_completeness_q02__e88f80f5c273", "report_id": "satellite_satellite-soil-moisture_completeness_q02", "dataset_id": "satellite-soil-moisture", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q02", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite soil moisture for drought monitoring in Europe (2023 case study) > Analysis and results > 2. Baseline completeness assessment", "title": "Satellite soil moisture for drought monitoring in Europe (2023 case study)", "chunk_index": 5, "token_count": 710, "text_raw": "To assess whether the dataset provides sufficient baseline support for computing standardized soil moisture anomalies (SSMA), the number of valid years available for each pixel and dekad was quantified over the 1991–2020 reference period, which corresponds to the standard 30-year climatological baseline recommended by the World Meteorological Organization (WMO)[[2]](https://community.wmo.int/site/knowledge-hub/programmes-and-initiatives/climate-services/wmo-climatological-normals). The most recent years (2021–2022) were not included to ensure consistency with established climatological reference periods and to avoid mixing the baseline with the period of analysis. In addition, recent years—particularly 2022—were characterized by pronounced drought conditions in Europe [[3]](https://doi.org/10.2760/998985); including such anomalous years in the baseline could bias the climatological mean and standard deviation, thereby reducing the sensitivity of the anomaly calculation.\n\nThe soil moisture time series was first subset to the baseline period and grouped by dekad of the year (36 dekads per year). For each dekad and pixel, the number of years containing a valid soil moisture value was counted. This results in a spatially explicit estimate of baseline sample size, ranging from 0 to 30 years, which represents the number of observations available to compute the climatological mean and standard deviation for that dekad. The resulting maps show the distribution of baseline support across selected dekads of the year.\n\nThis analysis is performed to evaluate whether the available baseline data are sufficient to support robust anomaly calculations. Although the dataset is provided as dekadal aggregates, the effective sample size underlying the climatology can vary substantially across space and time. In some regions and dekads, only a limited number of valid years are available, which may reduce the reliability of the derived anomalies. Therefore, assessing baseline support is a useful step to interpret the SSMA results within the scope of this drought monitoring use case.\n\nThis assessment focuses on the availability of valid dekadal values across years, rather than the number of daily observations contributing to each dekadal aggregate, as the anomaly calculation depends primarily on the robustness of the baseline climatology.\n\nKeep only the baseline period\nBuild a dekad-of-year index from the timestamp\nday 1 -> dekad 1 of month\nday 11 -> dekad 2 of month\nday 21 -> dekad 3 of month\nThen combine months into 1..36 across the year\nAttach year and dekad_of_year as coordinates\nCount valid observations across years for each dekad-of-year and pixel\nResult dims: dekad_of_year, latitude, longitude\nEnsure integer type for plotting / storage\n\nRepresentative dekads through the year\nOptional labels\nOptional: overlay borders if world_shape already exists\nEurope-focused extent; adjust if needed\nHide any unused axes\n\n*Figure 1. Number of valid baseline years per pixel (1991–2020) for selected dekads of the year.*\n\n(section-3)=", "text_with_prefix": "EQC Quality Assessment: \"Satellite soil moisture for drought monitoring in Europe (2023 case study)\"\nDataset: satellite-soil-moisture [CDS]\nAspect: completeness_q02 | Category: Satellite_ECVs\nSection: Satellite soil moisture for drought monitoring in Europe (2023 case study) > Analysis and results > 2. Baseline completeness assessment\n---\nTo assess whether the dataset provides sufficient baseline support for computing standardized soil moisture anomalies (SSMA), the number of valid years available for each pixel and dekad was quantified over the 1991–2020 reference period, which corresponds to the standard 30-year climatological baseline recommended by the World Meteorological Organization (WMO)[[2]](https://community.wmo.int/site/knowledge-hub/programmes-and-initiatives/climate-services/wmo-climatological-normals). The most recent years (2021–2022) were not included to ensure consistency with established climatological reference periods and to avoid mixing the baseline with the period of analysis. In addition, recent years—particularly 2022—were characterized by pronounced drought conditions in Europe [[3]](https://doi.org/10.2760/998985); including such anomalous years in the baseline could bias the climatological mean and standard deviation, thereby reducing the sensitivity of the anomaly calculation.\n\nThe soil moisture time series was first subset to the baseline period and grouped by dekad of the year (36 dekads per year). For each dekad and pixel, the number of years containing a valid soil moisture value was counted. This results in a spatially explicit estimate of baseline sample size, ranging from 0 to 30 years, which represents the number of observations available to compute the climatological mean and standard deviation for that dekad. The resulting maps show the distribution of baseline support across selected dekads of the year.\n\nThis analysis is performed to evaluate whether the available baseline data are sufficient to support robust anomaly calculations. Although the dataset is provided as dekadal aggregates, the effective sample size underlying the climatology can vary substantially across space and time. In some regions and dekads, only a limited number of valid years are available, which may reduce the reliability of the derived anomalies. Therefore, assessing baseline support is a useful step to interpret the SSMA results within the scope of this drought monitoring use case.\n\nThis assessment focuses on the availability of valid dekadal values across years, rather than the number of daily observations contributing to each dekadal aggregate, as the anomaly calculation depends primarily on the robustness of the baseline climatology.\n\nKeep only the baseline period\nBuild a dekad-of-year index from the timestamp\nday 1 -> dekad 1 of month\nday 11 -> dekad 2 of month\nday 21 -> dekad 3 of month\nThen combine months into 1..36 across the year\nAttach year and dekad_of_year as coordinates\nCount valid observations across years for each dekad-of-year and pixel\nResult dims: dekad_of_year, latitude, longitude\nEnsure integer type for plotting / storage\n\nRepresentative dekads through the year\nOptional labels\nOptional: overlay borders if world_shape already exists\nEurope-focused extent; adjust if needed\nHide any unused axes\n\n*Figure 1. Number of valid baseline years per pixel (1991–2020) for selected dekads of the year.*\n\n(section-3)="} {"chunk_id": "satellite_satellite-soil-moisture_completeness_q02__07e0d493810c", "report_id": "satellite_satellite-soil-moisture_completeness_q02", "dataset_id": "satellite-soil-moisture", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q02", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite soil moisture for drought monitoring in Europe (2023 case study) > Analysis and results > 3. Temporal completeness analysis", "title": "Satellite soil moisture for drought monitoring in Europe (2023 case study)", "chunk_index": 6, "token_count": 480, "text_raw": "Completeness was calculated for each year–dekad combination between 1991 and 2023 as the percentage of valid land pixels within the study domain. A land mask derived from the European land geometry was applied so that only inland pixels were included. The resulting heatmap shows how data availability changes through both the seasonal cycle and the observational record, and helps identify whether low completeness is concentrated in specific years or dekads.\n\nLand mask from first timestep\nKeep only land pixels\nConvert dekad number (1..36) to month and dekad-in-month\n\n*Figure 2. Heatmap of completeness over European land pixels by year and dekad (1991–2023), expressed as the percentage of valid observations.*\n\nThe results presented in Figures 1 and 2 are consistent with the spatial and temporal completeness assessment described in the Product Quality Assessment Report [(PQAR)](https://confluence.ecmwf.int/pages/viewpage.action?pageId=445290728).\n\nFigure 1 shows that completeness is lower in regions affected by seasonal soil moisture freeze–thaw processes, particularly at higher latitudes and during the winter season. Reduced completeness is also observed in mountainous areas, as well as in some urban regions (e.g. around London), which may be influenced by radio-frequency interference (RFI).\n\nFigure 2 indicates a clear seasonal signal, with generally higher completeness across Europe (approximately between 35°N and 70°N) during the May–September period. It can also be observed that completeness during the spring–summer months is lower in earlier years of the record, particularly prior to 2007. This is consistent with Figure 4 of the PQAR, which highlights that the addition of AMSR-E (from 2002) and ASCAT (from 2007) significantly improved spatial coverage. This improvement is reflected in Figure 2, where a noticeable increase in completeness is observed following the introduction of ASCAT in 2007.\n\n(section-4)=", "text_with_prefix": "EQC Quality Assessment: \"Satellite soil moisture for drought monitoring in Europe (2023 case study)\"\nDataset: satellite-soil-moisture [CDS]\nAspect: completeness_q02 | Category: Satellite_ECVs\nSection: Satellite soil moisture for drought monitoring in Europe (2023 case study) > Analysis and results > 3. Temporal completeness analysis\n---\nCompleteness was calculated for each year–dekad combination between 1991 and 2023 as the percentage of valid land pixels within the study domain. A land mask derived from the European land geometry was applied so that only inland pixels were included. The resulting heatmap shows how data availability changes through both the seasonal cycle and the observational record, and helps identify whether low completeness is concentrated in specific years or dekads.\n\nLand mask from first timestep\nKeep only land pixels\nConvert dekad number (1..36) to month and dekad-in-month\n\n*Figure 2. Heatmap of completeness over European land pixels by year and dekad (1991–2023), expressed as the percentage of valid observations.*\n\nThe results presented in Figures 1 and 2 are consistent with the spatial and temporal completeness assessment described in the Product Quality Assessment Report [(PQAR)](https://confluence.ecmwf.int/pages/viewpage.action?pageId=445290728).\n\nFigure 1 shows that completeness is lower in regions affected by seasonal soil moisture freeze–thaw processes, particularly at higher latitudes and during the winter season. Reduced completeness is also observed in mountainous areas, as well as in some urban regions (e.g. around London), which may be influenced by radio-frequency interference (RFI).\n\nFigure 2 indicates a clear seasonal signal, with generally higher completeness across Europe (approximately between 35°N and 70°N) during the May–September period. It can also be observed that completeness during the spring–summer months is lower in earlier years of the record, particularly prior to 2007. This is consistent with Figure 4 of the PQAR, which highlights that the addition of AMSR-E (from 2002) and ASCAT (from 2007) significantly improved spatial coverage. This improvement is reflected in Figure 2, where a noticeable increase in completeness is observed following the introduction of ASCAT in 2007.\n\n(section-4)="} {"chunk_id": "satellite_satellite-soil-moisture_completeness_q02__9f9edc3111a1", "report_id": "satellite_satellite-soil-moisture_completeness_q02", "dataset_id": "satellite-soil-moisture", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q02", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite soil moisture for drought monitoring in Europe (2023 case study) > Analysis and results > 4. SSMA computation under different thresholds", "title": "Satellite soil moisture for drought monitoring in Europe (2023 case study)", "chunk_index": 7, "token_count": 705, "text_raw": "The SSMA was calculated following a similar approach as the one explained by the European Drought Observatory (EDO) [[4]](https://drought.emergency.copernicus.eu/data/factsheets/factsheet_soilmoisture.pdf). Surface soil moisture values for each pixel and each dekad of the year 2023 were compared to the long-term climatological period (1991–2020, 30 years). The anomaly was computed using the following equation:\n\n$\\text{SMA} = \\frac{\\text{SMI}_t - \\overline{\\text{SMI}}}{\\delta_{\\text{SMI}}}$\n\nwhere:\n\n$\\text{SMI}_t$ = Surface soil moisture 2023 for a specific dekad for that pixel\n\n$\\overline{\\text{SMI}}$ = Mean surface soil moisture 1991-2020 for that same dekad for that pixel\n\n$\\delta_{\\text{SMI}}$ = Standard deviation surface soil moisture 1991-2020 for that same dekad for that pixel\n\nFurthermore, based on the completeness analysis presented in the previous section, a minimum baseline-availability threshold was introduced in the SSMA calculation. For each pixel and each dekad, the number of years within the 1991–2020 baseline period containing a valid soil moisture value was counted. A threshold was then applied requiring that a minimum fraction of these 30 years be available in order to compute the climatological mean and standard deviation.\n\nSpecifically, thresholds of 50% and 70% were tested, corresponding to at least 15 and 21 valid years, respectively, for a given pixel–dekad combination. If the number of valid years did not meet the selected threshold, the anomaly for that pixel and dekad was masked and not included in the analysis. A no-threshold case was also retained for comparison, where anomalies are computed regardless of baseline availability.\n\nThe threshold is applied at the dekadal level across years and does not account for the number of daily observations contributing to each dekadal aggregate. This is consistent with common practice in drought monitoring applications, where anomalies are typically computed from dekadal products, as also adopted in Joint Research Centre (JRC) Global Drought Observatory (GDO) analyses [[3]](https://doi.org/10.2760/998985) [[5]](https://doi.org/10.2760/575433) [[6]](https://doi.org/10.2760/928418). Given the focus of this use case on dekadal-scale drought conditions, this approach was considered appropriate.\n\nSelect long-term (baseline) and target period\nGroup baseline by day-of-year\nCount valid (non-NaN) years for each DOY at each pixel\nTotal number of years in the baseline\nBuild completeness mask\nCompute long-term mean and std\nAvoid division by zero\nGroup target year by DOY\nCompute standardized anomaly\nApply completeness mask\nRemove inf values\n\n(section-5)=", "text_with_prefix": "EQC Quality Assessment: \"Satellite soil moisture for drought monitoring in Europe (2023 case study)\"\nDataset: satellite-soil-moisture [CDS]\nAspect: completeness_q02 | Category: Satellite_ECVs\nSection: Satellite soil moisture for drought monitoring in Europe (2023 case study) > Analysis and results > 4. SSMA computation under different thresholds\n---\nThe SSMA was calculated following a similar approach as the one explained by the European Drought Observatory (EDO) [[4]](https://drought.emergency.copernicus.eu/data/factsheets/factsheet_soilmoisture.pdf). Surface soil moisture values for each pixel and each dekad of the year 2023 were compared to the long-term climatological period (1991–2020, 30 years). The anomaly was computed using the following equation:\n\n$\\text{SMA} = \\frac{\\text{SMI}_t - \\overline{\\text{SMI}}}{\\delta_{\\text{SMI}}}$\n\nwhere:\n\n$\\text{SMI}_t$ = Surface soil moisture 2023 for a specific dekad for that pixel\n\n$\\overline{\\text{SMI}}$ = Mean surface soil moisture 1991-2020 for that same dekad for that pixel\n\n$\\delta_{\\text{SMI}}$ = Standard deviation surface soil moisture 1991-2020 for that same dekad for that pixel\n\nFurthermore, based on the completeness analysis presented in the previous section, a minimum baseline-availability threshold was introduced in the SSMA calculation. For each pixel and each dekad, the number of years within the 1991–2020 baseline period containing a valid soil moisture value was counted. A threshold was then applied requiring that a minimum fraction of these 30 years be available in order to compute the climatological mean and standard deviation.\n\nSpecifically, thresholds of 50% and 70% were tested, corresponding to at least 15 and 21 valid years, respectively, for a given pixel–dekad combination. If the number of valid years did not meet the selected threshold, the anomaly for that pixel and dekad was masked and not included in the analysis. A no-threshold case was also retained for comparison, where anomalies are computed regardless of baseline availability.\n\nThe threshold is applied at the dekadal level across years and does not account for the number of daily observations contributing to each dekadal aggregate. This is consistent with common practice in drought monitoring applications, where anomalies are typically computed from dekadal products, as also adopted in Joint Research Centre (JRC) Global Drought Observatory (GDO) analyses [[3]](https://doi.org/10.2760/998985) [[5]](https://doi.org/10.2760/575433) [[6]](https://doi.org/10.2760/928418). Given the focus of this use case on dekadal-scale drought conditions, this approach was considered appropriate.\n\nSelect long-term (baseline) and target period\nGroup baseline by day-of-year\nCount valid (non-NaN) years for each DOY at each pixel\nTotal number of years in the baseline\nBuild completeness mask\nCompute long-term mean and std\nAvoid division by zero\nGroup target year by DOY\nCompute standardized anomaly\nApply completeness mask\nRemove inf values\n\n(section-5)="} {"chunk_id": "satellite_satellite-soil-moisture_completeness_q02__feade28f9b9b", "report_id": "satellite_satellite-soil-moisture_completeness_q02", "dataset_id": "satellite-soil-moisture", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q02", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite soil moisture for drought monitoring in Europe (2023 case study) > Analysis and results > 5. Spatial analysis of anomalies", "title": "Satellite soil moisture for drought monitoring in Europe (2023 case study)", "chunk_index": 8, "token_count": 945, "text_raw": "To support the interpretation of the SSMA results, spatial maps were generated for the third dekad of April, May, June, and July 2023. These four dekads were selected to enable direct comparison with the corresponding multi-panel figure presented in the JRC August 2023 drought report [[6]](https://doi.org/10.2760/928418), which shows the evolution of soil moisture anomaly conditions during late spring and early summer. The standardized anomalies were mapped over Europe using a consistent color scale, and predefined regions of interest were delineated to guide the subsequent regional analysis. The same figure layout was produced for different baseline-availability thresholds in order to visually compare their effect on the spatial patterns of the anomalies.\n\nThe comparison with JRC drought products should be interpreted as a qualitative assessment of pattern consistency rather than a strict validation. The C3S COMBINED dataset represents surface soil moisture, while the JRC Soil Moisture Anomaly is derived from LISFLOOD simulations of root-zone soil moisture [[4]](https://drought.emergency.copernicus.eu/data/factsheets/factsheet_soilmoisture.pdf). These differences in depth and methodology imply distinct temporal responses and spatial characteristics, and therefore some discrepancies are expected.\n\nDefine custom colormap and boundaries\nCreate colormap and norm\n\nThird dekad of April, May, June, July\nLat/lon ticks\nManual spacing instead of tight_layout\nColorbar placed lower\nMain title + year\n\n*Figure 3. SSMA over Europe for the third dekad of April, May, June, and July 2023. Anomalies are computed without applying a baseline completeness threshold. Boxes indicate the regions selected for subsequent regional analysis.*\n\n*Figure 4. SSMA over Europe for the third dekad of April, May, June, and July 2023. Anomalies are computed using a baseline completeness threshold of ≥50%. Boxes indicate the regions selected for subsequent regional analysis.*\n\n*Figure 5. SSMA over Europe for the third dekad of April, May, June, and July 2023. Anomalies are computed using a baseline completeness threshold of ≥70%. Boxes indicate the regions selected for subsequent regional analysis.*\n\n![JRGaugust23.jpg](attachment:004a20bf-a633-43eb-b74a-9cbebdc4ce91.jpg)\n\n*Figure 6. Soil Moisture Anomaly from April to July 2023. Extracted from: JRC (2023)[[6]](https://doi.org/10.2760/928418)*\n\nThe application of a baseline completeness threshold has a clear impact on the spatial coverage of the SSMA maps. Under the ≥70% threshold, missing values become more prominent in regions where long-term data availability is limited, particularly in areas with complex terrain such as the Pyrenees, Alps, Apennines, Dinaric Alps, Balkan Mountains, and Carpathians. Reduced coverage is also observed in parts of northern Spain and southern UK, potentially reflecting a combination of retrieval limitations over heterogeneous terrain and urban areas.\n\nCompared to the no-threshold and ≥50% cases, the ≥70% threshold leads to a noticeable reduction in spatial coverage over southern and southeastern Europe, especially across Italy and the Balkans. In these regions, areas that appear as neutral or positive anomalies in the lower-threshold cases are masked out when stricter baseline requirements are applied. This highlights the sensitivity of the spatial patterns to baseline data availability.\n\nDespite these differences, several large-scale features remain consistent with the patterns reported in the JRC drought analyses. In particular, persistent negative anomalies are observed across Scandinavia, and a transition towards more neutral or slightly positive anomalies is visible over parts of the British Isles in late July. This suggests that, even with reduced baseline support in some regions, the main spatial signals are broadly captured.\n\nThe largest discrepancies with respect to the JRC results are observed over central-western Europe (e.g. Benelux, Germany, and surrounding areas), where the satellite-based anomalies show weaker or more spatially fragmented drought signals. These differences may reflect both the sensitivity of surface soil moisture to short-term variability and the impact of data completeness on the anomaly calculation.\n\n(section-6)=", "text_with_prefix": "EQC Quality Assessment: \"Satellite soil moisture for drought monitoring in Europe (2023 case study)\"\nDataset: satellite-soil-moisture [CDS]\nAspect: completeness_q02 | Category: Satellite_ECVs\nSection: Satellite soil moisture for drought monitoring in Europe (2023 case study) > Analysis and results > 5. Spatial analysis of anomalies\n---\nTo support the interpretation of the SSMA results, spatial maps were generated for the third dekad of April, May, June, and July 2023. These four dekads were selected to enable direct comparison with the corresponding multi-panel figure presented in the JRC August 2023 drought report [[6]](https://doi.org/10.2760/928418), which shows the evolution of soil moisture anomaly conditions during late spring and early summer. The standardized anomalies were mapped over Europe using a consistent color scale, and predefined regions of interest were delineated to guide the subsequent regional analysis. The same figure layout was produced for different baseline-availability thresholds in order to visually compare their effect on the spatial patterns of the anomalies.\n\nThe comparison with JRC drought products should be interpreted as a qualitative assessment of pattern consistency rather than a strict validation. The C3S COMBINED dataset represents surface soil moisture, while the JRC Soil Moisture Anomaly is derived from LISFLOOD simulations of root-zone soil moisture [[4]](https://drought.emergency.copernicus.eu/data/factsheets/factsheet_soilmoisture.pdf). These differences in depth and methodology imply distinct temporal responses and spatial characteristics, and therefore some discrepancies are expected.\n\nDefine custom colormap and boundaries\nCreate colormap and norm\n\nThird dekad of April, May, June, July\nLat/lon ticks\nManual spacing instead of tight_layout\nColorbar placed lower\nMain title + year\n\n*Figure 3. SSMA over Europe for the third dekad of April, May, June, and July 2023. Anomalies are computed without applying a baseline completeness threshold. Boxes indicate the regions selected for subsequent regional analysis.*\n\n*Figure 4. SSMA over Europe for the third dekad of April, May, June, and July 2023. Anomalies are computed using a baseline completeness threshold of ≥50%. Boxes indicate the regions selected for subsequent regional analysis.*\n\n*Figure 5. SSMA over Europe for the third dekad of April, May, June, and July 2023. Anomalies are computed using a baseline completeness threshold of ≥70%. Boxes indicate the regions selected for subsequent regional analysis.*\n\n![JRGaugust23.jpg](attachment:004a20bf-a633-43eb-b74a-9cbebdc4ce91.jpg)\n\n*Figure 6. Soil Moisture Anomaly from April to July 2023. Extracted from: JRC (2023)[[6]](https://doi.org/10.2760/928418)*\n\nThe application of a baseline completeness threshold has a clear impact on the spatial coverage of the SSMA maps. Under the ≥70% threshold, missing values become more prominent in regions where long-term data availability is limited, particularly in areas with complex terrain such as the Pyrenees, Alps, Apennines, Dinaric Alps, Balkan Mountains, and Carpathians. Reduced coverage is also observed in parts of northern Spain and southern UK, potentially reflecting a combination of retrieval limitations over heterogeneous terrain and urban areas.\n\nCompared to the no-threshold and ≥50% cases, the ≥70% threshold leads to a noticeable reduction in spatial coverage over southern and southeastern Europe, especially across Italy and the Balkans. In these regions, areas that appear as neutral or positive anomalies in the lower-threshold cases are masked out when stricter baseline requirements are applied. This highlights the sensitivity of the spatial patterns to baseline data availability.\n\nDespite these differences, several large-scale features remain consistent with the patterns reported in the JRC drought analyses. In particular, persistent negative anomalies are observed across Scandinavia, and a transition towards more neutral or slightly positive anomalies is visible over parts of the British Isles in late July. This suggests that, even with reduced baseline support in some regions, the main spatial signals are broadly captured.\n\nThe largest discrepancies with respect to the JRC results are observed over central-western Europe (e.g. Benelux, Germany, and surrounding areas), where the satellite-based anomalies show weaker or more spatially fragmented drought signals. These differences may reflect both the sensitivity of surface soil moisture to short-term variability and the impact of data completeness on the anomaly calculation.\n\n(section-6)="} {"chunk_id": "satellite_satellite-soil-moisture_completeness_q02__599f86210441", "report_id": "satellite_satellite-soil-moisture_completeness_q02", "dataset_id": "satellite-soil-moisture", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q02", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite soil moisture for drought monitoring in Europe (2023 case study) > Analysis and results > 6. Regional SSMA analysis", "title": "Satellite soil moisture for drought monitoring in Europe (2023 case study)", "chunk_index": 9, "token_count": 1065, "text_raw": "The following section computes and visualizes regional SSMA statistics for selected European subregions. For each region, SSMA values are spatially subsetted to the corresponding bounding box and restricted to land pixels using a geographic mask. The analysis focuses on the period from March to August 2023, using dekadal time steps.\n\nThis period was selected to capture the onset, development, and peak of the 2023 European drought, while maintaining consistency with the temporal coverage of the JRC drought reports, which provide assessments for March [[3]](https://doi.org/10.2760/998985), June [[5]](https://doi.org/10.2760/575433), and August [[6]](https://doi.org/10.2760/928418). Focusing on this interval allows for a targeted evaluation of the dataset’s ability to represent the main phases of the drought event, while avoiding the inclusion of less relevant periods outside the core drought season.\n\nWithin each region and dekad, the distribution of SSMA across all land pixels is summarized using the median and interquartile range (IQR; 25th–75th percentiles), providing a robust representation of central tendency and spatial variability. In parallel, data completeness is quantified as the percentage of valid (non-missing) land pixels contributing to each dekad.\n\nThese statistics are computed separately for different baseline completeness thresholds (no threshold, ≥50%, ≥70%) to assess the sensitivity of the SSMA signal to data availability. The resulting time series are displayed as multi-panel figures, allowing direct comparison of SSMA behaviour and completeness across thresholds and regions.\n\nSubset region\nKeep only land pixels\nStack spatial dims\nCompleteness over land only\nLeft axis: median + IQR\nRight axis: completeness\nX ticks on dekad starts\nOnly one combined legend, placed in first panel\n\n*Figure 7. Mean SSMA over the Iberian Peninsula from spring to summer 2023, together with the corresponding data completeness (%). Panels (left to right) show results for different completeness thresholds: no threshold, ≥50%, and ≥70%.*\n\nOver the Iberian Peninsula, SSMA shows a clear seasonal evolution, with the most negative values occurring in mid-May, followed by a rapid transition towards positive anomalies in early June and a gradual stabilization towards near-neutral conditions during July and August. This pattern reflects the transition from peak drought conditions in late spring to partial recovery in early summer.\n\nThe temporal evolution of the median SSMA is highly consistent across all completeness thresholds, with only minor differences in magnitude. This indicates that the regional drought signal is robust to the choice of baseline filtering.\n\nData completeness remains high throughout the period. For the no-threshold and ≥50% cases, completeness exceeds 90% across most dekads, indicating strong spatial support. The ≥70% threshold results in slightly lower completeness (around 80%), but still maintains sufficient coverage to support the analysis.\n\nOverall, the strong agreement across thresholds, combined with high data availability, suggests that the SSMA estimates for the Iberian Peninsula are both stable and well-supported by the underlying data.\n\nThis evolution is consistent with the progression of drought conditions reported for the western Mediterranean in spring 2023, followed by partial recovery in early summer [[3]](https://doi.org/10.2760/998985), [[5]](https://doi.org/10.2760/575433), [[6]](https://doi.org/10.2760/928418), [[7]](https://joint-research-centre.ec.europa.eu/jrc-news-and-updates/severe-drought-western-mediterranean-faces-low-river-flows-and-crop-yields-earlier-ever-2023-06-13_en), [[8]](https://climate.copernicus.eu/precipitation-relative-humidity-and-soil-moisture-may-2023), and with studies highlighting the role of persistent soil moisture deficits in amplifying the record-breaking spring heatwave in the region [[9]](https://doi.org/10.1038/s41612-024-00569-6).\n\n*Figure 8. Mean SSMA over Great Britain and Ireland from spring to summer 2023, together with the corresponding data completeness (%). Panels (left to right) show results for different completeness thresholds: no threshold, ≥50%, and ≥70%.*\n\nOver Great Britain and Ireland, SSMA exhibits a clear minimum in June, indicating the peak of dry conditions, followed by a gradual recovery towards neutral and slight possitive anomalies conditions in July and August. The magnitude of the negative anomaly is slightly stronger when applying the ≥70% threshold, suggesting that stricter filtering removes pixels with limited baseline support, which can reduce noise and result in slightly stronger anomaly signals.", "text_with_prefix": "EQC Quality Assessment: \"Satellite soil moisture for drought monitoring in Europe (2023 case study)\"\nDataset: satellite-soil-moisture [CDS]\nAspect: completeness_q02 | Category: Satellite_ECVs\nSection: Satellite soil moisture for drought monitoring in Europe (2023 case study) > Analysis and results > 6. Regional SSMA analysis\n---\nThe following section computes and visualizes regional SSMA statistics for selected European subregions. For each region, SSMA values are spatially subsetted to the corresponding bounding box and restricted to land pixels using a geographic mask. The analysis focuses on the period from March to August 2023, using dekadal time steps.\n\nThis period was selected to capture the onset, development, and peak of the 2023 European drought, while maintaining consistency with the temporal coverage of the JRC drought reports, which provide assessments for March [[3]](https://doi.org/10.2760/998985), June [[5]](https://doi.org/10.2760/575433), and August [[6]](https://doi.org/10.2760/928418). Focusing on this interval allows for a targeted evaluation of the dataset’s ability to represent the main phases of the drought event, while avoiding the inclusion of less relevant periods outside the core drought season.\n\nWithin each region and dekad, the distribution of SSMA across all land pixels is summarized using the median and interquartile range (IQR; 25th–75th percentiles), providing a robust representation of central tendency and spatial variability. In parallel, data completeness is quantified as the percentage of valid (non-missing) land pixels contributing to each dekad.\n\nThese statistics are computed separately for different baseline completeness thresholds (no threshold, ≥50%, ≥70%) to assess the sensitivity of the SSMA signal to data availability. The resulting time series are displayed as multi-panel figures, allowing direct comparison of SSMA behaviour and completeness across thresholds and regions.\n\nSubset region\nKeep only land pixels\nStack spatial dims\nCompleteness over land only\nLeft axis: median + IQR\nRight axis: completeness\nX ticks on dekad starts\nOnly one combined legend, placed in first panel\n\n*Figure 7. Mean SSMA over the Iberian Peninsula from spring to summer 2023, together with the corresponding data completeness (%). Panels (left to right) show results for different completeness thresholds: no threshold, ≥50%, and ≥70%.*\n\nOver the Iberian Peninsula, SSMA shows a clear seasonal evolution, with the most negative values occurring in mid-May, followed by a rapid transition towards positive anomalies in early June and a gradual stabilization towards near-neutral conditions during July and August. This pattern reflects the transition from peak drought conditions in late spring to partial recovery in early summer.\n\nThe temporal evolution of the median SSMA is highly consistent across all completeness thresholds, with only minor differences in magnitude. This indicates that the regional drought signal is robust to the choice of baseline filtering.\n\nData completeness remains high throughout the period. For the no-threshold and ≥50% cases, completeness exceeds 90% across most dekads, indicating strong spatial support. The ≥70% threshold results in slightly lower completeness (around 80%), but still maintains sufficient coverage to support the analysis.\n\nOverall, the strong agreement across thresholds, combined with high data availability, suggests that the SSMA estimates for the Iberian Peninsula are both stable and well-supported by the underlying data.\n\nThis evolution is consistent with the progression of drought conditions reported for the western Mediterranean in spring 2023, followed by partial recovery in early summer [[3]](https://doi.org/10.2760/998985), [[5]](https://doi.org/10.2760/575433), [[6]](https://doi.org/10.2760/928418), [[7]](https://joint-research-centre.ec.europa.eu/jrc-news-and-updates/severe-drought-western-mediterranean-faces-low-river-flows-and-crop-yields-earlier-ever-2023-06-13_en), [[8]](https://climate.copernicus.eu/precipitation-relative-humidity-and-soil-moisture-may-2023), and with studies highlighting the role of persistent soil moisture deficits in amplifying the record-breaking spring heatwave in the region [[9]](https://doi.org/10.1038/s41612-024-00569-6).\n\n*Figure 8. Mean SSMA over Great Britain and Ireland from spring to summer 2023, together with the corresponding data completeness (%). Panels (left to right) show results for different completeness thresholds: no threshold, ≥50%, and ≥70%.*\n\nOver Great Britain and Ireland, SSMA exhibits a clear minimum in June, indicating the peak of dry conditions, followed by a gradual recovery towards neutral and slight possitive anomalies conditions in July and August. The magnitude of the negative anomaly is slightly stronger when applying the ≥70% threshold, suggesting that stricter filtering removes pixels with limited baseline support, which can reduce noise and result in slightly stronger anomaly signals."} {"chunk_id": "satellite_satellite-soil-moisture_completeness_q02__dc4f24688b4b", "report_id": "satellite_satellite-soil-moisture_completeness_q02", "dataset_id": "satellite-soil-moisture", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q02", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite soil moisture for drought monitoring in Europe (2023 case study) > Analysis and results > 6. Regional SSMA analysis", "title": "Satellite soil moisture for drought monitoring in Europe (2023 case study)", "chunk_index": 10, "token_count": 1031, "text_raw": "possitive anomalies conditions in July and August. The magnitude of the negative anomaly is slightly stronger when applying the ≥70% threshold, suggesting that stricter filtering removes pixels with limited baseline support, which can reduce noise and result in slightly stronger anomaly signals.\n\nData completeness remains high (>80%) for both the no-threshold and ≥50% cases, indicating robust spatial support. In contrast, the ≥70% threshold leads to a substantial reduction in completeness (approximately 40–60%), reflecting limited baseline availability in parts of the region. Despite this reduction, the temporal evolution of SSMA remains consistent across thresholds, indicating that the main drought signal is stable.\n\n*Figure 9. Mean SSMA over the Southern Balkans and Italy from spring to summer 2023, together with the corresponding data completeness (%). Panels (left to right) show results for different completeness thresholds: no threshold, ≥50%, and ≥70%.*\n\nIn the Southern Balkans and Italy, SSMA is predominantly positive throughout the analysis period, with peaks in mid-May and mid-June. This behaviour contrasts with drought-affected regions and is consistent with spatial patterns observed in the maps. This positive anomaly is consistent with reported rainfall surpluses in May and June over Italy, as well as parts of Slovakia, Hungary, Croatia, and Slovenia, which led to waterlogging of soils and increased pest pressure on crops [[10]](https://doi.org/10.2760/82821).\n\nCompleteness shows a stronger dependence on the applied threshold. While the no-threshold case increases from ~80% to near-complete coverage over time, the ≥50% threshold introduces moderate variability around ~80%. The ≥70% threshold results in lower and more stable completeness (approximately 40–60%), with slightly higher values in early spring.\n\nDespite these differences in data availability, the SSMA signal remains largely unchanged across thresholds, indicating that the positive anomaly pattern is robust.\n\n*Figure 10. Mean SSMA over the Nordic and Baltic region from spring to summer 2023, together with the corresponding data completeness (%). Panels (left to right) show results for different completeness thresholds: no threshold, ≥50%, and ≥70%.*\n\nThe Nordic and Baltic region shows a pronounced seasonal evolution, with the lowest SSMA values occurring in June and a transition to slight positive anomalies by early August. The magnitude of negative anomalies is slightly enhanced under the ≥70% threshold.\n\nCompleteness exhibits a strong seasonal increase, rising from very low values (<20%) in early spring to high values (>80%) in summer. This pattern is consistent across thresholds, although the ≥70% threshold delays the point at which sufficient coverage is reached (crossing ~60% only in early July). This behaviour reflects known limitations in satellite soil moisture retrievals under snow and frozen soil conditions.\n\nDespite the strong variation in completeness, the temporal SSMA pattern remains consistent across thresholds once sufficient data coverage is achieved.\n\n*Figure 11. Mean SSMA over the Germany–Benelux–Poland area from spring to summer 2023, together with the corresponding data completeness (%). Panels (left to right) show results for different completeness thresholds: no threshold, ≥50%, and ≥70%.*\n\nIn the Germany–Benelux–Poland region, SSMA reaches its minimum in early June and transitions to positive anomalies by early August. Compared to other regions, this area shows greater sensitivity to threshold selection, particularly in early spring.\n\nCompleteness is near-total (≈100%) in the no-threshold case and remains high (>90%) for the ≥50% threshold. Under the ≥70% threshold, completeness is initially low (~40% in early March) but increases rapidly, stabilizing around ~80% for the remainder of the period.\n\nAcross all regions, the median SSMA and its temporal evolution are broadly consistent across the different completeness thresholds, despite substantial differences in data availability. This indicates that the main drought signal is relatively robust to the choice of threshold. However, stricter thresholds (e.g. ≥70%) significantly reduce spatial coverage in several regions, particularly where baseline data availability is limited.\n\nThe close agreement between the no-threshold and ≥50% cases indicates that applying a 50% completeness requirement does not materially change the baseline composition. This is likely due to the uneven temporal distribution of valid observations, with a higher density of data in more recent years following the introduction of additional sensors (e.g. AMSR-E and ASCAT). Consequently, the ≥50% threshold effectively selects pixels for which a sufficient number of valid years is already concentrated in the later part of the record, leading to an overrepresentation of recent conditions in the baseline and a reduced contribution from earlier, less complete years.", "text_with_prefix": "EQC Quality Assessment: \"Satellite soil moisture for drought monitoring in Europe (2023 case study)\"\nDataset: satellite-soil-moisture [CDS]\nAspect: completeness_q02 | Category: Satellite_ECVs\nSection: Satellite soil moisture for drought monitoring in Europe (2023 case study) > Analysis and results > 6. Regional SSMA analysis\n---\npossitive anomalies conditions in July and August. The magnitude of the negative anomaly is slightly stronger when applying the ≥70% threshold, suggesting that stricter filtering removes pixels with limited baseline support, which can reduce noise and result in slightly stronger anomaly signals.\n\nData completeness remains high (>80%) for both the no-threshold and ≥50% cases, indicating robust spatial support. In contrast, the ≥70% threshold leads to a substantial reduction in completeness (approximately 40–60%), reflecting limited baseline availability in parts of the region. Despite this reduction, the temporal evolution of SSMA remains consistent across thresholds, indicating that the main drought signal is stable.\n\n*Figure 9. Mean SSMA over the Southern Balkans and Italy from spring to summer 2023, together with the corresponding data completeness (%). Panels (left to right) show results for different completeness thresholds: no threshold, ≥50%, and ≥70%.*\n\nIn the Southern Balkans and Italy, SSMA is predominantly positive throughout the analysis period, with peaks in mid-May and mid-June. This behaviour contrasts with drought-affected regions and is consistent with spatial patterns observed in the maps. This positive anomaly is consistent with reported rainfall surpluses in May and June over Italy, as well as parts of Slovakia, Hungary, Croatia, and Slovenia, which led to waterlogging of soils and increased pest pressure on crops [[10]](https://doi.org/10.2760/82821).\n\nCompleteness shows a stronger dependence on the applied threshold. While the no-threshold case increases from ~80% to near-complete coverage over time, the ≥50% threshold introduces moderate variability around ~80%. The ≥70% threshold results in lower and more stable completeness (approximately 40–60%), with slightly higher values in early spring.\n\nDespite these differences in data availability, the SSMA signal remains largely unchanged across thresholds, indicating that the positive anomaly pattern is robust.\n\n*Figure 10. Mean SSMA over the Nordic and Baltic region from spring to summer 2023, together with the corresponding data completeness (%). Panels (left to right) show results for different completeness thresholds: no threshold, ≥50%, and ≥70%.*\n\nThe Nordic and Baltic region shows a pronounced seasonal evolution, with the lowest SSMA values occurring in June and a transition to slight positive anomalies by early August. The magnitude of negative anomalies is slightly enhanced under the ≥70% threshold.\n\nCompleteness exhibits a strong seasonal increase, rising from very low values (<20%) in early spring to high values (>80%) in summer. This pattern is consistent across thresholds, although the ≥70% threshold delays the point at which sufficient coverage is reached (crossing ~60% only in early July). This behaviour reflects known limitations in satellite soil moisture retrievals under snow and frozen soil conditions.\n\nDespite the strong variation in completeness, the temporal SSMA pattern remains consistent across thresholds once sufficient data coverage is achieved.\n\n*Figure 11. Mean SSMA over the Germany–Benelux–Poland area from spring to summer 2023, together with the corresponding data completeness (%). Panels (left to right) show results for different completeness thresholds: no threshold, ≥50%, and ≥70%.*\n\nIn the Germany–Benelux–Poland region, SSMA reaches its minimum in early June and transitions to positive anomalies by early August. Compared to other regions, this area shows greater sensitivity to threshold selection, particularly in early spring.\n\nCompleteness is near-total (≈100%) in the no-threshold case and remains high (>90%) for the ≥50% threshold. Under the ≥70% threshold, completeness is initially low (~40% in early March) but increases rapidly, stabilizing around ~80% for the remainder of the period.\n\nAcross all regions, the median SSMA and its temporal evolution are broadly consistent across the different completeness thresholds, despite substantial differences in data availability. This indicates that the main drought signal is relatively robust to the choice of threshold. However, stricter thresholds (e.g. ≥70%) significantly reduce spatial coverage in several regions, particularly where baseline data availability is limited.\n\nThe close agreement between the no-threshold and ≥50% cases indicates that applying a 50% completeness requirement does not materially change the baseline composition. This is likely due to the uneven temporal distribution of valid observations, with a higher density of data in more recent years following the introduction of additional sensors (e.g. AMSR-E and ASCAT). Consequently, the ≥50% threshold effectively selects pixels for which a sufficient number of valid years is already concentrated in the later part of the record, leading to an overrepresentation of recent conditions in the baseline and a reduced contribution from earlier, less complete years."} {"chunk_id": "satellite_satellite-soil-moisture_completeness_q02__bd665b2d824a", "report_id": "satellite_satellite-soil-moisture_completeness_q02", "dataset_id": "satellite-soil-moisture", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q02", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite soil moisture for drought monitoring in Europe (2023 case study) > ℹ️ If you want to know more", "title": "Satellite soil moisture for drought monitoring in Europe (2023 case study)", "chunk_index": 11, "token_count": 351, "text_raw": "* Markonis, Y., Kumar, R., Hanel, M., Rakovec, O., Máca, P., & AghaKouchak, A. (2021). The rise of compound warm-season droughts in Europe. Science Advances, 7(6), eabb9668. [](https://doi.org/10.1126/sciadv.abb9668)\n* Ministerio para la Transición Ecológica y el Reto Demográfico (2023). El 14,6% del territorio está en emergencia por escasez de agua y el 27,4%, en alerta. Nota de prensa. [](https://www.miteco.gob.es/es/prensa/ultimas-noticias/2023/09/el-14-6--del-territorio-esta-en-emergencia-por-escasez-de-agua-y.html)\n* Laguardia, G. & Niemeyer, S. (2008). On the comparison between the LISFLOOD modelled and the ERS/SCAT derived soil moisture estimates. Hydrol. Earth Syst. Sci., 12, 1339–1351. [](https://hess.copernicus.org/articles/12/1339/2008/)", "text_with_prefix": "EQC Quality Assessment: \"Satellite soil moisture for drought monitoring in Europe (2023 case study)\"\nDataset: satellite-soil-moisture [CDS]\nAspect: completeness_q02 | Category: Satellite_ECVs\nSection: Satellite soil moisture for drought monitoring in Europe (2023 case study) > ℹ️ If you want to know more\n---\n* Markonis, Y., Kumar, R., Hanel, M., Rakovec, O., Máca, P., & AghaKouchak, A. (2021). The rise of compound warm-season droughts in Europe. Science Advances, 7(6), eabb9668. [](https://doi.org/10.1126/sciadv.abb9668)\n* Ministerio para la Transición Ecológica y el Reto Demográfico (2023). El 14,6% del territorio está en emergencia por escasez de agua y el 27,4%, en alerta. Nota de prensa. [](https://www.miteco.gob.es/es/prensa/ultimas-noticias/2023/09/el-14-6--del-territorio-esta-en-emergencia-por-escasez-de-agua-y.html)\n* Laguardia, G. & Niemeyer, S. (2008). On the comparison between the LISFLOOD modelled and the ERS/SCAT derived soil moisture estimates. Hydrol. Earth Syst. Sci., 12, 1339–1351. [](https://hess.copernicus.org/articles/12/1339/2008/)"} {"chunk_id": "satellite_satellite-soil-moisture_completeness_q02__f44ef86b8e22", "report_id": "satellite_satellite-soil-moisture_completeness_q02", "dataset_id": "satellite-soil-moisture", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q02", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite soil moisture for drought monitoring in Europe (2023 case study) > ℹ️ If you want to know more > Key resources", "title": "Satellite soil moisture for drought monitoring in Europe (2023 case study)", "chunk_index": 12, "token_count": 321, "text_raw": "Code libraries used:\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)\n\nDataset documentation:\n\n* [SM v202212: Product User Guide and Specification (PUGS)](https://confluence.ecmwf.int/pages/viewpage.action?pageId=355349314)\n\n* [SM v202312: Algorithm Theoretical Basis Document (ATBD)](https://confluence.ecmwf.int/pages/viewpage.action?pageId=453676750)\n\n* [SM v202312: Product Quality Assurance Document (PQAD)](https://confluence.ecmwf.int/pages/viewpage.action?pageId=445290971)\n\n* [SM v202312: Product Quality Assessment Report (PQAR)](https://confluence.ecmwf.int/pages/viewpage.action?pageId=445290728#SMv202312:ProductQualityAssessmentReport(PQAR)-Spatialandtemporalcompletenesss2.1)", "text_with_prefix": "EQC Quality Assessment: \"Satellite soil moisture for drought monitoring in Europe (2023 case study)\"\nDataset: satellite-soil-moisture [CDS]\nAspect: completeness_q02 | Category: Satellite_ECVs\nSection: Satellite soil moisture for drought monitoring in Europe (2023 case study) > ℹ️ If you want to know more > Key resources\n---\nCode libraries used:\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)\n\nDataset documentation:\n\n* [SM v202212: Product User Guide and Specification (PUGS)](https://confluence.ecmwf.int/pages/viewpage.action?pageId=355349314)\n\n* [SM v202312: Algorithm Theoretical Basis Document (ATBD)](https://confluence.ecmwf.int/pages/viewpage.action?pageId=453676750)\n\n* [SM v202312: Product Quality Assurance Document (PQAD)](https://confluence.ecmwf.int/pages/viewpage.action?pageId=445290971)\n\n* [SM v202312: Product Quality Assessment Report (PQAR)](https://confluence.ecmwf.int/pages/viewpage.action?pageId=445290728#SMv202312:ProductQualityAssessmentReport(PQAR)-Spatialandtemporalcompletenesss2.1)"} {"chunk_id": "satellite_satellite-soil-moisture_completeness_q02__e9848db38a73", "report_id": "satellite_satellite-soil-moisture_completeness_q02", "dataset_id": "satellite-soil-moisture", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q02", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite soil moisture for drought monitoring in Europe (2023 case study) > ℹ️ If you want to know more > References", "title": "Satellite soil moisture for drought monitoring in Europe (2023 case study)", "chunk_index": 13, "token_count": 991, "text_raw": "[[1]](https://ec.europa.eu/eurostat/web/gisco/geodata/administrative-units) Eurostat (Accessed on 2016). GISCO. Administrative units.\n\n[[2]](https://community.wmo.int/site/knowledge-hub/programmes-and-initiatives/climate-services/wmo-climatological-normals)) World Meteorological Organization. (Accessed on 2016). WMO Climatological Normals.\n\n[[3]](https://doi.org/10.2760/998985) Toreti, A., Bavera, D., Acosta Navarro, J., Arias Muñoz, C., Avanzi, F., Barbosa, P., de Jager, A., Di Ciollo, C., Ferraris, L., Fioravanti, G., Gabellani, S., Grimaldi, S., Hrast Essenfelder, A., Isabellon, M., Jonas, T., Maetens, W., Magni, D., Masante, D., Mazzeschi, M., McCormick, N., Meroni, M., Rossi, L., Salamon, P., Spinoni, J. (2023). Drought in Europe March 2023. Publications Office of the European Union, Luxembourg, doi:10.2760/998985, JRC133025.\n\n[[4]](https://drought.emergency.copernicus.eu/data/factsheets/factsheet_soilmoisture.pdf) European Commission (2019). EDO INDICATOR FACTSHEET Soil Moisture Anomaly (SMA). European Drought Observatory.\n\n[[5]](https://doi.org/10.2760/575433) Toreti, A., Bavera, D., Acosta Navarro, J., Arias Muñoz, C., Avanzi, F., Barbosa, P., de Jager, A., Di Ciollo, C., Ferraris, L., Fioravanti, G., Gabellani, S., Grimaldi, S., Hrast Essenfelder, A., Isabellon, M., Jonas, T., Maetens, W., Magni, D., Masante, D., Mazzeschi, M., McCormick, N., Rossi, L., Salamon, P. (2023). Drought in Europe June 2023, Publications Office of the European Union, Luxembourg, doi:10.2760/575433, JRC134492.\n\n[[6]](https://doi.org/10.2760/928418) Toreti, A., Bavera, D., Acosta Navarro, J., Arias Muñoz, C., Barbosa, P., de Jager, A., Di Ciollo, C., Fioravanti, G., , Grimaldi, S., Hrast Essenfelder, A., Maetens, W., Magni, D., Masante, D., Mazzeschi, M., McCormick, N., Salamon, P. (2023). Drought in Europe - August 2023, Publications Office of the European Union, Luxembourg, doi:10.2760/928418, JRC135032.\n\n[[7]](https://joint-research-centre.ec.europa.eu/jrc-news-and-updates/severe-drought-western-mediterranean-faces-low-river-flows-and-crop-yields-earlier-ever-2023-06-13_en) European Commission (2023). Severe drought: western Mediterranean faces low river flows and crop yields earlier than ever. European Drought Observatory. The Joint Research Centre: EU Science Hub.\n\n[[8]](https://climate.copernicus.eu/precipitation-relative-humidity-and-soil-moisture-may-2023) European Commission (2023). Precipitation, relative humidity and soil moisture for May 2023.\n\n[[9]](https://doi.org/10.1038/s41612-024-00569-6) Lemus-Canovas, M., Insua-Costa, D., Trigo, R. & Miralles, D. (2024). Record-shattering 2023 Spring heatwave in western Mediterranean amplified by long-term drought. npj Clim Atmos Sci 7, 25.", "text_with_prefix": "EQC Quality Assessment: \"Satellite soil moisture for drought monitoring in Europe (2023 case study)\"\nDataset: satellite-soil-moisture [CDS]\nAspect: completeness_q02 | Category: Satellite_ECVs\nSection: Satellite soil moisture for drought monitoring in Europe (2023 case study) > ℹ️ If you want to know more > References\n---\n[[1]](https://ec.europa.eu/eurostat/web/gisco/geodata/administrative-units) Eurostat (Accessed on 2016). GISCO. Administrative units.\n\n[[2]](https://community.wmo.int/site/knowledge-hub/programmes-and-initiatives/climate-services/wmo-climatological-normals)) World Meteorological Organization. (Accessed on 2016). WMO Climatological Normals.\n\n[[3]](https://doi.org/10.2760/998985) Toreti, A., Bavera, D., Acosta Navarro, J., Arias Muñoz, C., Avanzi, F., Barbosa, P., de Jager, A., Di Ciollo, C., Ferraris, L., Fioravanti, G., Gabellani, S., Grimaldi, S., Hrast Essenfelder, A., Isabellon, M., Jonas, T., Maetens, W., Magni, D., Masante, D., Mazzeschi, M., McCormick, N., Meroni, M., Rossi, L., Salamon, P., Spinoni, J. (2023). Drought in Europe March 2023. Publications Office of the European Union, Luxembourg, doi:10.2760/998985, JRC133025.\n\n[[4]](https://drought.emergency.copernicus.eu/data/factsheets/factsheet_soilmoisture.pdf) European Commission (2019). EDO INDICATOR FACTSHEET Soil Moisture Anomaly (SMA). European Drought Observatory.\n\n[[5]](https://doi.org/10.2760/575433) Toreti, A., Bavera, D., Acosta Navarro, J., Arias Muñoz, C., Avanzi, F., Barbosa, P., de Jager, A., Di Ciollo, C., Ferraris, L., Fioravanti, G., Gabellani, S., Grimaldi, S., Hrast Essenfelder, A., Isabellon, M., Jonas, T., Maetens, W., Magni, D., Masante, D., Mazzeschi, M., McCormick, N., Rossi, L., Salamon, P. (2023). Drought in Europe June 2023, Publications Office of the European Union, Luxembourg, doi:10.2760/575433, JRC134492.\n\n[[6]](https://doi.org/10.2760/928418) Toreti, A., Bavera, D., Acosta Navarro, J., Arias Muñoz, C., Barbosa, P., de Jager, A., Di Ciollo, C., Fioravanti, G., , Grimaldi, S., Hrast Essenfelder, A., Maetens, W., Magni, D., Masante, D., Mazzeschi, M., McCormick, N., Salamon, P. (2023). Drought in Europe - August 2023, Publications Office of the European Union, Luxembourg, doi:10.2760/928418, JRC135032.\n\n[[7]](https://joint-research-centre.ec.europa.eu/jrc-news-and-updates/severe-drought-western-mediterranean-faces-low-river-flows-and-crop-yields-earlier-ever-2023-06-13_en) European Commission (2023). Severe drought: western Mediterranean faces low river flows and crop yields earlier than ever. European Drought Observatory. The Joint Research Centre: EU Science Hub.\n\n[[8]](https://climate.copernicus.eu/precipitation-relative-humidity-and-soil-moisture-may-2023) European Commission (2023). Precipitation, relative humidity and soil moisture for May 2023.\n\n[[9]](https://doi.org/10.1038/s41612-024-00569-6) Lemus-Canovas, M., Insua-Costa, D., Trigo, R. & Miralles, D. (2024). Record-shattering 2023 Spring heatwave in western Mediterranean amplified by long-term drought. npj Clim Atmos Sci 7, 25."} {"chunk_id": "satellite_satellite-soil-moisture_completeness_q02__37ec74469ceb", "report_id": "satellite_satellite-soil-moisture_completeness_q02", "dataset_id": "satellite-soil-moisture", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q02", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite soil moisture for drought monitoring in Europe (2023 case study) > ℹ️ If you want to know more > References", "title": "Satellite soil moisture for drought monitoring in Europe (2023 case study)", "chunk_index": 14, "token_count": 375, "text_raw": "., Trigo, R. & Miralles, D. (2024). Record-shattering 2023 Spring heatwave in western Mediterranean amplified by long-term drought. npj Clim Atmos Sci 7, 25.\n\n[[10]](https://doi.org/10.2760/82821) Baruth, B., Ben Aoun, W., Biavetti, I., Bratu, M., Bussay, A., Cerrani, I., Chemin, Y., Claverie, M., De Palma, P., Fumagalli, D., Manfron, G., Morel, J., Nisini, L., Panarello, L., Rossi, M., Tarnavsky, E., Van Den Berg, M., Zajac, Z. and Zucchini, A., JRC MARS Bulletin - Crop monitoring in Europe - June 2023 - Vol. 31 No 6, Van Den Berg, M., Niemeyer, S. and Baruth, B. editor(s), Publications Office of the European Union, Luxembourg, 2023, doi:10.2760/82821, JRC133186.\n\n[[11]](https://www.ncei.noaa.gov/access/monitoring/monthly-report/global-drought/202306) NOAA National Centers for Environmental Information (2023). Monthly Climate Reports, Global Drought Narrative, June 2023.", "text_with_prefix": "EQC Quality Assessment: \"Satellite soil moisture for drought monitoring in Europe (2023 case study)\"\nDataset: satellite-soil-moisture [CDS]\nAspect: completeness_q02 | Category: Satellite_ECVs\nSection: Satellite soil moisture for drought monitoring in Europe (2023 case study) > ℹ️ If you want to know more > References\n---\n., Trigo, R. & Miralles, D. (2024). Record-shattering 2023 Spring heatwave in western Mediterranean amplified by long-term drought. npj Clim Atmos Sci 7, 25.\n\n[[10]](https://doi.org/10.2760/82821) Baruth, B., Ben Aoun, W., Biavetti, I., Bratu, M., Bussay, A., Cerrani, I., Chemin, Y., Claverie, M., De Palma, P., Fumagalli, D., Manfron, G., Morel, J., Nisini, L., Panarello, L., Rossi, M., Tarnavsky, E., Van Den Berg, M., Zajac, Z. and Zucchini, A., JRC MARS Bulletin - Crop monitoring in Europe - June 2023 - Vol. 31 No 6, Van Den Berg, M., Niemeyer, S. and Baruth, B. editor(s), Publications Office of the European Union, Luxembourg, 2023, doi:10.2760/82821, JRC133186.\n\n[[11]](https://www.ncei.noaa.gov/access/monitoring/monthly-report/global-drought/202306) NOAA National Centers for Environmental Information (2023). Monthly Climate Reports, Global Drought Narrative, June 2023."} {"chunk_id": "satellite_satellite-ocean-colour_completeness_q01__851da463616e", "report_id": "satellite_satellite-ocean-colour_completeness_q01", "dataset_id": "satellite-ocean-colour", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q01", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Completeness of ocean colour observations for biogeochemical models > Quality assessment question", "title": "Completeness of ocean colour observations for biogeochemical models", "chunk_index": 0, "token_count": 744, "text_raw": "**Is chlorophyll-a data sufficiently complete in time and space for integration into biogeochemical models?**\n\nThe Ocean Colour dataset version 6.0, as produced for the Copernicus Climate Change Service (C3S), includes the two variables: mass concentration of chlorophyll-a and remote sensing reflectance (Rrs) for six wavelengths from October 1997 to present [[1]](https://dast.copernicus-climate.eu/documents/satellite-ocean-colour/v6.0/WP2-FDDP-2022-04_C3S2-Lot3_PUGS-of-v6.0-OceanColour-product_v1.1_FINAL.pdf). \nChlorophyll-a concentration data, which are derived through specific algorithms using remote sensing reflectance (Rrs) [[2]](https://dast.copernicus-climate.eu/documents/satellite-ocean-colour/v6.0/WP2-FDDP-2022-04_C3S2-Lot3_ATBD-of-v6.0-OceanColour-product_v1.1_FINAL.pdf), are indispensable in biogeochemical modelling as they provide inputs for models initialisation and references to calibrate biogeochemical parameters and validate models results (e.g., [[3]](https://doi.org/10.1016/j.jmarsys.2013.02.007) [[4]](https://doi.org/10.3390/rs10101666) [[5]](https://doi.org/10.1038/s41467-019-08457-x)). Moreover, chlorophyll-a data can be assimilated into models, enabling continuous updates with observed conditions and significantly improve their accuracy (e.g., [[6]](https://doi.org/10.1029/2018JC014329)).\nThe dataset merges measurements carried out by six satellite sensors [[1]](https://dast.copernicus-climate.eu/documents/satellite-ocean-colour/v6.0/WP2-FDDP-2022-04_C3S2-Lot3_PUGS-of-v6.0-OceanColour-product_v1.1_FINAL.pdf): SeaWiFS, MERIS, MODIS-Aqua, VIIRS, OLCI-3A and OLCI-3B. Each sensor has specific design characteristics and viewing geometry, and no sensor was operational over the whole temporal coverage of the datasets [[1]](https://dast.copernicus-climate.eu/documents/satellite-ocean-colour/v6.0/WP2-FDDP-2022-04_C3S2-Lot3_PUGS-of-v6.0-OceanColour-product_v1.1_FINAL.pdf) [[2]](https://dast.copernicus-climate.eu/documents/satellite-ocean-colour/v6.0/WP2-FDDP-2022-04_C3S2-Lot3_ATBD-of-v6.0-OceanColour-product_v1.1_FINAL.pdf). Currently, only OLCI-3A and OLCI-3B are operational. \nHere, the goal is to assess the completeness of the chlorophyll-a dataset at both spatial and temporal scales in the Southern Ocean.", "text_with_prefix": "EQC Quality Assessment: \"Completeness of ocean colour observations for biogeochemical models\"\nDataset: satellite-ocean-colour [CDS]\nAspect: completeness_q01 | Category: Satellite_ECVs\nSection: Completeness of ocean colour observations for biogeochemical models > Quality assessment question\n---\n**Is chlorophyll-a data sufficiently complete in time and space for integration into biogeochemical models?**\n\nThe Ocean Colour dataset version 6.0, as produced for the Copernicus Climate Change Service (C3S), includes the two variables: mass concentration of chlorophyll-a and remote sensing reflectance (Rrs) for six wavelengths from October 1997 to present [[1]](https://dast.copernicus-climate.eu/documents/satellite-ocean-colour/v6.0/WP2-FDDP-2022-04_C3S2-Lot3_PUGS-of-v6.0-OceanColour-product_v1.1_FINAL.pdf). \nChlorophyll-a concentration data, which are derived through specific algorithms using remote sensing reflectance (Rrs) [[2]](https://dast.copernicus-climate.eu/documents/satellite-ocean-colour/v6.0/WP2-FDDP-2022-04_C3S2-Lot3_ATBD-of-v6.0-OceanColour-product_v1.1_FINAL.pdf), are indispensable in biogeochemical modelling as they provide inputs for models initialisation and references to calibrate biogeochemical parameters and validate models results (e.g., [[3]](https://doi.org/10.1016/j.jmarsys.2013.02.007) [[4]](https://doi.org/10.3390/rs10101666) [[5]](https://doi.org/10.1038/s41467-019-08457-x)). Moreover, chlorophyll-a data can be assimilated into models, enabling continuous updates with observed conditions and significantly improve their accuracy (e.g., [[6]](https://doi.org/10.1029/2018JC014329)).\nThe dataset merges measurements carried out by six satellite sensors [[1]](https://dast.copernicus-climate.eu/documents/satellite-ocean-colour/v6.0/WP2-FDDP-2022-04_C3S2-Lot3_PUGS-of-v6.0-OceanColour-product_v1.1_FINAL.pdf): SeaWiFS, MERIS, MODIS-Aqua, VIIRS, OLCI-3A and OLCI-3B. Each sensor has specific design characteristics and viewing geometry, and no sensor was operational over the whole temporal coverage of the datasets [[1]](https://dast.copernicus-climate.eu/documents/satellite-ocean-colour/v6.0/WP2-FDDP-2022-04_C3S2-Lot3_PUGS-of-v6.0-OceanColour-product_v1.1_FINAL.pdf) [[2]](https://dast.copernicus-climate.eu/documents/satellite-ocean-colour/v6.0/WP2-FDDP-2022-04_C3S2-Lot3_ATBD-of-v6.0-OceanColour-product_v1.1_FINAL.pdf). Currently, only OLCI-3A and OLCI-3B are operational. \nHere, the goal is to assess the completeness of the chlorophyll-a dataset at both spatial and temporal scales in the Southern Ocean."} {"chunk_id": "satellite_satellite-ocean-colour_completeness_q01__4bd0f66eef51", "report_id": "satellite_satellite-ocean-colour_completeness_q01", "dataset_id": "satellite-ocean-colour", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q01", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Completeness of ocean colour observations for biogeochemical models > Quality assessment statement", "title": "Completeness of ocean colour observations for biogeochemical models", "chunk_index": 1, "token_count": 1050, "text_raw": "These are the key outcomes of this assessment\n\nThe Ocean Colour dataset version 6.0:\n* provides a limited representation of the spatial distribution of chlorophyll-a concentrations in the oceanic region below latitude 45.5°S, based on less than 8% of valid pixels.\n* offers an acceptable representation of the spatial distribution of chlorophyll-a concentrations off the southeast coast of South America, based on more than 30% of valid pixels.\n* requires filtering based on the availability of valid pixels when analysing chlorophyll-a temporal distribution, to reduce potential bias due to limited sampling, especially during the austral winter (May-August).\n* allows for the calculation of monthly chlorophyll-a trends in the Southern Ocean, although these trends should be interpreted with caution as based on less than 15% of valid pixels.\n* can be used for calibrating, assimilating and validating biogeochemical models in the Southern Ocean from September to April, when valid observations are more abundant.\n\n## 📋 Methodology\n\nThis notebook provides an assessment of the ability of the Ocean Colour dataset version 6.0 to represent the spatial distribution and temporal variability of chlorophyll-a concentrations in the Indian, Pacific, and Atlantic Ocean sectors of the Permanent Open Ocean Zone in the Southern Ocean by analysing the number of valid pixels available.\n\nThe analysis and results are detailed in the sections below:\n\n[](section-1)\n * Import required packages\n * Define parameters to be analysed (time period, variables, regions) and data request\n\n[](section-2)\n * Define transform functions\n * Retrieve data for the selected variables, time period and regions\n\n[](section-3)\n * Define mapping function\n * Compute trends\n\n[](section-4)\n * Display results \n * Discussion\n\n## 📈 Analysis and results\n(section-1)=\n### 1. Choose the data to use and set up the code\n#### Import required packages\nBesides the standard libraries used to manage and analyse multidimensional arrays, the [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control) `c3s_eqc_automatic_quality_control`is imported to download data and calculate statistics.\n\n#### Define parameters to be analysed (time period, variables, regions) and data request\nThe analysis performed in this notebook examines the spatial and temporal distribution of chlorophyll-a data, and corresponding valid pixels, over a 21-year period (January 2003 - December 2023), which accounts for previous EQC information (https://cds.climate.copernicus.eu/datasets/satellite-ocean-colour?tab=quality_assurance_tab) and includes only complete years. The analysis is performed over three sectors within the oceanic region in the Southern Ocean located between latitudes 47.5°S and 63.5°S (i.e., the Permanent Open Ocean Zone, POOZ) [[7]](https://doi.org/10.1016/0967-0637%2895%2900021-W). The three POOZ sectors analysed are: the Indian Ocean sector (IO_POOZ), the Pacific Ocean sector (PO_POOZ) and the Atlantic Ocean sector (AO_POOZ), which extend between longitudes 20°E and 150°E, 150°E and 70°W, and 70°W and 20°E, respectively [[8]](https://doi.org/10.1029/2019GL083163).\n\nDefine data request\n\n(section-2)=\n### 2. Transform functions and data retrieval\n#### Define transform functions\nThe `monthly_weighted_log_mean` function is defined to calculate temporal averages of chlorophyll-a concentration (mg m-3) and the percentage of corresponding valid pixels for each POOZ sector by month and year, while the `weighted_log_map` function is defined to compute spatial averages across the entire POOZ.\nBoth functions account for the varying surface area at different latitudes. Averages for chlorophyll-a concentration are computed using log-transformed daily data, and then back-transformed [[9]](https://doi.org/10.1029/95JC00458). Daily chlorophyll-a concentrations outside the range 0.01-100 mg m-3 are excluded from the analysis [[10]](https://doi.org/10.3390/s19194285).\n\n#### Data retrieval\nThe `monthly_weighted_log_mean` and `weighted_log_map` functions are applied to the chlorophyll-a dataset in the selected regions, and temporal and spatial averages are downloaded as two separated arrays.", "text_with_prefix": "EQC Quality Assessment: \"Completeness of ocean colour observations for biogeochemical models\"\nDataset: satellite-ocean-colour [CDS]\nAspect: completeness_q01 | Category: Satellite_ECVs\nSection: Completeness of ocean colour observations for biogeochemical models > Quality assessment statement\n---\nThese are the key outcomes of this assessment\n\nThe Ocean Colour dataset version 6.0:\n* provides a limited representation of the spatial distribution of chlorophyll-a concentrations in the oceanic region below latitude 45.5°S, based on less than 8% of valid pixels.\n* offers an acceptable representation of the spatial distribution of chlorophyll-a concentrations off the southeast coast of South America, based on more than 30% of valid pixels.\n* requires filtering based on the availability of valid pixels when analysing chlorophyll-a temporal distribution, to reduce potential bias due to limited sampling, especially during the austral winter (May-August).\n* allows for the calculation of monthly chlorophyll-a trends in the Southern Ocean, although these trends should be interpreted with caution as based on less than 15% of valid pixels.\n* can be used for calibrating, assimilating and validating biogeochemical models in the Southern Ocean from September to April, when valid observations are more abundant.\n\n## 📋 Methodology\n\nThis notebook provides an assessment of the ability of the Ocean Colour dataset version 6.0 to represent the spatial distribution and temporal variability of chlorophyll-a concentrations in the Indian, Pacific, and Atlantic Ocean sectors of the Permanent Open Ocean Zone in the Southern Ocean by analysing the number of valid pixels available.\n\nThe analysis and results are detailed in the sections below:\n\n[](section-1)\n * Import required packages\n * Define parameters to be analysed (time period, variables, regions) and data request\n\n[](section-2)\n * Define transform functions\n * Retrieve data for the selected variables, time period and regions\n\n[](section-3)\n * Define mapping function\n * Compute trends\n\n[](section-4)\n * Display results \n * Discussion\n\n## 📈 Analysis and results\n(section-1)=\n### 1. Choose the data to use and set up the code\n#### Import required packages\nBesides the standard libraries used to manage and analyse multidimensional arrays, the [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control) `c3s_eqc_automatic_quality_control`is imported to download data and calculate statistics.\n\n#### Define parameters to be analysed (time period, variables, regions) and data request\nThe analysis performed in this notebook examines the spatial and temporal distribution of chlorophyll-a data, and corresponding valid pixels, over a 21-year period (January 2003 - December 2023), which accounts for previous EQC information (https://cds.climate.copernicus.eu/datasets/satellite-ocean-colour?tab=quality_assurance_tab) and includes only complete years. The analysis is performed over three sectors within the oceanic region in the Southern Ocean located between latitudes 47.5°S and 63.5°S (i.e., the Permanent Open Ocean Zone, POOZ) [[7]](https://doi.org/10.1016/0967-0637%2895%2900021-W). The three POOZ sectors analysed are: the Indian Ocean sector (IO_POOZ), the Pacific Ocean sector (PO_POOZ) and the Atlantic Ocean sector (AO_POOZ), which extend between longitudes 20°E and 150°E, 150°E and 70°W, and 70°W and 20°E, respectively [[8]](https://doi.org/10.1029/2019GL083163).\n\nDefine data request\n\n(section-2)=\n### 2. Transform functions and data retrieval\n#### Define transform functions\nThe `monthly_weighted_log_mean` function is defined to calculate temporal averages of chlorophyll-a concentration (mg m-3) and the percentage of corresponding valid pixels for each POOZ sector by month and year, while the `weighted_log_map` function is defined to compute spatial averages across the entire POOZ.\nBoth functions account for the varying surface area at different latitudes. Averages for chlorophyll-a concentration are computed using log-transformed daily data, and then back-transformed [[9]](https://doi.org/10.1029/95JC00458). Daily chlorophyll-a concentrations outside the range 0.01-100 mg m-3 are excluded from the analysis [[10]](https://doi.org/10.3390/s19194285).\n\n#### Data retrieval\nThe `monthly_weighted_log_mean` and `weighted_log_map` functions are applied to the chlorophyll-a dataset in the selected regions, and temporal and spatial averages are downloaded as two separated arrays."} {"chunk_id": "satellite_satellite-ocean-colour_completeness_q01__804e50c58538", "report_id": "satellite_satellite-ocean-colour_completeness_q01", "dataset_id": "satellite-ocean-colour", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q01", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Completeness of ocean colour observations for biogeochemical models > Quality assessment statement", "title": "Completeness of ocean colour observations for biogeochemical models", "chunk_index": 2, "token_count": 872, "text_raw": "19194285).\n\n#### Data retrieval\nThe `monthly_weighted_log_mean` and `weighted_log_map` functions are applied to the chlorophyll-a dataset in the selected regions, and temporal and spatial averages are downloaded as two separated arrays.\n\n(section-3)=\n### 3. Mapping function and trends calculation\n#### Define mapping function\nThe `plot_map` function is defined to visualise the maps of average chlorophyll-a concentrations and corresponding percentages of valid pixels across the POOZ.\n\n#### Compute trends \nThe `calculate_mk_trends` function is defined to calculate chlorophyll-a trends and their statistical significance for each individual month over the 21-year period analysed here in the three POOZ sectors.\n\n(section-4)=\n### 4. Display and discuss results\n#### Display results\nMaps showing the spatial distribution of valid pixel percentages and derived chlorophyll-a concentration in the POOZ are provided. Two sets of heatmaps are generated: one for the monthly percentage of valid chlorophyll-a pixels by POOZ sector, and another for the monthly chlorophyll-a concentrations. Here, chlorophyll-a data are displayed only when data coverage meets or exceeds the 21-year mean percentage of valid pixels across the POOZ (≈8%; see Discussion), and are referred to as filtered chlorophyll-a. Monthly climatologies of valid pixels and filtered chlorophyll-a concentrations are also shown. Trends in filtered chlorophyll-a and their statistical significance, calculated for each individual month over the 21-year period analysed, are summarised in a table, where statistically significant trends (p>0.05) are highlighted in bold and na indicates not enough data for trends calculations.\n\nDisplay maps\nFiltering chlorophyll-a based on valid pixel coverage\nDisplay heatmaps\nDisplay climatologies\nDisplay mk results\n\n#### Discussion\n\n**Chlorophyll-a spatial distribution:**\n\nThe mean availability of valid chlorophyll-a pixels over the 21-year period analysed is approximately 8% over the entire POOZ, indicating limited dataset coverage that may not fully capture the spatial variability of chlorophyll-a distribution in this region. A small difference in the number of valid pixels is observed across the three POOZ sectors: the PO_POOZ exhibits the greatest availability (9.5%), followed by the AO_POOZ (8.3%) and the IO_POOZ (7.3%). A significant number of valid pixels (above 30%) are detected off the southeastern coasts of South America. Therefore, the representation of the spatial distribution of chlorophyll-a in this area can be assumed to be more reliable than elsewhere in the POOZ. \nDifferent average chlorophyll-a concentrations are observed in the three POOZ sectors: the AO_POOZ sector is the less oligotrophic with an average value of about 0.2 mg m-3, whereas the IO_POOZ and PO_POOZ are characterised by lower averages (0.12 and 0.14 mg m-3, respectively), confirming what reported in previous studies (e.g., [[8]](https://doi.org/10.1029/2019GL083163)). While the subregional variability in terms of valid pixels likely depends on differences in sea ice extent and cloud coverage across the three POOZ sectors, differences in their average chlorophyll-a concentration are due to several key factors, such as nutrient supply through Patagonian dust deposition and/or upwelling (e.g., [[11]](https://doi.org/10.1029/2006jc004072) [[12]](https://doi.org/10.5670/oceanog.2018.408) [[13]](https://doi.org/10.3389/fmars.2024.1363088)).\n\n**Chlorophyll-a temporal distribution:**", "text_with_prefix": "EQC Quality Assessment: \"Completeness of ocean colour observations for biogeochemical models\"\nDataset: satellite-ocean-colour [CDS]\nAspect: completeness_q01 | Category: Satellite_ECVs\nSection: Completeness of ocean colour observations for biogeochemical models > Quality assessment statement\n---\n19194285).\n\n#### Data retrieval\nThe `monthly_weighted_log_mean` and `weighted_log_map` functions are applied to the chlorophyll-a dataset in the selected regions, and temporal and spatial averages are downloaded as two separated arrays.\n\n(section-3)=\n### 3. Mapping function and trends calculation\n#### Define mapping function\nThe `plot_map` function is defined to visualise the maps of average chlorophyll-a concentrations and corresponding percentages of valid pixels across the POOZ.\n\n#### Compute trends \nThe `calculate_mk_trends` function is defined to calculate chlorophyll-a trends and their statistical significance for each individual month over the 21-year period analysed here in the three POOZ sectors.\n\n(section-4)=\n### 4. Display and discuss results\n#### Display results\nMaps showing the spatial distribution of valid pixel percentages and derived chlorophyll-a concentration in the POOZ are provided. Two sets of heatmaps are generated: one for the monthly percentage of valid chlorophyll-a pixels by POOZ sector, and another for the monthly chlorophyll-a concentrations. Here, chlorophyll-a data are displayed only when data coverage meets or exceeds the 21-year mean percentage of valid pixels across the POOZ (≈8%; see Discussion), and are referred to as filtered chlorophyll-a. Monthly climatologies of valid pixels and filtered chlorophyll-a concentrations are also shown. Trends in filtered chlorophyll-a and their statistical significance, calculated for each individual month over the 21-year period analysed, are summarised in a table, where statistically significant trends (p>0.05) are highlighted in bold and na indicates not enough data for trends calculations.\n\nDisplay maps\nFiltering chlorophyll-a based on valid pixel coverage\nDisplay heatmaps\nDisplay climatologies\nDisplay mk results\n\n#### Discussion\n\n**Chlorophyll-a spatial distribution:**\n\nThe mean availability of valid chlorophyll-a pixels over the 21-year period analysed is approximately 8% over the entire POOZ, indicating limited dataset coverage that may not fully capture the spatial variability of chlorophyll-a distribution in this region. A small difference in the number of valid pixels is observed across the three POOZ sectors: the PO_POOZ exhibits the greatest availability (9.5%), followed by the AO_POOZ (8.3%) and the IO_POOZ (7.3%). A significant number of valid pixels (above 30%) are detected off the southeastern coasts of South America. Therefore, the representation of the spatial distribution of chlorophyll-a in this area can be assumed to be more reliable than elsewhere in the POOZ. \nDifferent average chlorophyll-a concentrations are observed in the three POOZ sectors: the AO_POOZ sector is the less oligotrophic with an average value of about 0.2 mg m-3, whereas the IO_POOZ and PO_POOZ are characterised by lower averages (0.12 and 0.14 mg m-3, respectively), confirming what reported in previous studies (e.g., [[8]](https://doi.org/10.1029/2019GL083163)). While the subregional variability in terms of valid pixels likely depends on differences in sea ice extent and cloud coverage across the three POOZ sectors, differences in their average chlorophyll-a concentration are due to several key factors, such as nutrient supply through Patagonian dust deposition and/or upwelling (e.g., [[11]](https://doi.org/10.1029/2006jc004072) [[12]](https://doi.org/10.5670/oceanog.2018.408) [[13]](https://doi.org/10.3389/fmars.2024.1363088)).\n\n**Chlorophyll-a temporal distribution:**"} {"chunk_id": "satellite_satellite-ocean-colour_completeness_q01__07a8d4cb7fe3", "report_id": "satellite_satellite-ocean-colour_completeness_q01", "dataset_id": "satellite-ocean-colour", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q01", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Completeness of ocean colour observations for biogeochemical models > Quality assessment statement", "title": "Completeness of ocean colour observations for biogeochemical models", "chunk_index": 3, "token_count": 1023, "text_raw": "8.408) [[13]](https://doi.org/10.3389/fmars.2024.1363088)).\n\n**Chlorophyll-a temporal distribution:**\n\nThe monthly climatological distribution of valid chlorophyll-a pixels is in line with expectations based on their dependence on the solar zenith angle: the highest percentage (up to about 15%) is observed between November and February, when the solar zenith angle is low, then it decreases to reach its minimum (almost 0%) during the polar night months (i.e. May-July) and gradually increases again towards the end of the year. It is interesting to note that, as shown in the valid pixel heatmap, the valid percentages increased between 2017 and 2019, reaching a maximum in 2018 (up to about 24%), when the highest number of observing systems were operational (i.e., MODIS-Aqua, VIIRS, OLCI-3A, and OLCI-3B)\n([[1]](https://dast.copernicus-climate.eu/documents/satellite-ocean-colour/v6.0/WP2-FDDP-2022-04_C3S2-Lot3_PUGS-of-v6.0-OceanColour-product_v1.1_FINAL.pdf)). \nFiltering chlorophyll-a data based on the availability of valid pixels allows to reduce the bias resulting from the lack of sampling at high latitudes, which leads to artefacts in the apparent seasonal cycle from ocean colour sensors [[14]](https://doi.org/10.1016/j.rse.2007.03.008). However, the temporal distribution for the filtered chlorophyll-a data alignes with typical observations in the Southern Ocean, where chlorophyll-a concentrations decrease during the austral winter (e.g., [[15]](https://doi.org/10.1038/s41467-020-19157-2)). \nLinear trend analysis for all months in all three POOZ sectors yields to results summarised in the table, indicating an overall chlorophyll-a increase in the POOZ, consistent with previous studies (e.g., [[8]](https://doi.org/10.1029/2019GL083163); [[16]](https://doi.org/10.1016/j.marenvres.2024.106856)). However, these trends should be interpreted with caution, as they are based on a low percentage of valid pixels (i.e., 8-15%). Including data prior to 2003 in the analysis would not change the sign of the significant chlorophyll-a trends but could affect their statistical significance. \nOverall, the results indicate that the dataset remains valuable for biogeochemical calibration and validation in the Southern Ocean during months with reliable observations (September to April). In contrast, using data from the local winter (May-August), when satellite-derived chlorophyll-a concentrations have been shown to be overestimated [[14]](https://doi.org/10.1016/j.rse.2007.03.008), could lead to biased model parameterisation. This in turn may result in inflated estimates of primary production and carbon fluxes, as well as a misrepresentation of the ecosystem functioning, including nutrient cycling and grazing pressure. To address these issues, chlorophyll-a data should be filtered based on the availability of valid pixels used for their retrieval and corrected using in situ measurements (e.g., Argo floats, ship-based observations), or through assimilation techniques that incorporate satellite chlorophyll-a data from reliable seasons, enabling the model to estimate missing data based on physical and biogeochemical constraints.\n\n## ℹ️ If you want to know more\n\n### Key resources\n* CDS catalogue entry used in this notebook is the [Ocean colour daily data from 1997 to present derived from satellite observations](https://cds.climate.copernicus.eu/datasets/satellite-ocean-colour?tab=overview)\n* Data download is from [CDS](https://cds.climate.copernicus.eu/datasets/satellite-ocean-colour?tab=form)\n* Product Documentation is available on the [CDS](https://cds.climate.copernicus.eu/datasets/satellite-ocean-colour?tab=documentation) website\n\nCode libraries used:\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)", "text_with_prefix": "EQC Quality Assessment: \"Completeness of ocean colour observations for biogeochemical models\"\nDataset: satellite-ocean-colour [CDS]\nAspect: completeness_q01 | Category: Satellite_ECVs\nSection: Completeness of ocean colour observations for biogeochemical models > Quality assessment statement\n---\n8.408) [[13]](https://doi.org/10.3389/fmars.2024.1363088)).\n\n**Chlorophyll-a temporal distribution:**\n\nThe monthly climatological distribution of valid chlorophyll-a pixels is in line with expectations based on their dependence on the solar zenith angle: the highest percentage (up to about 15%) is observed between November and February, when the solar zenith angle is low, then it decreases to reach its minimum (almost 0%) during the polar night months (i.e. May-July) and gradually increases again towards the end of the year. It is interesting to note that, as shown in the valid pixel heatmap, the valid percentages increased between 2017 and 2019, reaching a maximum in 2018 (up to about 24%), when the highest number of observing systems were operational (i.e., MODIS-Aqua, VIIRS, OLCI-3A, and OLCI-3B)\n([[1]](https://dast.copernicus-climate.eu/documents/satellite-ocean-colour/v6.0/WP2-FDDP-2022-04_C3S2-Lot3_PUGS-of-v6.0-OceanColour-product_v1.1_FINAL.pdf)). \nFiltering chlorophyll-a data based on the availability of valid pixels allows to reduce the bias resulting from the lack of sampling at high latitudes, which leads to artefacts in the apparent seasonal cycle from ocean colour sensors [[14]](https://doi.org/10.1016/j.rse.2007.03.008). However, the temporal distribution for the filtered chlorophyll-a data alignes with typical observations in the Southern Ocean, where chlorophyll-a concentrations decrease during the austral winter (e.g., [[15]](https://doi.org/10.1038/s41467-020-19157-2)). \nLinear trend analysis for all months in all three POOZ sectors yields to results summarised in the table, indicating an overall chlorophyll-a increase in the POOZ, consistent with previous studies (e.g., [[8]](https://doi.org/10.1029/2019GL083163); [[16]](https://doi.org/10.1016/j.marenvres.2024.106856)). However, these trends should be interpreted with caution, as they are based on a low percentage of valid pixels (i.e., 8-15%). Including data prior to 2003 in the analysis would not change the sign of the significant chlorophyll-a trends but could affect their statistical significance. \nOverall, the results indicate that the dataset remains valuable for biogeochemical calibration and validation in the Southern Ocean during months with reliable observations (September to April). In contrast, using data from the local winter (May-August), when satellite-derived chlorophyll-a concentrations have been shown to be overestimated [[14]](https://doi.org/10.1016/j.rse.2007.03.008), could lead to biased model parameterisation. This in turn may result in inflated estimates of primary production and carbon fluxes, as well as a misrepresentation of the ecosystem functioning, including nutrient cycling and grazing pressure. To address these issues, chlorophyll-a data should be filtered based on the availability of valid pixels used for their retrieval and corrected using in situ measurements (e.g., Argo floats, ship-based observations), or through assimilation techniques that incorporate satellite chlorophyll-a data from reliable seasons, enabling the model to estimate missing data based on physical and biogeochemical constraints.\n\n## ℹ️ If you want to know more\n\n### Key resources\n* CDS catalogue entry used in this notebook is the [Ocean colour daily data from 1997 to present derived from satellite observations](https://cds.climate.copernicus.eu/datasets/satellite-ocean-colour?tab=overview)\n* Data download is from [CDS](https://cds.climate.copernicus.eu/datasets/satellite-ocean-colour?tab=form)\n* Product Documentation is available on the [CDS](https://cds.climate.copernicus.eu/datasets/satellite-ocean-colour?tab=documentation) website\n\nCode libraries used:\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)"} {"chunk_id": "satellite_satellite-ocean-colour_completeness_q01__dfbeea2b375e", "report_id": "satellite_satellite-ocean-colour_completeness_q01", "dataset_id": "satellite-ocean-colour", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q01", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Completeness of ocean colour observations for biogeochemical models > Quality assessment statement", "title": "Completeness of ocean colour observations for biogeochemical models", "chunk_index": 4, "token_count": 1019, "text_raw": "open/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)\n\nFurther readings:\n* Ocean Colour Monitoring Portals: [CMEMS](https://marine.copernicus.eu/), [ESA-GlobColour Project](https://hermes.acri.fr), [ESA-OC-CCI](https://climate.esa.int/en/projects/ocean-colour/), [ESA-Sentinel 3](https://sentiwiki.copernicus.eu/web/s3-mission), [IMOS](https://imos.org.au/facility/satellite-remote-sensing/ocean-colour), [NASA](https://oceancolor.gsfc.nasa.gov/#), [NOAA CoastWatch](https://coastwatch.noaa.gov/cwn/index.html)\n* Use cases: [EUMETSAT](https://user.eumetsat.int/search-view?sort=startDate%20desc&facets=%7B%22theme%22:%5B%22Marine%7COcean%20biogeochemistry%22%5D,%22contentTypes%22:%5B%22Resources%7CCase%20studies%22%5D%7D), [C3S](https://climate.copernicus.eu/ocean)\n* Groom, S., et al. (2019). [Satellite Ocean Colour: Current Status and Future Perspective](https://www.frontiersin.org/journals/marine-science/articles/10.3389/fmars.2019.00485/full). Frontiers in Marine Science, 6.\n* [International Ocean Colour Coordinating Group (ICCG)](https://ioccg.org/)\n* [Southern Ocean Carbon and Climate Observations and Modeling (SOCCOM)](https://soccom.princeton.edu)\n\n### References\n\n[[1]](https://dast.copernicus-climate.eu/documents/satellite-ocean-colour/v6.0/WP2-FDDP-2022-04_C3S2-Lot3_PUGS-of-v6.0-OceanColour-product_v1.1_FINAL.pdf) Jackson, T., et al. (2023). C3S Ocean Colour Version 6.0: Product User Guide and Specification. Issue 1.1. E.U. Copernicus Climate Change Service. Document ref. WP2-FDDP-2022-04_C3S2-Lot3_PUGS-of-v6.0-OceanColour-product.\n\n[[2]](https://dast.copernicus-climate.eu/documents/satellite-ocean-colour/v6.0/WP2-FDDP-2022-04_C3S2-Lot3_ATBD-of-v6.0-OceanColour-product_v1.1_FINAL.pdf) Jackson, T., et al. (2022). C3S Ocean Colour Version 6.0: Algorithm Theoretical Basis Document. Issue 1.1. E.U. Copernicus Climate Change Service. Document ref. WP2-FDDP-2022-04_C3S2-Lot3_ATBD-of-v6.0-OceanColour-product.\n\n[[3]](https://doi.org/10.1016/j.jmarsys.2013.02.007) Doron, M., Brasseur, P., Brankart, J.-M., Losa, S.N., & Melet, A. (2013). Stochastic estimation of biogeochemical parameters from Globcolour ocean colour satellite data in a North Atlantic 3D ocean coupled physical–biogeochemical model. Journal of Marine Systems, 117-118, 91-95.\n\n[[4]](https://doi.org/10.3390/rs10101666) Sammartino, M., Marullo, S., Santoleri, R., & Scardi, M. (2018). Modelling the Vertical Distribution of Phytoplankton Biomass in the Mediterranean Sea from Satellite Data: A Neural Network Approach. Remote Sensing, 10(10), 1666.\n\n[[5]](https://doi.org/10.1038/s41467-019-08457-x) Dutkiewicz, S., Hickman, A.E., Jahn, O., Henson, S., Beaulieu, C., & Monier, E. (2019) Ocean colour signature of climate change. Nature Communications, 10, 578.", "text_with_prefix": "EQC Quality Assessment: \"Completeness of ocean colour observations for biogeochemical models\"\nDataset: satellite-ocean-colour [CDS]\nAspect: completeness_q01 | Category: Satellite_ECVs\nSection: Completeness of ocean colour observations for biogeochemical models > Quality assessment statement\n---\nopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)\n\nFurther readings:\n* Ocean Colour Monitoring Portals: [CMEMS](https://marine.copernicus.eu/), [ESA-GlobColour Project](https://hermes.acri.fr), [ESA-OC-CCI](https://climate.esa.int/en/projects/ocean-colour/), [ESA-Sentinel 3](https://sentiwiki.copernicus.eu/web/s3-mission), [IMOS](https://imos.org.au/facility/satellite-remote-sensing/ocean-colour), [NASA](https://oceancolor.gsfc.nasa.gov/#), [NOAA CoastWatch](https://coastwatch.noaa.gov/cwn/index.html)\n* Use cases: [EUMETSAT](https://user.eumetsat.int/search-view?sort=startDate%20desc&facets=%7B%22theme%22:%5B%22Marine%7COcean%20biogeochemistry%22%5D,%22contentTypes%22:%5B%22Resources%7CCase%20studies%22%5D%7D), [C3S](https://climate.copernicus.eu/ocean)\n* Groom, S., et al. (2019). [Satellite Ocean Colour: Current Status and Future Perspective](https://www.frontiersin.org/journals/marine-science/articles/10.3389/fmars.2019.00485/full). Frontiers in Marine Science, 6.\n* [International Ocean Colour Coordinating Group (ICCG)](https://ioccg.org/)\n* [Southern Ocean Carbon and Climate Observations and Modeling (SOCCOM)](https://soccom.princeton.edu)\n\n### References\n\n[[1]](https://dast.copernicus-climate.eu/documents/satellite-ocean-colour/v6.0/WP2-FDDP-2022-04_C3S2-Lot3_PUGS-of-v6.0-OceanColour-product_v1.1_FINAL.pdf) Jackson, T., et al. (2023). C3S Ocean Colour Version 6.0: Product User Guide and Specification. Issue 1.1. E.U. Copernicus Climate Change Service. Document ref. WP2-FDDP-2022-04_C3S2-Lot3_PUGS-of-v6.0-OceanColour-product.\n\n[[2]](https://dast.copernicus-climate.eu/documents/satellite-ocean-colour/v6.0/WP2-FDDP-2022-04_C3S2-Lot3_ATBD-of-v6.0-OceanColour-product_v1.1_FINAL.pdf) Jackson, T., et al. (2022). C3S Ocean Colour Version 6.0: Algorithm Theoretical Basis Document. Issue 1.1. E.U. Copernicus Climate Change Service. Document ref. WP2-FDDP-2022-04_C3S2-Lot3_ATBD-of-v6.0-OceanColour-product.\n\n[[3]](https://doi.org/10.1016/j.jmarsys.2013.02.007) Doron, M., Brasseur, P., Brankart, J.-M., Losa, S.N., & Melet, A. (2013). Stochastic estimation of biogeochemical parameters from Globcolour ocean colour satellite data in a North Atlantic 3D ocean coupled physical–biogeochemical model. Journal of Marine Systems, 117-118, 91-95.\n\n[[4]](https://doi.org/10.3390/rs10101666) Sammartino, M., Marullo, S., Santoleri, R., & Scardi, M. (2018). Modelling the Vertical Distribution of Phytoplankton Biomass in the Mediterranean Sea from Satellite Data: A Neural Network Approach. Remote Sensing, 10(10), 1666.\n\n[[5]](https://doi.org/10.1038/s41467-019-08457-x) Dutkiewicz, S., Hickman, A.E., Jahn, O., Henson, S., Beaulieu, C., & Monier, E. (2019) Ocean colour signature of climate change. Nature Communications, 10, 578."} {"chunk_id": "satellite_satellite-ocean-colour_completeness_q01__b47dc34b835d", "report_id": "satellite_satellite-ocean-colour_completeness_q01", "dataset_id": "satellite-ocean-colour", "store": "CDS", "doc_type": "EQC_QA", "aspect": "completeness_q01", "aspect_base": "completeness", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Completeness of ocean colour observations for biogeochemical models > Quality assessment statement", "title": "Completeness of ocean colour observations for biogeochemical models", "chunk_index": 5, "token_count": 1009, "text_raw": "A.E., Jahn, O., Henson, S., Beaulieu, C., & Monier, E. (2019) Ocean colour signature of climate change. Nature Communications, 10, 578.\n\n[[6]](https://doi.org/10.1029/2018JC014329) Pradhan, H.K., Völker, C., Losa, S. N., Bracher, A., & Nerger, L. (2019). Assimilation of global total chlorophyll OC-CCI data and its impact on individual phytoplankton fields. Journal of Geophysical Research: Oceans, 124, 470-490.\n\n[[7]](https://doi.org/10.1016/0967-0637%2895%2900021-W) Orsi, A.H., Whitworth, T., & Nowlin, W.D. (1995). On the meridional extent and fronts of the Antarctic Circumpolar Current. Deep Sea Research Part I: Oceanographic Research Papers, 42(5), 641-673.\n\n[[8]](https://doi.org/10.1029/2019GL083163) Del Castillo, C.E., Signorini, S.R., Karaköylü, E.M., & Rivero-Calle, S. (2019). Is the Southern Ocean Getting Greener? Geophysical Research Letters, 46, 6034-6040.\n\n[[9]](https://doi.org/10.1029/95JC00458) Campbell, J.W. (1995). The lognormal distribution as a model for bio-optical variability in the sea. Journal of Geophysical Research, 100, 13237-13254.\n\n[[10]](https://doi.org/10.3390/s19194285) Sathyendranath, S., et al. (2019). An Ocean-Colour time series for use in climate studies: the experience of the Ocean-Colour Climate Change Initiative (OC-CCI). Sensors, 19, 4285.\n\n[[11]](https://doi.org/10.1029/2006jc004072) Sokolov, S., & Rintoul, S.R. (2007). On the relationship between fronts of the Antarctic Circumpolar Current and surface chlorophyll concentrations in the Southern Ocean, Journal of Geophysical Research, 112, C07030.\n\n[[12]](https://doi.org/10.5670/oceanog.2018.408) Paparazzo, F.E., Crespi-Abril, A.C., Gonçalves, R.J., Barbieri, E.S., Gracia Villalobos, L.L., Solís, M.E., & Soria, G. (2018). Patagonian dust as a source of macronutrients in the Southwest Atlantic Ocean. Oceanography, 31(4), 33-39.\n\n[[13]](https://doi.org/10.3389/fmars.2024.1363088) Demasy, C., Boye, M., Lai, B., Burckel, P., Feng, Y., Losno, R., Borensztajn, S., & Besson, P. (2024). Iron dissolution from Patagonian dust in the Southern Ocean: under present and future conditions. Frontiers in Marine Science, 11, 1363088.\n\n[[14]](https://doi.org/10.1016/j.rse.2007.03.008) Gregg, W.W., & Casey, N.W. (2007). Sampling biases in MODIS and SeaWiFS ocean chlorophyll data. Remote Sensing of Environment, 111(1), 25-35.\n\n[[15]](https://doi.org/10.1038/s41467-020-19157-2) Arteaga, L.A., Boss, E., Behrenfeld, M.J., Westberry, T.K., & Sarmiento, J.L. (2020). Seasonal modulation of phytoplankton biomass in the Southern Ocean. Nature Communications, 11, 5364.\n\n[[16]](https://doi.org/10.1016/j.marenvres.2024.106856) Venegas, R.M., Rivas, D., & Treml, E. (2025). Global climate-driven sea surface temperature and chlorophyll dynamics. Marine Environmental Research, 204, 106856", "text_with_prefix": "EQC Quality Assessment: \"Completeness of ocean colour observations for biogeochemical models\"\nDataset: satellite-ocean-colour [CDS]\nAspect: completeness_q01 | Category: Satellite_ECVs\nSection: Completeness of ocean colour observations for biogeochemical models > Quality assessment statement\n---\nA.E., Jahn, O., Henson, S., Beaulieu, C., & Monier, E. (2019) Ocean colour signature of climate change. Nature Communications, 10, 578.\n\n[[6]](https://doi.org/10.1029/2018JC014329) Pradhan, H.K., Völker, C., Losa, S. N., Bracher, A., & Nerger, L. (2019). Assimilation of global total chlorophyll OC-CCI data and its impact on individual phytoplankton fields. Journal of Geophysical Research: Oceans, 124, 470-490.\n\n[[7]](https://doi.org/10.1016/0967-0637%2895%2900021-W) Orsi, A.H., Whitworth, T., & Nowlin, W.D. (1995). On the meridional extent and fronts of the Antarctic Circumpolar Current. Deep Sea Research Part I: Oceanographic Research Papers, 42(5), 641-673.\n\n[[8]](https://doi.org/10.1029/2019GL083163) Del Castillo, C.E., Signorini, S.R., Karaköylü, E.M., & Rivero-Calle, S. (2019). Is the Southern Ocean Getting Greener? Geophysical Research Letters, 46, 6034-6040.\n\n[[9]](https://doi.org/10.1029/95JC00458) Campbell, J.W. (1995). The lognormal distribution as a model for bio-optical variability in the sea. Journal of Geophysical Research, 100, 13237-13254.\n\n[[10]](https://doi.org/10.3390/s19194285) Sathyendranath, S., et al. (2019). An Ocean-Colour time series for use in climate studies: the experience of the Ocean-Colour Climate Change Initiative (OC-CCI). Sensors, 19, 4285.\n\n[[11]](https://doi.org/10.1029/2006jc004072) Sokolov, S., & Rintoul, S.R. (2007). On the relationship between fronts of the Antarctic Circumpolar Current and surface chlorophyll concentrations in the Southern Ocean, Journal of Geophysical Research, 112, C07030.\n\n[[12]](https://doi.org/10.5670/oceanog.2018.408) Paparazzo, F.E., Crespi-Abril, A.C., Gonçalves, R.J., Barbieri, E.S., Gracia Villalobos, L.L., Solís, M.E., & Soria, G. (2018). Patagonian dust as a source of macronutrients in the Southwest Atlantic Ocean. Oceanography, 31(4), 33-39.\n\n[[13]](https://doi.org/10.3389/fmars.2024.1363088) Demasy, C., Boye, M., Lai, B., Burckel, P., Feng, Y., Losno, R., Borensztajn, S., & Besson, P. (2024). Iron dissolution from Patagonian dust in the Southern Ocean: under present and future conditions. Frontiers in Marine Science, 11, 1363088.\n\n[[14]](https://doi.org/10.1016/j.rse.2007.03.008) Gregg, W.W., & Casey, N.W. (2007). Sampling biases in MODIS and SeaWiFS ocean chlorophyll data. Remote Sensing of Environment, 111(1), 25-35.\n\n[[15]](https://doi.org/10.1038/s41467-020-19157-2) Arteaga, L.A., Boss, E., Behrenfeld, M.J., Westberry, T.K., & Sarmiento, J.L. (2020). Seasonal modulation of phytoplankton biomass in the Southern Ocean. Nature Communications, 11, 5364.\n\n[[16]](https://doi.org/10.1016/j.marenvres.2024.106856) Venegas, R.M., Rivas, D., & Treml, E. (2025). Global climate-driven sea surface temperature and chlorophyll dynamics. Marine Environmental Research, 204, 106856"} {"chunk_id": "satellite_satellite-ocean-colour_consistency-assessment_q01__ee2de7eb833b", "report_id": "satellite_satellite-ocean-colour_consistency-assessment_q01", "dataset_id": "satellite-ocean-colour", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency-assessment_q01", "aspect_base": "consistency-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Impact of using successive satellites on Ocean Colour time series", "title": "Impact of using successive satellites on Ocean Colour time series", "chunk_index": 0, "token_count": 95, "text_raw": "Production Date: 16-03-2024 \nProduced by: Chiara Volta (ENEA, Italy) and Salvatore Marullo (CNR, Italy)", "text_with_prefix": "EQC Quality Assessment: \"Impact of using successive satellites on Ocean Colour time series\"\nDataset: satellite-ocean-colour [CDS]\nAspect: consistency-assessment_q01 | Category: Satellite_ECVs\nSection: Impact of using successive satellites on Ocean Colour time series\n---\nProduction Date: 16-03-2024 \nProduced by: Chiara Volta (ENEA, Italy) and Salvatore Marullo (CNR, Italy)"} {"chunk_id": "satellite_satellite-ocean-colour_consistency-assessment_q01__2c7bc2ffdb1f", "report_id": "satellite_satellite-ocean-colour_consistency-assessment_q01", "dataset_id": "satellite-ocean-colour", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency-assessment_q01", "aspect_base": "consistency-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Impact of using successive satellites on Ocean Colour time series > Quality assessment question", "title": "Impact of using successive satellites on Ocean Colour time series", "chunk_index": 1, "token_count": 519, "text_raw": "**Does the consistency of chlorophyll-a and remote sensing reflectance get affected by the succession of satellite sensors over time?**\n\nThe Ocean Colour dataset version 6.0, as produced for the Copernicus Climate Change Service (C3S), includes the variables, mass concentration of chlorophyll-a and remote sensing reflectance (Rrs) for six wavelengths from October 1997 to present [[1]](https://dast.copernicus-climate.eu/documents/satellite-ocean-colour/v6.0/WP2-FDDP-2022-04_C3S2-Lot3_PUGS-of-v6.0-OceanColour-product_v1.1_FINAL.pdf). \nChlorophyll-a concentration is typically used as proxy for algal biomass and fisheries production [[2]](https://doi.org/10.4319/lo.1997.42.7.1479) [[3]](https://doi.org/10.1371/journal.pone.0028945) and is derived through specific algorithms which rely on remote sensing reflectance (Rrs) [[4]](https://dast.copernicus-climate.eu/documents/satellite-ocean-colour/v6.0/WP2-FDDP-2022-04_C3S2-Lot3_ATBD-of-v6.0-OceanColour-product_v1.1_FINAL.pdf). \nThe dataset combines measurements carried out by six satellite sensors [[1]](https://dast.copernicus-climate.eu/documents/satellite-ocean-colour/v6.0/WP2-FDDP-2022-04_C3S2-Lot3_PUGS-of-v6.0-OceanColour-product_v1.1_FINAL.pdf): SeaWiFS, MERIS, MODIS-Aqua, VIIRS, OLCI-3A and OLCI-3B. No sensor was operational over the whole temporal coverage of the datasets and only OLCI-3A and OLCI-3B are still operational. \nHere, the goal is to assess if the stability of Ocean Colour time series is affected by the transition of different satellite sensors over time.", "text_with_prefix": "EQC Quality Assessment: \"Impact of using successive satellites on Ocean Colour time series\"\nDataset: satellite-ocean-colour [CDS]\nAspect: consistency-assessment_q01 | Category: Satellite_ECVs\nSection: Impact of using successive satellites on Ocean Colour time series > Quality assessment question\n---\n**Does the consistency of chlorophyll-a and remote sensing reflectance get affected by the succession of satellite sensors over time?**\n\nThe Ocean Colour dataset version 6.0, as produced for the Copernicus Climate Change Service (C3S), includes the variables, mass concentration of chlorophyll-a and remote sensing reflectance (Rrs) for six wavelengths from October 1997 to present [[1]](https://dast.copernicus-climate.eu/documents/satellite-ocean-colour/v6.0/WP2-FDDP-2022-04_C3S2-Lot3_PUGS-of-v6.0-OceanColour-product_v1.1_FINAL.pdf). \nChlorophyll-a concentration is typically used as proxy for algal biomass and fisheries production [[2]](https://doi.org/10.4319/lo.1997.42.7.1479) [[3]](https://doi.org/10.1371/journal.pone.0028945) and is derived through specific algorithms which rely on remote sensing reflectance (Rrs) [[4]](https://dast.copernicus-climate.eu/documents/satellite-ocean-colour/v6.0/WP2-FDDP-2022-04_C3S2-Lot3_ATBD-of-v6.0-OceanColour-product_v1.1_FINAL.pdf). \nThe dataset combines measurements carried out by six satellite sensors [[1]](https://dast.copernicus-climate.eu/documents/satellite-ocean-colour/v6.0/WP2-FDDP-2022-04_C3S2-Lot3_PUGS-of-v6.0-OceanColour-product_v1.1_FINAL.pdf): SeaWiFS, MERIS, MODIS-Aqua, VIIRS, OLCI-3A and OLCI-3B. No sensor was operational over the whole temporal coverage of the datasets and only OLCI-3A and OLCI-3B are still operational. \nHere, the goal is to assess if the stability of Ocean Colour time series is affected by the transition of different satellite sensors over time."} {"chunk_id": "satellite_satellite-ocean-colour_consistency-assessment_q01__3eb4a4ffdaf1", "report_id": "satellite_satellite-ocean-colour_consistency-assessment_q01", "dataset_id": "satellite-ocean-colour", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency-assessment_q01", "aspect_base": "consistency-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Impact of using successive satellites on Ocean Colour time series > Quality assessment statement", "title": "Impact of using successive satellites on Ocean Colour time series", "chunk_index": 2, "token_count": 1035, "text_raw": "These are the key outcomes of this assessment\n\nThe Ocean Colour dataset version 6.0: \n* provides a time consistent record of remote sensing reflectance at 443 and 560 nm, and derived chlorophyll-a concentrations, which is not affected by the succession of different satellite sensors. However, data prior to March 2002 should be carefully evaluated. \n* would overestimate chlorophyll-a concentrations before March 2002, especially when high-latitude, more productive oceanic regions are included in the analysis. \n* indicates decreasing chlorophyll-a trends in the more productive regions analysed, while the most oligotrophic area exhibits an increasing trend.\n\n![Graphical_abstract.png](attachment:4244bac9-1f7d-45b9-a358-ee1e46162ae8.png)\n
Fig.1 The top panel shows the 48-month rolling mean of chlorophyll-a concentrations calculated in the four regions selected for the analysis, as well as the name and duration of each satellite mission used to produce the dataset. Mean chlorophyll-a concentrations are also provided in the legend. The lower panel shows a global map of the 25-year average chlorophyll-a concentration. Dashed, dotted and yellow boxes on the map indicate the oceanic region between latitudes 50°S-50°N, the NASTG's region and the STG's region, respectively.
\n\n## 📋 Methodology\n\nThis notebook provides an assessment of the ability of Ocean Colour dataset version 6.0 to describe the temporal variability of chlorophyll-a concentrations and Rrs at 443 and 560 nm over different oceanic regions.\n\nThe analysis and results are detailed in the sections below:\n\n[](section-1)\n * Import required packages\n * Define parameters to be analysed (time period, variables, regions) and data request\n\n[](section-2)\n * Area average function\n * Retrieve daily averages for the selected variables, time period and regions\n\n[](section-3)\n * Compute statistics and define plot function\n * Define satellite missions\n\n[](section-4)\n * Display results \n * Discussion\n\n## 📈 Analysis and results\n(section-1)=\n### 1. Choose the data to use and set up the code\n#### Import required packages\nBesides the standard libraries used to manage and analyse multidimensional arrays, the [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control) `c3s_eqc_automatic_quality_control`is imported to download data and calculate statistics.\n\n#### Define parameters to be analysed (time period, variables, regions) and data request\nThe analysis performed in this notebook focuses on the time series of chlorophyll-a concentration and Rrs at 443 and 560 nm, the most commonly used bands to derive chlorophyll-a data, over a 25-year period (January 1998 - December 2022), in four regions: the global ocean, the oceanic region between 50°S and 50°N, the region where the North Atlantic SubTropical Gyre (NASTG) is located [[5]](https://doi.org/10.1029/2021GL096965) and a 10°x10° area, within the South Pacific Gyre (SPG), over which the inter-sensor bias-correction process was validated [[6]](https://dast.copernicus-climate.eu/documents/satellite-ocean-colour/v6.0/WP2-FDDP-2022-04_C3S2-Lot3_PQAR-of-v6.0-OceanColour-product_v1.1_FINAL.pdf).\n\nEnforce sorting as original data\nDefine data request\n\n(section-2)=\n### 2. Area average function and data retrieval\n#### Area average function\nThe `regionalised_spatial_weighted_mean` function, which accounts for the varying surface area at different latitudes, is defined to calculate the area averages over the selected regions. Area averages for chlorophyll-a concentration are computed using log-transformed daily data, and then unlogged [[7]](https://doi.org/10.1029/95JC00458).\n\n#### Data retrieval\nThe `regionalised_spatial_weighted_mean` is applied to the selected variables and regions, and the time series are downloaded as a single array. Chlorophyll-a concentrations outside the range 0.01-100 mg m-3 are excluded from the download and the analysis [[8]](https://doi.org/10.3390/s19194285).", "text_with_prefix": "EQC Quality Assessment: \"Impact of using successive satellites on Ocean Colour time series\"\nDataset: satellite-ocean-colour [CDS]\nAspect: consistency-assessment_q01 | Category: Satellite_ECVs\nSection: Impact of using successive satellites on Ocean Colour time series > Quality assessment statement\n---\nThese are the key outcomes of this assessment\n\nThe Ocean Colour dataset version 6.0: \n* provides a time consistent record of remote sensing reflectance at 443 and 560 nm, and derived chlorophyll-a concentrations, which is not affected by the succession of different satellite sensors. However, data prior to March 2002 should be carefully evaluated. \n* would overestimate chlorophyll-a concentrations before March 2002, especially when high-latitude, more productive oceanic regions are included in the analysis. \n* indicates decreasing chlorophyll-a trends in the more productive regions analysed, while the most oligotrophic area exhibits an increasing trend.\n\n![Graphical_abstract.png](attachment:4244bac9-1f7d-45b9-a358-ee1e46162ae8.png)\n
Fig.1 The top panel shows the 48-month rolling mean of chlorophyll-a concentrations calculated in the four regions selected for the analysis, as well as the name and duration of each satellite mission used to produce the dataset. Mean chlorophyll-a concentrations are also provided in the legend. The lower panel shows a global map of the 25-year average chlorophyll-a concentration. Dashed, dotted and yellow boxes on the map indicate the oceanic region between latitudes 50°S-50°N, the NASTG's region and the STG's region, respectively.
\n\n## 📋 Methodology\n\nThis notebook provides an assessment of the ability of Ocean Colour dataset version 6.0 to describe the temporal variability of chlorophyll-a concentrations and Rrs at 443 and 560 nm over different oceanic regions.\n\nThe analysis and results are detailed in the sections below:\n\n[](section-1)\n * Import required packages\n * Define parameters to be analysed (time period, variables, regions) and data request\n\n[](section-2)\n * Area average function\n * Retrieve daily averages for the selected variables, time period and regions\n\n[](section-3)\n * Compute statistics and define plot function\n * Define satellite missions\n\n[](section-4)\n * Display results \n * Discussion\n\n## 📈 Analysis and results\n(section-1)=\n### 1. Choose the data to use and set up the code\n#### Import required packages\nBesides the standard libraries used to manage and analyse multidimensional arrays, the [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control) `c3s_eqc_automatic_quality_control`is imported to download data and calculate statistics.\n\n#### Define parameters to be analysed (time period, variables, regions) and data request\nThe analysis performed in this notebook focuses on the time series of chlorophyll-a concentration and Rrs at 443 and 560 nm, the most commonly used bands to derive chlorophyll-a data, over a 25-year period (January 1998 - December 2022), in four regions: the global ocean, the oceanic region between 50°S and 50°N, the region where the North Atlantic SubTropical Gyre (NASTG) is located [[5]](https://doi.org/10.1029/2021GL096965) and a 10°x10° area, within the South Pacific Gyre (SPG), over which the inter-sensor bias-correction process was validated [[6]](https://dast.copernicus-climate.eu/documents/satellite-ocean-colour/v6.0/WP2-FDDP-2022-04_C3S2-Lot3_PQAR-of-v6.0-OceanColour-product_v1.1_FINAL.pdf).\n\nEnforce sorting as original data\nDefine data request\n\n(section-2)=\n### 2. Area average function and data retrieval\n#### Area average function\nThe `regionalised_spatial_weighted_mean` function, which accounts for the varying surface area at different latitudes, is defined to calculate the area averages over the selected regions. Area averages for chlorophyll-a concentration are computed using log-transformed daily data, and then unlogged [[7]](https://doi.org/10.1029/95JC00458).\n\n#### Data retrieval\nThe `regionalised_spatial_weighted_mean` is applied to the selected variables and regions, and the time series are downloaded as a single array. Chlorophyll-a concentrations outside the range 0.01-100 mg m-3 are excluded from the download and the analysis [[8]](https://doi.org/10.3390/s19194285)."} {"chunk_id": "satellite_satellite-ocean-colour_consistency-assessment_q01__46578ed647bf", "report_id": "satellite_satellite-ocean-colour_consistency-assessment_q01", "dataset_id": "satellite-ocean-colour", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency-assessment_q01", "aspect_base": "consistency-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Impact of using successive satellites on Ocean Colour time series > Quality assessment statement", "title": "Impact of using successive satellites on Ocean Colour time series", "chunk_index": 3, "token_count": 1051, "text_raw": "ophyll-a concentrations outside the range 0.01-100 mg m-3 are excluded from the download and the analysis [[8]](https://doi.org/10.3390/s19194285).\n\n(section-3)=\n### 3. Plot and statistics functions\n#### Compute statistics and define plot function\nFunctions to calculate 48-month rolling mean and linear trend for the selected variables and regions are embedded in the plot function. Linear trend equations and their statistical significance (p) are specified in the legend of each figure. Time-averaged values are also computed for the selected variables and regions.\n\nCompute running mean\nCompute linear trend\nFinal settings\n\n#### Define satellite missions\nThe name and data acquiring period are specified for each satellite sensor used to produce the dataset [[1]](https://dast.copernicus-climate.eu/documents/satellite-ocean-colour/v6.0/WP2-FDDP-2022-04_C3S2-Lot3_PUGS-of-v6.0-OceanColour-product_v1.1_FINAL.pdf). This information is displayed in the time series figures (see below).\n\n(section-4)=\n### 4. Display and discuss results\n#### Display results\nTime series of chlorophyll-a concentrations and Rrs at 443 and 560 nm are displayed for each region analysed, together with the duration of every satellite mission used to retrieve data. Linear trend equations and their statistical significance are indicated in the legends. Data are plotted on the same zoomed-in scale to focus on long-term chlorophyll-a and Rrs trends in the four regions analysed and facilitate their comparison.\n\n#### Discussion\nResults show that the Ocean Colour dataset version 6.0 captures both the seasonal and interannual variability of the variables analysed in the four selected regions.\n\n**Chlorophyll-a:** \nChlorophyll-a timeseries in the global ocean and the region between 50°S and 50°N show similar seasonal and interannual variabilities, and trends. Long-term trends obtained through a 48-month rolling mean show a maxima around 2002 in both the global ocean and the oceanic region between latitudes 50°S and 50°N, followed by a sharp decrease until 2004. Afterwards, a progressive decrease until around 2015 and a slight increase until 2022 are observed.\nThe largest chlorophyll-a seasonal variability is observed in the NASTG's region, due to the strong seasonal expansion/contraction cycle of the gyre [[5]](https://doi.org/10.1029/2021GL096965), together with a rather constant long-term decreasing trend.\nThe lowest chlorophyll-a concentrations and seasonal variability are observed in the SPG's region and indicate a high stability in the area [[6]](https://dast.copernicus-climate.eu/documents/satellite-ocean-colour/v6.0/WP2-FDDP-2022-04_C3S2-Lot3_PQAR-of-v6.0-OceanColour-product_v1.1_FINAL.pdf). Linear trends are statistically significant (p<0.05) in all regions analyzed here, indicating that chlorophyll-a is decreasing everywhere except in the SPG. These results align with the recent paper [[9]](https://doi.org/10.1038/s41586-023-06321-z), where decreasing (increasing) chlorophyll-a is associated with increasing (decreasing) sea surface temperature trends.\n\n**Rrs:** \nCompared to chlorophyll-a, average Rrs at 443 nm shows reverse patterns and statistically significant (p<0.05) trends in all selected regions, which are consistent with results in [[9]](https://doi.org/10.1038/s41586-023-06321-z). It is worth noting that results for the SPG's region are consistent with the post-bias correction validation [[6]](https://dast.copernicus-climate.eu/documents/satellite-ocean-colour/v6.0/WP2-FDDP-2022-04_C3S2-Lot3_PQAR-of-v6.0-OceanColour-product_v1.1_FINAL.pdf) and show no suspicious artefacts along with the satellite succession.\nRrs at 560 nm trends are rather linear in all the regions analysed here, except the NASTG, where a progressive increase between 1998 and 2017, followed by a rapid decrease until 2022, is observed. Linear trends calculated here for Rrs at 560 nm are not statistically significant (p>0.05), although an increasing tendency is observed everywhere, except in the SPG where Rrs at 560 nm decreases over time.", "text_with_prefix": "EQC Quality Assessment: \"Impact of using successive satellites on Ocean Colour time series\"\nDataset: satellite-ocean-colour [CDS]\nAspect: consistency-assessment_q01 | Category: Satellite_ECVs\nSection: Impact of using successive satellites on Ocean Colour time series > Quality assessment statement\n---\nophyll-a concentrations outside the range 0.01-100 mg m-3 are excluded from the download and the analysis [[8]](https://doi.org/10.3390/s19194285).\n\n(section-3)=\n### 3. Plot and statistics functions\n#### Compute statistics and define plot function\nFunctions to calculate 48-month rolling mean and linear trend for the selected variables and regions are embedded in the plot function. Linear trend equations and their statistical significance (p) are specified in the legend of each figure. Time-averaged values are also computed for the selected variables and regions.\n\nCompute running mean\nCompute linear trend\nFinal settings\n\n#### Define satellite missions\nThe name and data acquiring period are specified for each satellite sensor used to produce the dataset [[1]](https://dast.copernicus-climate.eu/documents/satellite-ocean-colour/v6.0/WP2-FDDP-2022-04_C3S2-Lot3_PUGS-of-v6.0-OceanColour-product_v1.1_FINAL.pdf). This information is displayed in the time series figures (see below).\n\n(section-4)=\n### 4. Display and discuss results\n#### Display results\nTime series of chlorophyll-a concentrations and Rrs at 443 and 560 nm are displayed for each region analysed, together with the duration of every satellite mission used to retrieve data. Linear trend equations and their statistical significance are indicated in the legends. Data are plotted on the same zoomed-in scale to focus on long-term chlorophyll-a and Rrs trends in the four regions analysed and facilitate their comparison.\n\n#### Discussion\nResults show that the Ocean Colour dataset version 6.0 captures both the seasonal and interannual variability of the variables analysed in the four selected regions.\n\n**Chlorophyll-a:** \nChlorophyll-a timeseries in the global ocean and the region between 50°S and 50°N show similar seasonal and interannual variabilities, and trends. Long-term trends obtained through a 48-month rolling mean show a maxima around 2002 in both the global ocean and the oceanic region between latitudes 50°S and 50°N, followed by a sharp decrease until 2004. Afterwards, a progressive decrease until around 2015 and a slight increase until 2022 are observed.\nThe largest chlorophyll-a seasonal variability is observed in the NASTG's region, due to the strong seasonal expansion/contraction cycle of the gyre [[5]](https://doi.org/10.1029/2021GL096965), together with a rather constant long-term decreasing trend.\nThe lowest chlorophyll-a concentrations and seasonal variability are observed in the SPG's region and indicate a high stability in the area [[6]](https://dast.copernicus-climate.eu/documents/satellite-ocean-colour/v6.0/WP2-FDDP-2022-04_C3S2-Lot3_PQAR-of-v6.0-OceanColour-product_v1.1_FINAL.pdf). Linear trends are statistically significant (p<0.05) in all regions analyzed here, indicating that chlorophyll-a is decreasing everywhere except in the SPG. These results align with the recent paper [[9]](https://doi.org/10.1038/s41586-023-06321-z), where decreasing (increasing) chlorophyll-a is associated with increasing (decreasing) sea surface temperature trends.\n\n**Rrs:** \nCompared to chlorophyll-a, average Rrs at 443 nm shows reverse patterns and statistically significant (p<0.05) trends in all selected regions, which are consistent with results in [[9]](https://doi.org/10.1038/s41586-023-06321-z). It is worth noting that results for the SPG's region are consistent with the post-bias correction validation [[6]](https://dast.copernicus-climate.eu/documents/satellite-ocean-colour/v6.0/WP2-FDDP-2022-04_C3S2-Lot3_PQAR-of-v6.0-OceanColour-product_v1.1_FINAL.pdf) and show no suspicious artefacts along with the satellite succession.\nRrs at 560 nm trends are rather linear in all the regions analysed here, except the NASTG, where a progressive increase between 1998 and 2017, followed by a rapid decrease until 2022, is observed. Linear trends calculated here for Rrs at 560 nm are not statistically significant (p>0.05), although an increasing tendency is observed everywhere, except in the SPG where Rrs at 560 nm decreases over time."} {"chunk_id": "satellite_satellite-ocean-colour_consistency-assessment_q01__371a0bee1778", "report_id": "satellite_satellite-ocean-colour_consistency-assessment_q01", "dataset_id": "satellite-ocean-colour", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency-assessment_q01", "aspect_base": "consistency-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Impact of using successive satellites on Ocean Colour time series > Quality assessment statement", "title": "Impact of using successive satellites on Ocean Colour time series", "chunk_index": 4, "token_count": 1060, "text_raw": ". Linear trends calculated here for Rrs at 560 nm are not statistically significant (p>0.05), although an increasing tendency is observed everywhere, except in the SPG where Rrs at 560 nm decreases over time.\n\nRelatively high chlorophyll-a concentrations and low Rrs at 443 nm are observed when SeaWiFS was the only active satellite (i.e., from January 1998 to March 2002) in both the global ocean and the oceanic region between 50°S and 50°N, but not in the NASTG's and SPG's regions. \nOverall, these results suggest that, prior to March 2002, Rrs at 443 would be underestimated, especially when the analysis is extended to include high-latitude, more productive oceanic zones, and trigger chlorophyll-a overestimates.\nDespite the extensive time series analyzed here and the relatively short time series length required to detect climate-driven variations in low-latitude and/or oligotrophic regions, such as the NASTG and the SPG [[9]](https://doi.org/10.1038/s41586-023-06321-z) [[10]](https://doi.org/10.5194/bg-7-621-2010), it is not yet possible to unequivocally associate the trends calculated here with climate change.\nAn additional analysis, performed using the same notebook, but excluding SeaWiFS data prior to March 2002, indicates that they do not significantly modify the long-term trends computed for the variables analysed here nor their statistics in all selected regions.\n\n## ℹ️ If you want to know more\n\n### Key resources\n* CDS catalogue entry used in this notebook is the [Ocean colour daily data from 1997 to present derived from satellite observations](https://cds.climate.copernicus.eu/datasets/satellite-ocean-colour?tab=overview)\n* Data download is from [CDS](https://cds.climate.copernicus.eu/datasets/satellite-ocean-colour?tab=form)\n* Product Documentation is available on the [CDS](https://cds.climate.copernicus.eu/datasets/satellite-ocean-colour?tab=doc) website\n\nCode libraries used:\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)\n\nFurther readings:\n* Ocean Colour Monitoring Portals: [CMEMS](https://marine.copernicus.eu/), [ESA-GlobColour Project](https://hermes.acri.fr/), [ESA-OC-CCI](https://climate.esa.int/en/projects/ocean-colour/), [ESA-Sentinel 3](https://sentiwiki.copernicus.eu/web/s3-mission), [IMOS](https://imos.org.au/facility/satellite-remote-sensing/ocean-colour), [NASA](https://oceancolor.gsfc.nasa.gov/#), [NOAA CoastWatch](https://coastwatch.noaa.gov/cwn/index.html)\n* Use cases: [EUMETSAT](https://user.eumetsat.int/search-view?sort=startDate%20desc&facets=%7B%22theme%22:%5B%22Marine%7COcean%20biogeochemistry%22%5D,%22contentTypes%22:%5B%22Resources%7CCase%20studies%22%5D%7D), [C3S](https://climate.copernicus.eu/ocean)\n* Groom, S., et al. (2019). [Satellite Ocean Colour: Current Status and Future Perspective](https://doi.org/10.3389/fmars.2019.00485). Frontiers in Marine Science, 6.\n* [International Ocean Colour Coordinating Group (ICCG)](https://ioccg.org/)\n\n### References\n\n[[1]](https://dast.copernicus-climate.eu/documents/satellite-ocean-colour/v6.0/WP2-FDDP-2022-04_C3S2-Lot3_PUGS-of-v6.0-OceanColour-product_v1.1_FINAL.pdf) Jackson, T., et al. (2023). C3S Ocean Colour Version 6.0: Product User Guide and Specification. Issue 1.1. E.U. Copernicus Climate Change Service. Document ref. WP2-FDDP-2022-04_C3S2-Lot3_PUGS-of-v6.0-OceanColour-product.", "text_with_prefix": "EQC Quality Assessment: \"Impact of using successive satellites on Ocean Colour time series\"\nDataset: satellite-ocean-colour [CDS]\nAspect: consistency-assessment_q01 | Category: Satellite_ECVs\nSection: Impact of using successive satellites on Ocean Colour time series > Quality assessment statement\n---\n. Linear trends calculated here for Rrs at 560 nm are not statistically significant (p>0.05), although an increasing tendency is observed everywhere, except in the SPG where Rrs at 560 nm decreases over time.\n\nRelatively high chlorophyll-a concentrations and low Rrs at 443 nm are observed when SeaWiFS was the only active satellite (i.e., from January 1998 to March 2002) in both the global ocean and the oceanic region between 50°S and 50°N, but not in the NASTG's and SPG's regions. \nOverall, these results suggest that, prior to March 2002, Rrs at 443 would be underestimated, especially when the analysis is extended to include high-latitude, more productive oceanic zones, and trigger chlorophyll-a overestimates.\nDespite the extensive time series analyzed here and the relatively short time series length required to detect climate-driven variations in low-latitude and/or oligotrophic regions, such as the NASTG and the SPG [[9]](https://doi.org/10.1038/s41586-023-06321-z) [[10]](https://doi.org/10.5194/bg-7-621-2010), it is not yet possible to unequivocally associate the trends calculated here with climate change.\nAn additional analysis, performed using the same notebook, but excluding SeaWiFS data prior to March 2002, indicates that they do not significantly modify the long-term trends computed for the variables analysed here nor their statistics in all selected regions.\n\n## ℹ️ If you want to know more\n\n### Key resources\n* CDS catalogue entry used in this notebook is the [Ocean colour daily data from 1997 to present derived from satellite observations](https://cds.climate.copernicus.eu/datasets/satellite-ocean-colour?tab=overview)\n* Data download is from [CDS](https://cds.climate.copernicus.eu/datasets/satellite-ocean-colour?tab=form)\n* Product Documentation is available on the [CDS](https://cds.climate.copernicus.eu/datasets/satellite-ocean-colour?tab=doc) website\n\nCode libraries used:\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)\n\nFurther readings:\n* Ocean Colour Monitoring Portals: [CMEMS](https://marine.copernicus.eu/), [ESA-GlobColour Project](https://hermes.acri.fr/), [ESA-OC-CCI](https://climate.esa.int/en/projects/ocean-colour/), [ESA-Sentinel 3](https://sentiwiki.copernicus.eu/web/s3-mission), [IMOS](https://imos.org.au/facility/satellite-remote-sensing/ocean-colour), [NASA](https://oceancolor.gsfc.nasa.gov/#), [NOAA CoastWatch](https://coastwatch.noaa.gov/cwn/index.html)\n* Use cases: [EUMETSAT](https://user.eumetsat.int/search-view?sort=startDate%20desc&facets=%7B%22theme%22:%5B%22Marine%7COcean%20biogeochemistry%22%5D,%22contentTypes%22:%5B%22Resources%7CCase%20studies%22%5D%7D), [C3S](https://climate.copernicus.eu/ocean)\n* Groom, S., et al. (2019). [Satellite Ocean Colour: Current Status and Future Perspective](https://doi.org/10.3389/fmars.2019.00485). Frontiers in Marine Science, 6.\n* [International Ocean Colour Coordinating Group (ICCG)](https://ioccg.org/)\n\n### References\n\n[[1]](https://dast.copernicus-climate.eu/documents/satellite-ocean-colour/v6.0/WP2-FDDP-2022-04_C3S2-Lot3_PUGS-of-v6.0-OceanColour-product_v1.1_FINAL.pdf) Jackson, T., et al. (2023). C3S Ocean Colour Version 6.0: Product User Guide and Specification. Issue 1.1. E.U. Copernicus Climate Change Service. Document ref. WP2-FDDP-2022-04_C3S2-Lot3_PUGS-of-v6.0-OceanColour-product."} {"chunk_id": "satellite_satellite-ocean-colour_consistency-assessment_q01__2c6dc3852474", "report_id": "satellite_satellite-ocean-colour_consistency-assessment_q01", "dataset_id": "satellite-ocean-colour", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency-assessment_q01", "aspect_base": "consistency-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Impact of using successive satellites on Ocean Colour time series > Quality assessment statement", "title": "Impact of using successive satellites on Ocean Colour time series", "chunk_index": 5, "token_count": 856, "text_raw": "1.1. E.U. Copernicus Climate Change Service. Document ref. WP2-FDDP-2022-04_C3S2-Lot3_PUGS-of-v6.0-OceanColour-product.\n\n[[2]](https://doi.org/10.4319/lo.1997.42.7.1479) Behrenfeld, M.J., & Falkowski, P.G. (1997). A consumer's guide to phytoplankton primary productivity models. Limnology and Oceanography, 42, 1479-1491. 9.\n\n[[3]](https://doi.org/10.1371/journal.pone.0028945) Friedland, K.D., et al. (2012). Pathways between Primary Production and Fisheries Yields of Large Marine Ecosystems. PLoS ONE, 7(1), e28945.\n\n[[4]](https://dast.copernicus-climate.eu/documents/satellite-ocean-colour/v6.0/WP2-FDDP-2022-04_C3S2-Lot3_ATBD-of-v6.0-OceanColour-product_v1.1_FINAL.pdf) Jackson, T., et al. (2022). C3S Ocean Colour Version 6.0: Algorithm Theoretical Basis Document. Issue 1.1. E.U. Copernicus Climate Change Service. Document ref. WP2-FDDP-2022-04_C3S2-Lot3_ATBD-of-v6.0-OceanColour-product.\n\n[[5]](https://doi.org/10.1029/2021GL096965) Leonelli, F.E., et al. (2022). Ultra-oligotrophic waters expansion in the North Atlantic Subtropical Gyre revealed by 21 years of satellite observations. Geophysical Research Letters, 49, e2021GL096965.\n\n[[6]](https://dast.copernicus-climate.eu/documents/satellite-ocean-colour/v6.0/WP2-FDDP-2022-04_C3S2-Lot3_PQAR-of-v6.0-OceanColour-product_v1.1_FINAL.pdf) Jackson, T., et al. (2023) C3S Ocean Colour Version 6.0: Product Quality Assessment Report. Issue 1.1. E.U. Copernicus Climate Change Service. Document ref. WP2-FDDP-2022-04_C3S2-Lot3_ PQAR-of-v6.0-OceanColour-product.\n\n[[7]](https://doi.org/10.1029/95JC00458) Campbell, J.W. (1995). The lognormal distribution as a model for bio-optical variability in the sea. Journal of Geophysical Research, 100, 13237-13254.\n\n[[8]](https://doi.org/10.3390/s19194285) Sathyendranath, S., et al. (2019). An Ocean-Colour time series for use in climate studies: the experience of the Ocean-Colour Climate Change Initiative (OC-CCI). Sensors, 19, 4285.\n\n[[9]](https://doi.org/10.1038/s41586-023-06321-z) Cael, B.B., et al. (2023). Global climate-change trends detected in indicators of ocean ecology. Nature, 619, 551-554.\n\n[[10]](https://doi.org/10.5194/bg-7-621-2010) Henson, S.A., et al. (2010). Detection of anthropogenic climate change in satellite records of ocean chlorophyll and productivity. Biogeosciences, 7, 621-640.", "text_with_prefix": "EQC Quality Assessment: \"Impact of using successive satellites on Ocean Colour time series\"\nDataset: satellite-ocean-colour [CDS]\nAspect: consistency-assessment_q01 | Category: Satellite_ECVs\nSection: Impact of using successive satellites on Ocean Colour time series > Quality assessment statement\n---\n1.1. E.U. Copernicus Climate Change Service. Document ref. WP2-FDDP-2022-04_C3S2-Lot3_PUGS-of-v6.0-OceanColour-product.\n\n[[2]](https://doi.org/10.4319/lo.1997.42.7.1479) Behrenfeld, M.J., & Falkowski, P.G. (1997). A consumer's guide to phytoplankton primary productivity models. Limnology and Oceanography, 42, 1479-1491. 9.\n\n[[3]](https://doi.org/10.1371/journal.pone.0028945) Friedland, K.D., et al. (2012). Pathways between Primary Production and Fisheries Yields of Large Marine Ecosystems. PLoS ONE, 7(1), e28945.\n\n[[4]](https://dast.copernicus-climate.eu/documents/satellite-ocean-colour/v6.0/WP2-FDDP-2022-04_C3S2-Lot3_ATBD-of-v6.0-OceanColour-product_v1.1_FINAL.pdf) Jackson, T., et al. (2022). C3S Ocean Colour Version 6.0: Algorithm Theoretical Basis Document. Issue 1.1. E.U. Copernicus Climate Change Service. Document ref. WP2-FDDP-2022-04_C3S2-Lot3_ATBD-of-v6.0-OceanColour-product.\n\n[[5]](https://doi.org/10.1029/2021GL096965) Leonelli, F.E., et al. (2022). Ultra-oligotrophic waters expansion in the North Atlantic Subtropical Gyre revealed by 21 years of satellite observations. Geophysical Research Letters, 49, e2021GL096965.\n\n[[6]](https://dast.copernicus-climate.eu/documents/satellite-ocean-colour/v6.0/WP2-FDDP-2022-04_C3S2-Lot3_PQAR-of-v6.0-OceanColour-product_v1.1_FINAL.pdf) Jackson, T., et al. (2023) C3S Ocean Colour Version 6.0: Product Quality Assessment Report. Issue 1.1. E.U. Copernicus Climate Change Service. Document ref. WP2-FDDP-2022-04_C3S2-Lot3_ PQAR-of-v6.0-OceanColour-product.\n\n[[7]](https://doi.org/10.1029/95JC00458) Campbell, J.W. (1995). The lognormal distribution as a model for bio-optical variability in the sea. Journal of Geophysical Research, 100, 13237-13254.\n\n[[8]](https://doi.org/10.3390/s19194285) Sathyendranath, S., et al. (2019). An Ocean-Colour time series for use in climate studies: the experience of the Ocean-Colour Climate Change Initiative (OC-CCI). Sensors, 19, 4285.\n\n[[9]](https://doi.org/10.1038/s41586-023-06321-z) Cael, B.B., et al. (2023). Global climate-change trends detected in indicators of ocean ecology. Nature, 619, 551-554.\n\n[[10]](https://doi.org/10.5194/bg-7-621-2010) Henson, S.A., et al. (2010). Detection of anthropogenic climate change in satellite records of ocean chlorophyll and productivity. Biogeosciences, 7, 621-640."} {"chunk_id": "satellite_satellite-sea-ice-edge-type_intercomparison_q01__1eeed55192c2", "report_id": "satellite_satellite-sea-ice-edge-type_intercomparison_q01", "dataset_id": "satellite-sea-ice-edge-type", "store": "CDS", "doc_type": "EQC_QA", "aspect": "intercomparison_q01", "aspect_base": "intercomparison", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Intercomparison of satellite multi-year sea ice extent estimates", "title": "Intercomparison of satellite multi-year sea ice extent estimates", "chunk_index": 0, "token_count": 110, "text_raw": "Production date: 19-05-2025\n\nProduced by: Timothy Williams (Nansen Environmental and Remote Sensing Center; NERSC) and Fabio Mangini (Nansen Environmental and Remote Sensing Center; NERSC)", "text_with_prefix": "EQC Quality Assessment: \"Intercomparison of satellite multi-year sea ice extent estimates\"\nDataset: satellite-sea-ice-edge-type [CDS]\nAspect: intercomparison_q01 | Category: Satellite_ECVs\nSection: Intercomparison of satellite multi-year sea ice extent estimates\n---\nProduction date: 19-05-2025\n\nProduced by: Timothy Williams (Nansen Environmental and Remote Sensing Center; NERSC) and Fabio Mangini (Nansen Environmental and Remote Sensing Center; NERSC)"} {"chunk_id": "satellite_satellite-sea-ice-edge-type_intercomparison_q01__31aa09584834", "report_id": "satellite_satellite-sea-ice-edge-type_intercomparison_q01", "dataset_id": "satellite-sea-ice-edge-type", "store": "CDS", "doc_type": "EQC_QA", "aspect": "intercomparison_q01", "aspect_base": "intercomparison", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Intercomparison of satellite multi-year sea ice extent estimates > Quality assessment question", "title": "Intercomparison of satellite multi-year sea ice extent estimates", "chunk_index": 1, "token_count": 428, "text_raw": "* **How consistent is the evolution of multi-year sea-ice extent in the Arctic across different satellite sea-ice products?**\n\nThis study examines the geographical distribution of multi-year sea ice (MYI) in the Arctic, providing an analysis of its evolution since the early 1990s. The assessment focuses on multi-year sea ice, which is sea ice that has survived at least one melting season, because its extent in the Arctic has significantly declined over the last few decades (e.g., [[1]](https://doi.org/10.1029/2011GL047735)). Changes in MYI extent are of interest because of the repercussions they have on the Earth's climate which, in turn, can impact the Arctic ecosystem and affect human activities in the region (e.g., [[2]](https://www.cambridge.org/core/books/ocean-and-cryosphere-in-a-changing-climate/polar-regions/8D76B8865B796C16991F7A9FB6271C2D)).\n\nThe assessment uses two different sea-ice products. This choice helps determine the robustness of the results and highlight the properties of the two datasets. The first dataset, referred to as \"CDS dataset\", provides information on sea-ice types in the region, while the second, the \"NERSC dataset\", classifies sea ice into age categories.\n\nBoth datasets indicate a general decrease in Arctic multi-year sea ice extent over the last three decades, especially in the Beaufort Sea, the Chukchi Sea, and the Eurasian Basin. However, they show several differences on a regional scale. For example, compared to the CDS dataset, the NERSC dataset reports a consistent higher presence of multi-year sea ice in the Greenland Sea away from the coast.", "text_with_prefix": "EQC Quality Assessment: \"Intercomparison of satellite multi-year sea ice extent estimates\"\nDataset: satellite-sea-ice-edge-type [CDS]\nAspect: intercomparison_q01 | Category: Satellite_ECVs\nSection: Intercomparison of satellite multi-year sea ice extent estimates > Quality assessment question\n---\n* **How consistent is the evolution of multi-year sea-ice extent in the Arctic across different satellite sea-ice products?**\n\nThis study examines the geographical distribution of multi-year sea ice (MYI) in the Arctic, providing an analysis of its evolution since the early 1990s. The assessment focuses on multi-year sea ice, which is sea ice that has survived at least one melting season, because its extent in the Arctic has significantly declined over the last few decades (e.g., [[1]](https://doi.org/10.1029/2011GL047735)). Changes in MYI extent are of interest because of the repercussions they have on the Earth's climate which, in turn, can impact the Arctic ecosystem and affect human activities in the region (e.g., [[2]](https://www.cambridge.org/core/books/ocean-and-cryosphere-in-a-changing-climate/polar-regions/8D76B8865B796C16991F7A9FB6271C2D)).\n\nThe assessment uses two different sea-ice products. This choice helps determine the robustness of the results and highlight the properties of the two datasets. The first dataset, referred to as \"CDS dataset\", provides information on sea-ice types in the region, while the second, the \"NERSC dataset\", classifies sea ice into age categories.\n\nBoth datasets indicate a general decrease in Arctic multi-year sea ice extent over the last three decades, especially in the Beaufort Sea, the Chukchi Sea, and the Eurasian Basin. However, they show several differences on a regional scale. For example, compared to the CDS dataset, the NERSC dataset reports a consistent higher presence of multi-year sea ice in the Greenland Sea away from the coast."} {"chunk_id": "satellite_satellite-sea-ice-edge-type_intercomparison_q01__fe04bb7c3ad7", "report_id": "satellite_satellite-sea-ice-edge-type_intercomparison_q01", "dataset_id": "satellite-sea-ice-edge-type", "store": "CDS", "doc_type": "EQC_QA", "aspect": "intercomparison_q01", "aspect_base": "intercomparison", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Intercomparison of satellite multi-year sea ice extent estimates > Quality assessment statement", "title": "Intercomparison of satellite multi-year sea ice extent estimates", "chunk_index": 2, "token_count": 270, "text_raw": "These are the key outcomes of this assessment\n\n* Both the CDS and the NERSC datasets reveal an overall decrease in multi-year sea ice extent over the Arctic since 1991, with the most significant reductions occurring in the Beaufort Sea, the Chukchi Sea, and the Eurasian Basin.\n\n* Despite their consistent conclusions, notable differences exist between the two datasets. Specifically, the CDS dataset tends to overestimate multi-year sea ice presence in the Beaufort Sea and underestimate it in the Greenland Sea when compared to the NERSC dataset. It is shown that these discrepancies tend to be minimum in October and to become more pronounced as each winter progresses.\n\n* The inclusion of \"ambiguous\" data points in the CDS dataset has two main implications on the estimates of multi-year sea ice extent in the Arctic. At first, it affects the total amount of multi-year sea ice in the region, even though it only little alters its geographic distribution. Secondly, it increases its day-to-day variability. \n```", "text_with_prefix": "EQC Quality Assessment: \"Intercomparison of satellite multi-year sea ice extent estimates\"\nDataset: satellite-sea-ice-edge-type [CDS]\nAspect: intercomparison_q01 | Category: Satellite_ECVs\nSection: Intercomparison of satellite multi-year sea ice extent estimates > Quality assessment statement\n---\nThese are the key outcomes of this assessment\n\n* Both the CDS and the NERSC datasets reveal an overall decrease in multi-year sea ice extent over the Arctic since 1991, with the most significant reductions occurring in the Beaufort Sea, the Chukchi Sea, and the Eurasian Basin.\n\n* Despite their consistent conclusions, notable differences exist between the two datasets. Specifically, the CDS dataset tends to overestimate multi-year sea ice presence in the Beaufort Sea and underestimate it in the Greenland Sea when compared to the NERSC dataset. It is shown that these discrepancies tend to be minimum in October and to become more pronounced as each winter progresses.\n\n* The inclusion of \"ambiguous\" data points in the CDS dataset has two main implications on the estimates of multi-year sea ice extent in the Arctic. At first, it affects the total amount of multi-year sea ice in the region, even though it only little alters its geographic distribution. Secondly, it increases its day-to-day variability. \n```"} {"chunk_id": "satellite_satellite-sea-ice-edge-type_intercomparison_q01__65ce0458c03f", "report_id": "satellite_satellite-sea-ice-edge-type_intercomparison_q01", "dataset_id": "satellite-sea-ice-edge-type", "store": "CDS", "doc_type": "EQC_QA", "aspect": "intercomparison_q01", "aspect_base": "intercomparison", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Intercomparison of satellite multi-year sea ice extent estimates > Methodology", "title": "Intercomparison of satellite multi-year sea ice extent estimates", "chunk_index": 3, "token_count": 1030, "text_raw": "The evolution of multi-year sea ice extent in the Arctic is based on two different datasets. This choice helps show the properties of the two products as well as assessing the robustness of the results. The first dataset is freely available on the Copernicus Climate Data Store (CDS) website under the name [\"Sea ice edge and type daily gridded data from 1978 to present derived from satellite observations\"](https://cds.climate.copernicus.eu/datasets/satellite-sea-ice-edge-type?tab=overview). The data provides information on sea-ice types in the Arctic Ocean by classifying sea ice into four categories: \"open water\", \"first-year ice\", \"multi-year ice\", and \"ambiguous\". The classification primarily relies on brightness temperature observed by passive microwave radiometers. For the purposes of this Notebook, only the subset of the dataset labelled as Climate Data Records (CDRs) is considered, thereby excluding the Interim Climate Data Records (ICDRs). The ICDRs have been excluded from this work because their consistency has not been as thoroughly verified as that of the CDRs.\n\nThe second dataset is described by [[3]](https://tc.copernicus.org/articles/12/2073/2018/) and is provided by Dr. Anton Korosov (anton.korosov@nersc.no) on request. This dataset differs from the previous one as it uses an Eulerian advection scheme, together with passive-microwave-derived concentration and drift estimations, to classify sea ice into age categories, instead of ice types. Specifically, the dataset provides an estimate of sea-ice age over the Arctic, and an estimate of individual ice-age fractions at each grid point. For convenience, the former dataset will be referred to as \"CDS dataset\" whereas the latter as \"NERSC dataset\".\n\nThe assessment uses two distinct approaches to identify the grid points characterized by multi-year sea ice in the CDS and in the NERSC datasets. In the CDS dataset, the presence of multi-year sea ice is determined using two different criteria: first, by selecting the pixels classified as \"multi-year ice\", and second, by considering the pixels classified as both \"multi-year ice\" and \"ambiguous\". The results obtained with these two criteria are then compared to highlight the role of the pixels classified as ambiguous. Instead, in the NERSC dataset, multi-year sea ice is identified by selecting the pixels where the concentration of first-year sea ice (FYI) does not exceed that of sea ice that is older than one year, and where the sum of the two exceeds 30% (where there is some ice present). Note that this threshold, which is usually 15%, is chosen to be consistent with the CDS product.\n\nThe analysis is performed on the overlapping interval of the two datasets during the boreal winters between October 1991 and April 2020, where with boreal winter we refer to the period ranging from the 1st of October of one year to the 30th of April of the following year. The upper limit is set by the CDS dataset, which provides information on sea-ice types for the boreal winter months between October 1978 and December 2020, whereas the lower limit is set by the NERSC dataset, which covers the period from September 1991 to May 2023.\n\nThe \"Analysis and results\" section is structured as follows:\n\n**[](section-1)**\n\n**[](section-2)**\n\n**[](section-3)**\n\n     **[](section-3.1)**\n\n     We present and compare the temporal evolution of the area covered by multi-year sea ice over the Arctic Basin as provided by the CDS and the NERSC datasets. This section also quantifies the mismatch between the two datasets by analysing the bias and the integrated ice type error (IITE) of the NERSC dataset relative to CDS dataset. The bias is the difference between the area where the CDS dataset reports multi-year sea ice while the NERSC dataset does not, and the area where the CDS dataset does not report multi-year sea ice while the NERSC dataset does. Instead, the integrated ice type error is the sum of these two areas.\n\n     **[](section-3.2)**\n\n     We investigate further the temporal evolution of multi-year sea ice in the Arctic with the help of maps. Specifically, we present and compare, for each dataset, month and pixel, the percentage of days associated with multi-year sea ice over three decades.", "text_with_prefix": "EQC Quality Assessment: \"Intercomparison of satellite multi-year sea ice extent estimates\"\nDataset: satellite-sea-ice-edge-type [CDS]\nAspect: intercomparison_q01 | Category: Satellite_ECVs\nSection: Intercomparison of satellite multi-year sea ice extent estimates > Methodology\n---\nThe evolution of multi-year sea ice extent in the Arctic is based on two different datasets. This choice helps show the properties of the two products as well as assessing the robustness of the results. The first dataset is freely available on the Copernicus Climate Data Store (CDS) website under the name [\"Sea ice edge and type daily gridded data from 1978 to present derived from satellite observations\"](https://cds.climate.copernicus.eu/datasets/satellite-sea-ice-edge-type?tab=overview). The data provides information on sea-ice types in the Arctic Ocean by classifying sea ice into four categories: \"open water\", \"first-year ice\", \"multi-year ice\", and \"ambiguous\". The classification primarily relies on brightness temperature observed by passive microwave radiometers. For the purposes of this Notebook, only the subset of the dataset labelled as Climate Data Records (CDRs) is considered, thereby excluding the Interim Climate Data Records (ICDRs). The ICDRs have been excluded from this work because their consistency has not been as thoroughly verified as that of the CDRs.\n\nThe second dataset is described by [[3]](https://tc.copernicus.org/articles/12/2073/2018/) and is provided by Dr. Anton Korosov (anton.korosov@nersc.no) on request. This dataset differs from the previous one as it uses an Eulerian advection scheme, together with passive-microwave-derived concentration and drift estimations, to classify sea ice into age categories, instead of ice types. Specifically, the dataset provides an estimate of sea-ice age over the Arctic, and an estimate of individual ice-age fractions at each grid point. For convenience, the former dataset will be referred to as \"CDS dataset\" whereas the latter as \"NERSC dataset\".\n\nThe assessment uses two distinct approaches to identify the grid points characterized by multi-year sea ice in the CDS and in the NERSC datasets. In the CDS dataset, the presence of multi-year sea ice is determined using two different criteria: first, by selecting the pixels classified as \"multi-year ice\", and second, by considering the pixels classified as both \"multi-year ice\" and \"ambiguous\". The results obtained with these two criteria are then compared to highlight the role of the pixels classified as ambiguous. Instead, in the NERSC dataset, multi-year sea ice is identified by selecting the pixels where the concentration of first-year sea ice (FYI) does not exceed that of sea ice that is older than one year, and where the sum of the two exceeds 30% (where there is some ice present). Note that this threshold, which is usually 15%, is chosen to be consistent with the CDS product.\n\nThe analysis is performed on the overlapping interval of the two datasets during the boreal winters between October 1991 and April 2020, where with boreal winter we refer to the period ranging from the 1st of October of one year to the 30th of April of the following year. The upper limit is set by the CDS dataset, which provides information on sea-ice types for the boreal winter months between October 1978 and December 2020, whereas the lower limit is set by the NERSC dataset, which covers the period from September 1991 to May 2023.\n\nThe \"Analysis and results\" section is structured as follows:\n\n**[](section-1)**\n\n**[](section-2)**\n\n**[](section-3)**\n\n     **[](section-3.1)**\n\n     We present and compare the temporal evolution of the area covered by multi-year sea ice over the Arctic Basin as provided by the CDS and the NERSC datasets. This section also quantifies the mismatch between the two datasets by analysing the bias and the integrated ice type error (IITE) of the NERSC dataset relative to CDS dataset. The bias is the difference between the area where the CDS dataset reports multi-year sea ice while the NERSC dataset does not, and the area where the CDS dataset does not report multi-year sea ice while the NERSC dataset does. Instead, the integrated ice type error is the sum of these two areas.\n\n     **[](section-3.2)**\n\n     We investigate further the temporal evolution of multi-year sea ice in the Arctic with the help of maps. Specifically, we present and compare, for each dataset, month and pixel, the percentage of days associated with multi-year sea ice over three decades."} {"chunk_id": "satellite_satellite-sea-ice-edge-type_intercomparison_q01__377cc8b1e56f", "report_id": "satellite_satellite-sea-ice-edge-type_intercomparison_q01", "dataset_id": "satellite-sea-ice-edge-type", "store": "CDS", "doc_type": "EQC_QA", "aspect": "intercomparison_q01", "aspect_base": "intercomparison", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Intercomparison of satellite multi-year sea ice extent estimates > Analysis and results > 1. Parameters, requests and functions definition", "title": "Intercomparison of satellite multi-year sea ice extent estimates", "chunk_index": 4, "token_count": 117, "text_raw": "- Define the parameters and formulate the requests for downloading with the EQC toolbox.\n- Define the functions to process and reduce the size of the downloaded data.\n- Define the functions to post-process and visualize the data.", "text_with_prefix": "EQC Quality Assessment: \"Intercomparison of satellite multi-year sea ice extent estimates\"\nDataset: satellite-sea-ice-edge-type [CDS]\nAspect: intercomparison_q01 | Category: Satellite_ECVs\nSection: Intercomparison of satellite multi-year sea ice extent estimates > Analysis and results > 1. Parameters, requests and functions definition\n---\n- Define the parameters and formulate the requests for downloading with the EQC toolbox.\n- Define the functions to process and reduce the size of the downloaded data.\n- Define the functions to post-process and visualize the data."} {"chunk_id": "satellite_satellite-sea-ice-edge-type_intercomparison_q01__f47412af8ed6", "report_id": "satellite_satellite-sea-ice-edge-type_intercomparison_q01", "dataset_id": "satellite-sea-ice-edge-type", "store": "CDS", "doc_type": "EQC_QA", "aspect": "intercomparison_q01", "aspect_base": "intercomparison", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Intercomparison of satellite multi-year sea ice extent estimates > Analysis and results > 1. Parameters, requests and functions definition > 1.2 Set Parameters", "title": "Intercomparison of satellite multi-year sea ice extent estimates", "chunk_index": 5, "token_count": 180, "text_raw": "- set the years to use for the climatologies with `periods`. These years are also used to define the length of the time series.\n- set `conc_threshold`, the concentration threshold to determine the presence of sea ice in the NERSC product.\n- set the path to the NERSC ice age product with `NERSC_path`.\n\nDefine time periods\nConcentration threshold to determine the presence of sea ice in the NERSC product\nDefine path to NERSC data", "text_with_prefix": "EQC Quality Assessment: \"Intercomparison of satellite multi-year sea ice extent estimates\"\nDataset: satellite-sea-ice-edge-type [CDS]\nAspect: intercomparison_q01 | Category: Satellite_ECVs\nSection: Intercomparison of satellite multi-year sea ice extent estimates > Analysis and results > 1. Parameters, requests and functions definition > 1.2 Set Parameters\n---\n- set the years to use for the climatologies with `periods`. These years are also used to define the length of the time series.\n- set `conc_threshold`, the concentration threshold to determine the presence of sea ice in the NERSC product.\n- set the path to the NERSC ice age product with `NERSC_path`.\n\nDefine time periods\nConcentration threshold to determine the presence of sea ice in the NERSC product\nDefine path to NERSC data"} {"chunk_id": "satellite_satellite-sea-ice-edge-type_intercomparison_q01__04bba9f662af", "report_id": "satellite_satellite-sea-ice-edge-type_intercomparison_q01", "dataset_id": "satellite-sea-ice-edge-type", "store": "CDS", "doc_type": "EQC_QA", "aspect": "intercomparison_q01", "aspect_base": "intercomparison", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Intercomparison of satellite multi-year sea ice extent estimates > Analysis and results > 1. Parameters, requests and functions definition > 1.4 Functions to load and classify NERSC data", "title": "Intercomparison of satellite multi-year sea ice extent estimates", "chunk_index": 6, "token_count": 140, "text_raw": "- `get_nersc_data` loads the NERSC data for a period corresponding to a set CDS data.\n- `get_nersc_multiyear_ice` processes the NERSC data to produce a mask of multi-year sea ice.", "text_with_prefix": "EQC Quality Assessment: \"Intercomparison of satellite multi-year sea ice extent estimates\"\nDataset: satellite-sea-ice-edge-type [CDS]\nAspect: intercomparison_q01 | Category: Satellite_ECVs\nSection: Intercomparison of satellite multi-year sea ice extent estimates > Analysis and results > 1. Parameters, requests and functions definition > 1.4 Functions to load and classify NERSC data\n---\n- `get_nersc_data` loads the NERSC data for a period corresponding to a set CDS data.\n- `get_nersc_multiyear_ice` processes the NERSC data to produce a mask of multi-year sea ice."} {"chunk_id": "satellite_satellite-sea-ice-edge-type_intercomparison_q01__bcbb57ea18e6", "report_id": "satellite_satellite-sea-ice-edge-type_intercomparison_q01", "dataset_id": "satellite-sea-ice-edge-type", "store": "CDS", "doc_type": "EQC_QA", "aspect": "intercomparison_q01", "aspect_base": "intercomparison", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Intercomparison of satellite multi-year sea ice extent estimates > Analysis and results > 1. Parameters, requests and functions definition > 1.5 Functions to produce time series", "title": "Intercomparison of satellite multi-year sea ice extent estimates", "chunk_index": 7, "token_count": 245, "text_raw": "- `get_classification_mask` processes the CDS data to produce a mask of multi-year sea ice.\n- `mask_outside_central_arctic` defines a mask based on a manually defined set of maximum latitudes. This mask is `False` if a pixel should be included in the calculation of MYI extent and error statistics and `True` if it shouldn't.\n- `compute_spatial_sum` computes the area corresponding to a given boolean array on an equal-area grid.\n- `compute_sea_ice_evaluation_diagnostics` computes the multi-year sea ice extent for the NERSC and the CDS products, the bias and the integrated ice type error between them.\n\ngrid cell area of sea ice edge grid\nMask for inside central Arctic\nMasks for MYI\nFill variables", "text_with_prefix": "EQC Quality Assessment: \"Intercomparison of satellite multi-year sea ice extent estimates\"\nDataset: satellite-sea-ice-edge-type [CDS]\nAspect: intercomparison_q01 | Category: Satellite_ECVs\nSection: Intercomparison of satellite multi-year sea ice extent estimates > Analysis and results > 1. Parameters, requests and functions definition > 1.5 Functions to produce time series\n---\n- `get_classification_mask` processes the CDS data to produce a mask of multi-year sea ice.\n- `mask_outside_central_arctic` defines a mask based on a manually defined set of maximum latitudes. This mask is `False` if a pixel should be included in the calculation of MYI extent and error statistics and `True` if it shouldn't.\n- `compute_spatial_sum` computes the area corresponding to a given boolean array on an equal-area grid.\n- `compute_sea_ice_evaluation_diagnostics` computes the multi-year sea ice extent for the NERSC and the CDS products, the bias and the integrated ice type error between them.\n\ngrid cell area of sea ice edge grid\nMask for inside central Arctic\nMasks for MYI\nFill variables"} {"chunk_id": "satellite_satellite-sea-ice-edge-type_intercomparison_q01__7410e0fc3798", "report_id": "satellite_satellite-sea-ice-edge-type_intercomparison_q01", "dataset_id": "satellite-sea-ice-edge-type", "store": "CDS", "doc_type": "EQC_QA", "aspect": "intercomparison_q01", "aspect_base": "intercomparison", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Intercomparison of satellite multi-year sea ice extent estimates > Analysis and results > 1. Parameters, requests and functions definition > 1.6 Functions to post-process and plot time series", "title": "Intercomparison of satellite multi-year sea ice extent estimates", "chunk_index": 8, "token_count": 175, "text_raw": "- `rearrange_year_vs_monthday` groups a long time series by year and day of year.\n- `split_dataset_ambiguous` splits a dataset into one dataset that depends on whether the ambiguous pixels in the CDS dataset are used, and another dataset that does not depend on this.\n- `plot_against_monthday` plots the timeseries against the day-of-year by year, with a different colour for each year.", "text_with_prefix": "EQC Quality Assessment: \"Intercomparison of satellite multi-year sea ice extent estimates\"\nDataset: satellite-sea-ice-edge-type [CDS]\nAspect: intercomparison_q01 | Category: Satellite_ECVs\nSection: Intercomparison of satellite multi-year sea ice extent estimates > Analysis and results > 1. Parameters, requests and functions definition > 1.6 Functions to post-process and plot time series\n---\n- `rearrange_year_vs_monthday` groups a long time series by year and day of year.\n- `split_dataset_ambiguous` splits a dataset into one dataset that depends on whether the ambiguous pixels in the CDS dataset are used, and another dataset that does not depend on this.\n- `plot_against_monthday` plots the timeseries against the day-of-year by year, with a different colour for each year."} {"chunk_id": "satellite_satellite-sea-ice-edge-type_intercomparison_q01__41bd800f5ae6", "report_id": "satellite_satellite-sea-ice-edge-type_intercomparison_q01", "dataset_id": "satellite-sea-ice-edge-type", "store": "CDS", "doc_type": "EQC_QA", "aspect": "intercomparison_q01", "aspect_base": "intercomparison", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Intercomparison of satellite multi-year sea ice extent estimates > Analysis and results > 1. Parameters, requests and functions definition > 1.7 Functions to produce climatologies", "title": "Intercomparison of satellite multi-year sea ice extent estimates", "chunk_index": 9, "token_count": 175, "text_raw": "- `compute_multiyear_ice_percentage` groups daily MYI masks by month and calculates the proportion of each month that a pixel contains MYI.\n- `compute_cds_multiyear_ice_percentage` is a wrapper to call `compute_multiyear_ice_percentage` for CDS data.\n- `compute_nersc_multiyear_ice_percentage` is a wrapper to call `compute_multiyear_ice_percentage` for NERSC data.", "text_with_prefix": "EQC Quality Assessment: \"Intercomparison of satellite multi-year sea ice extent estimates\"\nDataset: satellite-sea-ice-edge-type [CDS]\nAspect: intercomparison_q01 | Category: Satellite_ECVs\nSection: Intercomparison of satellite multi-year sea ice extent estimates > Analysis and results > 1. Parameters, requests and functions definition > 1.7 Functions to produce climatologies\n---\n- `compute_multiyear_ice_percentage` groups daily MYI masks by month and calculates the proportion of each month that a pixel contains MYI.\n- `compute_cds_multiyear_ice_percentage` is a wrapper to call `compute_multiyear_ice_percentage` for CDS data.\n- `compute_nersc_multiyear_ice_percentage` is a wrapper to call `compute_multiyear_ice_percentage` for NERSC data."} {"chunk_id": "satellite_satellite-sea-ice-edge-type_intercomparison_q01__001c96047f92", "report_id": "satellite_satellite-sea-ice-edge-type_intercomparison_q01", "dataset_id": "satellite-sea-ice-edge-type", "store": "CDS", "doc_type": "EQC_QA", "aspect": "intercomparison_q01", "aspect_base": "intercomparison", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Intercomparison of satellite multi-year sea ice extent estimates > Analysis and results > 1. Parameters, requests and functions definition > 1.8 Functions to plot climatologies", "title": "Intercomparison of satellite multi-year sea ice extent estimates", "chunk_index": 10, "token_count": 194, "text_raw": "- `plot_maps` plots multiple maps of either MYI percentage or the difference between the CDS and NERSC products.\n- `plot_cds_maps` is a wrapper that calls `plot_maps` for CDS data.\n- `plot_masked_data_cds` plots an example map to show the masked region used to calculate the time series.\n- `plot_bias_maps` plots multiple maps of the difference between the CDS and NERSC products.\n\nselect a map to plot to illustrate the masked area\n\n(section-2)=", "text_with_prefix": "EQC Quality Assessment: \"Intercomparison of satellite multi-year sea ice extent estimates\"\nDataset: satellite-sea-ice-edge-type [CDS]\nAspect: intercomparison_q01 | Category: Satellite_ECVs\nSection: Intercomparison of satellite multi-year sea ice extent estimates > Analysis and results > 1. Parameters, requests and functions definition > 1.8 Functions to plot climatologies\n---\n- `plot_maps` plots multiple maps of either MYI percentage or the difference between the CDS and NERSC products.\n- `plot_cds_maps` is a wrapper that calls `plot_maps` for CDS data.\n- `plot_masked_data_cds` plots an example map to show the masked region used to calculate the time series.\n- `plot_bias_maps` plots multiple maps of the difference between the CDS and NERSC products.\n\nselect a map to plot to illustrate the masked area\n\n(section-2)="} {"chunk_id": "satellite_satellite-sea-ice-edge-type_intercomparison_q01__fe06149a08cd", "report_id": "satellite_satellite-sea-ice-edge-type_intercomparison_q01", "dataset_id": "satellite-sea-ice-edge-type", "store": "CDS", "doc_type": "EQC_QA", "aspect": "intercomparison_q01", "aspect_base": "intercomparison", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Intercomparison of satellite multi-year sea ice extent estimates > Analysis and results > 2. Download and transform data", "title": "Intercomparison of satellite multi-year sea ice extent estimates", "chunk_index": 11, "token_count": 116, "text_raw": "Here we download the sea ice type data and transform it to create the time series and climatology maps. The NERSC data should already be available locally.\n\nBias compared to NERSC product\n\n(section-3)=", "text_with_prefix": "EQC Quality Assessment: \"Intercomparison of satellite multi-year sea ice extent estimates\"\nDataset: satellite-sea-ice-edge-type [CDS]\nAspect: intercomparison_q01 | Category: Satellite_ECVs\nSection: Intercomparison of satellite multi-year sea ice extent estimates > Analysis and results > 2. Download and transform data\n---\nHere we download the sea ice type data and transform it to create the time series and climatology maps. The NERSC data should already be available locally.\n\nBias compared to NERSC product\n\n(section-3)="} {"chunk_id": "satellite_satellite-sea-ice-edge-type_intercomparison_q01__638d72bc4629", "report_id": "satellite_satellite-sea-ice-edge-type_intercomparison_q01", "dataset_id": "satellite-sea-ice-edge-type", "store": "CDS", "doc_type": "EQC_QA", "aspect": "intercomparison_q01", "aspect_base": "intercomparison", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Intercomparison of satellite multi-year sea ice extent estimates > Analysis and results > 3. Results", "title": "Intercomparison of satellite multi-year sea ice extent estimates", "chunk_index": 12, "token_count": 143, "text_raw": "The temporal evolution of multi-year sea ice extent over the Arctic is analysed using a combination of plots and maps. The plots help quantify and compare the temporal evolution of the Arctic multi-year sea ice extent provided by the two datasets, while the maps help explain the differences between the two datasets by identifying the regions where the two products differ the most.\n\n(section-3.1)=", "text_with_prefix": "EQC Quality Assessment: \"Intercomparison of satellite multi-year sea ice extent estimates\"\nDataset: satellite-sea-ice-edge-type [CDS]\nAspect: intercomparison_q01 | Category: Satellite_ECVs\nSection: Intercomparison of satellite multi-year sea ice extent estimates > Analysis and results > 3. Results\n---\nThe temporal evolution of multi-year sea ice extent over the Arctic is analysed using a combination of plots and maps. The plots help quantify and compare the temporal evolution of the Arctic multi-year sea ice extent provided by the two datasets, while the maps help explain the differences between the two datasets by identifying the regions where the two products differ the most.\n\n(section-3.1)="} {"chunk_id": "satellite_satellite-sea-ice-edge-type_intercomparison_q01__b202c3a53b9d", "report_id": "satellite_satellite-sea-ice-edge-type_intercomparison_q01", "dataset_id": "satellite-sea-ice-edge-type", "store": "CDS", "doc_type": "EQC_QA", "aspect": "intercomparison_q01", "aspect_base": "intercomparison", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Intercomparison of satellite multi-year sea ice extent estimates > Analysis and results > 3. Results > 3.1 Temporal evolution of the total area covered by multi-year sea ice in the Arctic Basin", "title": "Intercomparison of satellite multi-year sea ice extent estimates", "chunk_index": 13, "token_count": 986, "text_raw": "The map below shows the area over which the Arctic MYI extent is calculated to produce the plots. We should note that the MYI extent is computed over the Arctic Basin, therefore excluding the narrow straits and the Greenland Sea. This choice was made to ensure that the NERSC dataset was able to permit MYI in the selected region. For example, it may not be capable of advecting multi-year sea ice through narrow straits, such as the Nares Strait west of Greenland and the northern Canadian Archipelago.\n\nThe figure below consists of two plots. The plot on the left displays, for each boreal winter between October 1991 and April 2020, the daily area where the sea-ice type from the CDS dataset is set to ambiguous. The plot on the right follows the same structure, but shows the daily MYI extent as provided by the NERSC dataset.\n\nThe ambiguous MYI extent increases as each boreal winter progresses, with the rise accelerating in spring. This results from the CDS product performing better during the cold and stable mid-winter months than in the warmer spring months, when melting and other factors disrupt the distinction between classes. In addition, the CDS product uses drift tracking from the summer minimum to confirm the presence of MYI, and the error in this drift would also accumulate over the winter.\n\nInterestingly, the ambiguous MYI extent does not experience a long-term change, despite the MYI extent decreasing over the period considered, as shown by the figure to the right and the ones below (introduced and discussed later).\n\nThe figure below consists of six plots, arranged in a grid of three rows and two columns. The first row displays the daily MYI extent for each boreal winter between October 1991 and April 2020, estimated from the CDS dataset. The second row shows the difference (\"bias\") between the CDS and the NERSC datasets, while the third row shows the corresponding integrated ice type error. The columns distinguish between the two definitions of multi-year sea ice for the CDS dataset. The right column shows the results based solely on the grid points labelled as \"multi-year ice\", whereas the left column shows results based on the grid points labelled as either \"multi-year ice\" or \"ambiguous\".\n\nThe comparison between the MYI extent from the CDS and the NERSC datasets yields comparable results, and the inclusion of the grid points labelled as \"ambiguous\" does not significantly alter the main conclusions. The results indicate that the spatial extent of multi-year sea ice in the Arctic has decreased since October 1991. Specifically, the October MYI extent dropped by approximately 1-2 million km$^2$ between the 5-year periods 1991-1995 and 2015-2019. A comparable result is found for the MYI extent in April. Selecting the ambiguous points does not significantly alter the results.\n\nBoth datasets also suggest that the MYI extent tends to decrease throughout each boreal winter. This is expected because MYI tends to be both exported and compressed through ridging as winter progresses (the reader is referred to [[4]](https://tc.copernicus.org/articles/17/1873/2023/) for a comprehensive explanation of sources and sinks of multi-year sea ice).\n\nDespite their similarities, the datasets present some significant differences. Notably, the CDS dataset exhibits a slightly greater day-to-day variability compared to the NERSC dataset when the pixels labelled as \"ambiguous\" are included into the analysis. A thorough understanding of this discrepancy goes beyond the scope of this study. However, we hypothesize it to be partially attributed to atmospheric variability, which can affect the brightness temperature of sea ice and, therefore, the classification into first-year sea ice and multi-year sea ice. For example, warmer weather events pose a challenge to the classification of sea ice into ice types as they modify the surface of multi-year sea ice, making it appear similar to first-year sea ice (Product User Guide and Specification of the CDS dataset). This interpretation is corroborated by the day-to-day variability in the CDS dataset decreasing when the grid points labelled as \"ambiguous\" are not considered. The NERSC dataset on the other hand is based on relatively smooth, low-resolution sea ice concentration and drift products derived from lower-frequency (SSMIS) passive microwave measurements which are not as affected by atmospheric variability.", "text_with_prefix": "EQC Quality Assessment: \"Intercomparison of satellite multi-year sea ice extent estimates\"\nDataset: satellite-sea-ice-edge-type [CDS]\nAspect: intercomparison_q01 | Category: Satellite_ECVs\nSection: Intercomparison of satellite multi-year sea ice extent estimates > Analysis and results > 3. Results > 3.1 Temporal evolution of the total area covered by multi-year sea ice in the Arctic Basin\n---\nThe map below shows the area over which the Arctic MYI extent is calculated to produce the plots. We should note that the MYI extent is computed over the Arctic Basin, therefore excluding the narrow straits and the Greenland Sea. This choice was made to ensure that the NERSC dataset was able to permit MYI in the selected region. For example, it may not be capable of advecting multi-year sea ice through narrow straits, such as the Nares Strait west of Greenland and the northern Canadian Archipelago.\n\nThe figure below consists of two plots. The plot on the left displays, for each boreal winter between October 1991 and April 2020, the daily area where the sea-ice type from the CDS dataset is set to ambiguous. The plot on the right follows the same structure, but shows the daily MYI extent as provided by the NERSC dataset.\n\nThe ambiguous MYI extent increases as each boreal winter progresses, with the rise accelerating in spring. This results from the CDS product performing better during the cold and stable mid-winter months than in the warmer spring months, when melting and other factors disrupt the distinction between classes. In addition, the CDS product uses drift tracking from the summer minimum to confirm the presence of MYI, and the error in this drift would also accumulate over the winter.\n\nInterestingly, the ambiguous MYI extent does not experience a long-term change, despite the MYI extent decreasing over the period considered, as shown by the figure to the right and the ones below (introduced and discussed later).\n\nThe figure below consists of six plots, arranged in a grid of three rows and two columns. The first row displays the daily MYI extent for each boreal winter between October 1991 and April 2020, estimated from the CDS dataset. The second row shows the difference (\"bias\") between the CDS and the NERSC datasets, while the third row shows the corresponding integrated ice type error. The columns distinguish between the two definitions of multi-year sea ice for the CDS dataset. The right column shows the results based solely on the grid points labelled as \"multi-year ice\", whereas the left column shows results based on the grid points labelled as either \"multi-year ice\" or \"ambiguous\".\n\nThe comparison between the MYI extent from the CDS and the NERSC datasets yields comparable results, and the inclusion of the grid points labelled as \"ambiguous\" does not significantly alter the main conclusions. The results indicate that the spatial extent of multi-year sea ice in the Arctic has decreased since October 1991. Specifically, the October MYI extent dropped by approximately 1-2 million km$^2$ between the 5-year periods 1991-1995 and 2015-2019. A comparable result is found for the MYI extent in April. Selecting the ambiguous points does not significantly alter the results.\n\nBoth datasets also suggest that the MYI extent tends to decrease throughout each boreal winter. This is expected because MYI tends to be both exported and compressed through ridging as winter progresses (the reader is referred to [[4]](https://tc.copernicus.org/articles/17/1873/2023/) for a comprehensive explanation of sources and sinks of multi-year sea ice).\n\nDespite their similarities, the datasets present some significant differences. Notably, the CDS dataset exhibits a slightly greater day-to-day variability compared to the NERSC dataset when the pixels labelled as \"ambiguous\" are included into the analysis. A thorough understanding of this discrepancy goes beyond the scope of this study. However, we hypothesize it to be partially attributed to atmospheric variability, which can affect the brightness temperature of sea ice and, therefore, the classification into first-year sea ice and multi-year sea ice. For example, warmer weather events pose a challenge to the classification of sea ice into ice types as they modify the surface of multi-year sea ice, making it appear similar to first-year sea ice (Product User Guide and Specification of the CDS dataset). This interpretation is corroborated by the day-to-day variability in the CDS dataset decreasing when the grid points labelled as \"ambiguous\" are not considered. The NERSC dataset on the other hand is based on relatively smooth, low-resolution sea ice concentration and drift products derived from lower-frequency (SSMIS) passive microwave measurements which are not as affected by atmospheric variability."} {"chunk_id": "satellite_satellite-sea-ice-edge-type_intercomparison_q01__145d5956a471", "report_id": "satellite_satellite-sea-ice-edge-type_intercomparison_q01", "dataset_id": "satellite-sea-ice-edge-type", "store": "CDS", "doc_type": "EQC_QA", "aspect": "intercomparison_q01", "aspect_base": "intercomparison", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Intercomparison of satellite multi-year sea ice extent estimates > Analysis and results > 3. Results > 3.1 Temporal evolution of the total area covered by multi-year sea ice in the Arctic Basin", "title": "Intercomparison of satellite multi-year sea ice extent estimates", "chunk_index": 14, "token_count": 521, "text_raw": "The NERSC dataset on the other hand is based on relatively smooth, low-resolution sea ice concentration and drift products derived from lower-frequency (SSMIS) passive microwave measurements which are not as affected by atmospheric variability.\n\nThe differences between the two datasets also appear in the four plots at the bottom, which show the temporal evolution of the bias and the integrated ice type error. Both diagnostics tend to progressively diverge from zero as each boreal winter evolves. This is not entirely surprising as at the end of summer, the Arctic by definition contains only multi-year ice, but the potential for mis-classification (by both the CDS and NERSC products) grows later in the winter as more FYI forms. With the NERSC product, the drift tracking algorithm will accumulate errors further into the winter; with the CDS product, the distinction between classes is disrupted in the spring months as warmer temperatures, melting, and dynamics (eg. increased deformation of FYI) disturb the emissivity of the sea ice.\n\nThe figure above indicates that the exclusion from the analysis of the pixels labelled as \"ambiguous\" leads to stronger decrease of multi-year sea ice extent throughout each boreal winter. This result better agrees with that provided by the NERSC dataset, with the exception of the month of April. In April, the penultimate subplot on the right shows a drop in the bias, especially during the first decade of the analysis, as indicated by the blue lines. Unfortunately, it is difficult to understand the causes behind this issue. On the one hand, the discrepancy might indicate that, in the month of April, some of the grid points labelled as \"ambiguous\" contain multi-year sea ice. On the other hand, it might indicate that the NERSC dataset interprets part of first-year sea ice as multi-year sea ice. This might occur because, as noticed in [[3]](https://tc.copernicus.org/articles/12/2073/2018/), the algorithm might prioritize small concentration of multi-year sea ice, therefore overstimating the age of sea ice at each pixel.\n\n(section-3.2)=", "text_with_prefix": "EQC Quality Assessment: \"Intercomparison of satellite multi-year sea ice extent estimates\"\nDataset: satellite-sea-ice-edge-type [CDS]\nAspect: intercomparison_q01 | Category: Satellite_ECVs\nSection: Intercomparison of satellite multi-year sea ice extent estimates > Analysis and results > 3. Results > 3.1 Temporal evolution of the total area covered by multi-year sea ice in the Arctic Basin\n---\nThe NERSC dataset on the other hand is based on relatively smooth, low-resolution sea ice concentration and drift products derived from lower-frequency (SSMIS) passive microwave measurements which are not as affected by atmospheric variability.\n\nThe differences between the two datasets also appear in the four plots at the bottom, which show the temporal evolution of the bias and the integrated ice type error. Both diagnostics tend to progressively diverge from zero as each boreal winter evolves. This is not entirely surprising as at the end of summer, the Arctic by definition contains only multi-year ice, but the potential for mis-classification (by both the CDS and NERSC products) grows later in the winter as more FYI forms. With the NERSC product, the drift tracking algorithm will accumulate errors further into the winter; with the CDS product, the distinction between classes is disrupted in the spring months as warmer temperatures, melting, and dynamics (eg. increased deformation of FYI) disturb the emissivity of the sea ice.\n\nThe figure above indicates that the exclusion from the analysis of the pixels labelled as \"ambiguous\" leads to stronger decrease of multi-year sea ice extent throughout each boreal winter. This result better agrees with that provided by the NERSC dataset, with the exception of the month of April. In April, the penultimate subplot on the right shows a drop in the bias, especially during the first decade of the analysis, as indicated by the blue lines. Unfortunately, it is difficult to understand the causes behind this issue. On the one hand, the discrepancy might indicate that, in the month of April, some of the grid points labelled as \"ambiguous\" contain multi-year sea ice. On the other hand, it might indicate that the NERSC dataset interprets part of first-year sea ice as multi-year sea ice. This might occur because, as noticed in [[3]](https://tc.copernicus.org/articles/12/2073/2018/), the algorithm might prioritize small concentration of multi-year sea ice, therefore overstimating the age of sea ice at each pixel.\n\n(section-3.2)="} {"chunk_id": "satellite_satellite-sea-ice-edge-type_intercomparison_q01__541072483da7", "report_id": "satellite_satellite-sea-ice-edge-type_intercomparison_q01", "dataset_id": "satellite-sea-ice-edge-type", "store": "CDS", "doc_type": "EQC_QA", "aspect": "intercomparison_q01", "aspect_base": "intercomparison", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Intercomparison of satellite multi-year sea ice extent estimates > Analysis and results > 3.2 Spatial and temporal evolution of the region covered by multi-year sea ice in the Arctic", "title": "Intercomparison of satellite multi-year sea ice extent estimates", "chunk_index": 15, "token_count": 136, "text_raw": "- from Oct 1991 to Apr 2001 (10 boreal winters)\n- from Oct 2001 to Apr 2011 (10 boreal winters)\n- from Oct 2011 to Apr 2020 (9 boreal winters)", "text_with_prefix": "EQC Quality Assessment: \"Intercomparison of satellite multi-year sea ice extent estimates\"\nDataset: satellite-sea-ice-edge-type [CDS]\nAspect: intercomparison_q01 | Category: Satellite_ECVs\nSection: Intercomparison of satellite multi-year sea ice extent estimates > Analysis and results > 3.2 Spatial and temporal evolution of the region covered by multi-year sea ice in the Arctic\n---\n- from Oct 1991 to Apr 2001 (10 boreal winters)\n- from Oct 2001 to Apr 2011 (10 boreal winters)\n- from Oct 2011 to Apr 2020 (9 boreal winters)"} {"chunk_id": "satellite_satellite-sea-ice-edge-type_intercomparison_q01__bd5e094d02c9", "report_id": "satellite_satellite-sea-ice-edge-type_intercomparison_q01", "dataset_id": "satellite-sea-ice-edge-type", "store": "CDS", "doc_type": "EQC_QA", "aspect": "intercomparison_q01", "aspect_base": "intercomparison", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Intercomparison of satellite multi-year sea ice extent estimates > Analysis and results > 3.2 Spatial and temporal evolution of the region covered by multi-year sea ice in the Arctic > Maps of multi-year sea ice from NERSC dataset", "title": "Intercomparison of satellite multi-year sea ice extent estimates", "chunk_index": 16, "token_count": 364, "text_raw": "The maps below visually complement the previous plots, illustrating the geographical distribution of multi-year sea ice in the Arctic, its evolution since the early 1990s, and the regions where the CDS and NERSC datasets diverge the most.\n\nThe initial set of maps refers to the NERSC dataset and displays, for each month and grid point, the percentage of days associated with multi-year sea ice across three distinct periods: from October 1991 to April 2001, from October 2001 to April 2011, and from October 2011 to April 2020. The figure comprises 21 maps, with each column corresponding to one of the three periods and each row to a different month.\n\nConsistent with the plots above, the maps show a significant decline in the multi-year sea ice extent since 1991, as well as throughout the boreal winters. The decline in multi-year sea ice appears to have mostly occurred in the Beaufort Sea, the Chukchi Sea, and the Eurasian Basin. On the contrary, multi-year sea ice has consistently been present in the Canadian sector and in the Greenland Sea throughout the entire period. The presence of multi-year sea ice off the coast of Greenland is noticeable and presumably due to the advection of multi-year sea ice through the Fram Strait.", "text_with_prefix": "EQC Quality Assessment: \"Intercomparison of satellite multi-year sea ice extent estimates\"\nDataset: satellite-sea-ice-edge-type [CDS]\nAspect: intercomparison_q01 | Category: Satellite_ECVs\nSection: Intercomparison of satellite multi-year sea ice extent estimates > Analysis and results > 3.2 Spatial and temporal evolution of the region covered by multi-year sea ice in the Arctic > Maps of multi-year sea ice from NERSC dataset\n---\nThe maps below visually complement the previous plots, illustrating the geographical distribution of multi-year sea ice in the Arctic, its evolution since the early 1990s, and the regions where the CDS and NERSC datasets diverge the most.\n\nThe initial set of maps refers to the NERSC dataset and displays, for each month and grid point, the percentage of days associated with multi-year sea ice across three distinct periods: from October 1991 to April 2001, from October 2001 to April 2011, and from October 2011 to April 2020. The figure comprises 21 maps, with each column corresponding to one of the three periods and each row to a different month.\n\nConsistent with the plots above, the maps show a significant decline in the multi-year sea ice extent since 1991, as well as throughout the boreal winters. The decline in multi-year sea ice appears to have mostly occurred in the Beaufort Sea, the Chukchi Sea, and the Eurasian Basin. On the contrary, multi-year sea ice has consistently been present in the Canadian sector and in the Greenland Sea throughout the entire period. The presence of multi-year sea ice off the coast of Greenland is noticeable and presumably due to the advection of multi-year sea ice through the Fram Strait."} {"chunk_id": "satellite_satellite-sea-ice-edge-type_intercomparison_q01__25dcf8fb03ec", "report_id": "satellite_satellite-sea-ice-edge-type_intercomparison_q01", "dataset_id": "satellite-sea-ice-edge-type", "store": "CDS", "doc_type": "EQC_QA", "aspect": "intercomparison_q01", "aspect_base": "intercomparison", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Intercomparison of satellite multi-year sea ice extent estimates > Analysis and results > 3.2 Spatial and temporal evolution of the region covered by multi-year sea ice in the Arctic > Maps of multi-year sea ice from CDS and \"CDS - NERSC\"", "title": "Intercomparison of satellite multi-year sea ice extent estimates", "chunk_index": 17, "token_count": 748, "text_raw": "The Jupyter Notebook further presents four sets of maps. These are meant to help visualize the differences between the CDS and the NERSC datasets. Each set follows the same structure as the figure above. However, the first and the second sets of maps respectively show the percentage of multi-year sea ice provided by the CDS dataset for the case when the presence of multi-year sea ice is only detected using the pixels labelled as 'multi-year ice' and when the 'ambiguous' grid points are considered. The remaining sets of maps show the differences between the percentage of multi-year sea ice provided by the CDS and the NERSC datasets. The third set is produced only using the grid points of the CDS dataset that are classified as ‘multi-year ice’, whereas the one underneath uses the grid points classified as 'multi-year ice' and 'ambiguous'.\n\nThe inclusion of \"ambiguous\" data points primarily affects the magnitude of the difference, rather than the patterns. More precisely, when \"ambiguous\" values are included into the analysis, the grid points of the CDS dataset are more likely associated with multi-year sea ice and, therefore, the difference between the CDS and the NERSC datasets becomes more positive.\n\nCompared to the NERSC dataset, the CDS product has less multi-year sea ice in the Greenland Sea, away from the coast of Greenland. Such a discrepancy might result from the difficulty of the NERSC dataset to well capture the impact of local dynamics on multi-year sea ice. Indeed, the drifts that were used to create the NERSC dataset have a daily temporal resolution. Thus, they might not fully represent the effects of the ocean since it is highly dynamic in the region. However, the discrepancy might also highlight issues with the CDS dataset as [[3]](https://tc.copernicus.org/articles/12/2073/2018/) showed the ability of the NERSC dataset to capture multi-year sea ice in the Greenland Sea through a comparison with Sentinel-1 SAR images, used as an independent dataset, on the 1st January 2016. Therefore, further analysis is necessary to determine what causes such a discrepancy and, therefore, which dataset most reliably provides information on multi-year sea ice in the region.\n\nMore difficult is to determine the quality of the two datasets in other sectors of the Arctic. For example, compared to the NERSC dataset, the CDS dataset shows higher percentages of multi-year sea ice in the Beaufort Sea, the Chukchi Sea, and the Eurasian Basin. We might speculate that, in the Eurasian Basin, the discrepancy might occur because the CDS dataset might overestimate the amount of MYI due to its binary classification, which labels a grid point as MYI even if contains substantial amount of FYI. In contrast, when analysing the NERSC dataset we have classified a pixel as MYI if the concentration of MYI is greater than the concentration of FYI, which may lead to there being less MYI in this product than in the one by CDS. However, further analysis, for example focusing on individual years, is needed to assess the causes behind these disagreements.", "text_with_prefix": "EQC Quality Assessment: \"Intercomparison of satellite multi-year sea ice extent estimates\"\nDataset: satellite-sea-ice-edge-type [CDS]\nAspect: intercomparison_q01 | Category: Satellite_ECVs\nSection: Intercomparison of satellite multi-year sea ice extent estimates > Analysis and results > 3.2 Spatial and temporal evolution of the region covered by multi-year sea ice in the Arctic > Maps of multi-year sea ice from CDS and \"CDS - NERSC\"\n---\nThe Jupyter Notebook further presents four sets of maps. These are meant to help visualize the differences between the CDS and the NERSC datasets. Each set follows the same structure as the figure above. However, the first and the second sets of maps respectively show the percentage of multi-year sea ice provided by the CDS dataset for the case when the presence of multi-year sea ice is only detected using the pixels labelled as 'multi-year ice' and when the 'ambiguous' grid points are considered. The remaining sets of maps show the differences between the percentage of multi-year sea ice provided by the CDS and the NERSC datasets. The third set is produced only using the grid points of the CDS dataset that are classified as ‘multi-year ice’, whereas the one underneath uses the grid points classified as 'multi-year ice' and 'ambiguous'.\n\nThe inclusion of \"ambiguous\" data points primarily affects the magnitude of the difference, rather than the patterns. More precisely, when \"ambiguous\" values are included into the analysis, the grid points of the CDS dataset are more likely associated with multi-year sea ice and, therefore, the difference between the CDS and the NERSC datasets becomes more positive.\n\nCompared to the NERSC dataset, the CDS product has less multi-year sea ice in the Greenland Sea, away from the coast of Greenland. Such a discrepancy might result from the difficulty of the NERSC dataset to well capture the impact of local dynamics on multi-year sea ice. Indeed, the drifts that were used to create the NERSC dataset have a daily temporal resolution. Thus, they might not fully represent the effects of the ocean since it is highly dynamic in the region. However, the discrepancy might also highlight issues with the CDS dataset as [[3]](https://tc.copernicus.org/articles/12/2073/2018/) showed the ability of the NERSC dataset to capture multi-year sea ice in the Greenland Sea through a comparison with Sentinel-1 SAR images, used as an independent dataset, on the 1st January 2016. Therefore, further analysis is necessary to determine what causes such a discrepancy and, therefore, which dataset most reliably provides information on multi-year sea ice in the region.\n\nMore difficult is to determine the quality of the two datasets in other sectors of the Arctic. For example, compared to the NERSC dataset, the CDS dataset shows higher percentages of multi-year sea ice in the Beaufort Sea, the Chukchi Sea, and the Eurasian Basin. We might speculate that, in the Eurasian Basin, the discrepancy might occur because the CDS dataset might overestimate the amount of MYI due to its binary classification, which labels a grid point as MYI even if contains substantial amount of FYI. In contrast, when analysing the NERSC dataset we have classified a pixel as MYI if the concentration of MYI is greater than the concentration of FYI, which may lead to there being less MYI in this product than in the one by CDS. However, further analysis, for example focusing on individual years, is needed to assess the causes behind these disagreements."} {"chunk_id": "satellite_satellite-sea-ice-edge-type_intercomparison_q01__065886ddefa2", "report_id": "satellite_satellite-sea-ice-edge-type_intercomparison_q01", "dataset_id": "satellite-sea-ice-edge-type", "store": "CDS", "doc_type": "EQC_QA", "aspect": "intercomparison_q01", "aspect_base": "intercomparison", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Intercomparison of satellite multi-year sea ice extent estimates > ℹ️ If you want to know more > Key resources", "title": "Intercomparison of satellite multi-year sea ice extent estimates", "chunk_index": 18, "token_count": 483, "text_raw": "CDS dataset used:\n\n* [Sea ice edge and type daily gridded data from 1978 to present derived from satellite observations](https://cds.climate.copernicus.eu/datasets/satellite-sea-ice-edge-type?tab=overview)\n\nAdditional resources:\n\n* Section \"SEA-ICE\" in the [State of the Global Climate 2023](https://library.wmo.int/records/item/68835-state-of-the-global-climate-2023)\n* [Learning material on sea ice provided by the National Snow and Ice Data Center](https://nsidc.org/learn/parts-cryosphere/sea-ice)\n* [Arctic multi-year sea ice visualization](https://svs.gsfc.nasa.gov/4251) provided by NASA\n* [Sea ice as a climate indicator](https://19january2021snapshot.epa.gov/climate-indicators/climate-change-indicators-arctic-sea-ice_.html )\n* [How sea ice is observed](https://www.metoffice.gov.uk/research/climate/cryosphere-oceans/sea-ice/measure)\n* [NARVAL, a visualization portal for Earth observation and model data in the high-latitude oceans](https://narval.nersc.no), offers visualizations of various sea-ice variables\n\nCode libraries used:\n\n* [pandas](https://pandas.pydata.org/) \n* [xarray](https://docs.xarray.dev/en/stable/) \n* [matplotlib](https://matplotlib.org/) \n* [cmocean](https://matplotlib.org/cmocean/) \n* [cartopy](https://scitools.org.uk/cartopy/docs/latest/) \n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)", "text_with_prefix": "EQC Quality Assessment: \"Intercomparison of satellite multi-year sea ice extent estimates\"\nDataset: satellite-sea-ice-edge-type [CDS]\nAspect: intercomparison_q01 | Category: Satellite_ECVs\nSection: Intercomparison of satellite multi-year sea ice extent estimates > ℹ️ If you want to know more > Key resources\n---\nCDS dataset used:\n\n* [Sea ice edge and type daily gridded data from 1978 to present derived from satellite observations](https://cds.climate.copernicus.eu/datasets/satellite-sea-ice-edge-type?tab=overview)\n\nAdditional resources:\n\n* Section \"SEA-ICE\" in the [State of the Global Climate 2023](https://library.wmo.int/records/item/68835-state-of-the-global-climate-2023)\n* [Learning material on sea ice provided by the National Snow and Ice Data Center](https://nsidc.org/learn/parts-cryosphere/sea-ice)\n* [Arctic multi-year sea ice visualization](https://svs.gsfc.nasa.gov/4251) provided by NASA\n* [Sea ice as a climate indicator](https://19january2021snapshot.epa.gov/climate-indicators/climate-change-indicators-arctic-sea-ice_.html )\n* [How sea ice is observed](https://www.metoffice.gov.uk/research/climate/cryosphere-oceans/sea-ice/measure)\n* [NARVAL, a visualization portal for Earth observation and model data in the high-latitude oceans](https://narval.nersc.no), offers visualizations of various sea-ice variables\n\nCode libraries used:\n\n* [pandas](https://pandas.pydata.org/) \n* [xarray](https://docs.xarray.dev/en/stable/) \n* [matplotlib](https://matplotlib.org/) \n* [cmocean](https://matplotlib.org/cmocean/) \n* [cartopy](https://scitools.org.uk/cartopy/docs/latest/) \n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)"} {"chunk_id": "satellite_satellite-sea-ice-edge-type_intercomparison_q01__41de9dc12359", "report_id": "satellite_satellite-sea-ice-edge-type_intercomparison_q01", "dataset_id": "satellite-sea-ice-edge-type", "store": "CDS", "doc_type": "EQC_QA", "aspect": "intercomparison_q01", "aspect_base": "intercomparison", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Intercomparison of satellite multi-year sea ice extent estimates > ℹ️ If you want to know more > References", "title": "Intercomparison of satellite multi-year sea ice extent estimates", "chunk_index": 19, "token_count": 567, "text_raw": "1. Maslanik, J., Stroeve, J., Fowler, C., and Emery, W.: Distribution and trends in Arctic sea ice age through spring 2011, Geophys. Res. Lett., 38, L13502, https://doi.org/10.1029/2011GL047735, 2011.\n2. Meredith, M., Sommerkorn, M., Cassotta, S., Derksen, C., Ekaykin, A., Hollowed, A., Kofinas, G., Mackintosh, A., Melbourne-Thomas, J., Muelbert, M. M. C., Ottersen, G., Pritchard, H., and Schuur, E.: Polar Regions, in: IPCC Special Report on the Ocean and Cryosphere in a Changing Climate, edited by: Pörtner, H.-O., Roberts, D. C., Masson-Delmotte, V., Zhai, P., Tignor, M., Poloczanska, E., Mintenbeck, K., Alegría, A., Nicolai, M., Okem, A., Petzold, J., Rama, B., and Weyer, N. M., chap. 3, 203–320, Cambridge University Press, Cambridge, UK and New York, NY, USA, https://doi.org/10.1017/9781009157964.005, 2019.\n3. Korosov, A. A., Rampal, P., Pedersen, L. T., Saldo, R., Ye, Y., Heygster, G., Lavergne, T., Aaboe, S., and Girard-Ardhuin, F.: A new tracking algorithm for sea ice age distribution estimation, The Cryosphere, 12, 2073–2085, https://doi.org/10.5194/tc-12-2073-2018, 2018.\n4. Regan, H., Rampal, P., Ólason, E., Boutin, G., Korosov, A.: Modelling the evolution of Arctic multiyear sea ice over 2000-2018, The Cryosphere, 17, 1873-1893, https://doi.org/10.5194/tc-17-1873-2023, 2023.", "text_with_prefix": "EQC Quality Assessment: \"Intercomparison of satellite multi-year sea ice extent estimates\"\nDataset: satellite-sea-ice-edge-type [CDS]\nAspect: intercomparison_q01 | Category: Satellite_ECVs\nSection: Intercomparison of satellite multi-year sea ice extent estimates > ℹ️ If you want to know more > References\n---\n1. Maslanik, J., Stroeve, J., Fowler, C., and Emery, W.: Distribution and trends in Arctic sea ice age through spring 2011, Geophys. Res. Lett., 38, L13502, https://doi.org/10.1029/2011GL047735, 2011.\n2. Meredith, M., Sommerkorn, M., Cassotta, S., Derksen, C., Ekaykin, A., Hollowed, A., Kofinas, G., Mackintosh, A., Melbourne-Thomas, J., Muelbert, M. M. C., Ottersen, G., Pritchard, H., and Schuur, E.: Polar Regions, in: IPCC Special Report on the Ocean and Cryosphere in a Changing Climate, edited by: Pörtner, H.-O., Roberts, D. C., Masson-Delmotte, V., Zhai, P., Tignor, M., Poloczanska, E., Mintenbeck, K., Alegría, A., Nicolai, M., Okem, A., Petzold, J., Rama, B., and Weyer, N. M., chap. 3, 203–320, Cambridge University Press, Cambridge, UK and New York, NY, USA, https://doi.org/10.1017/9781009157964.005, 2019.\n3. Korosov, A. A., Rampal, P., Pedersen, L. T., Saldo, R., Ye, Y., Heygster, G., Lavergne, T., Aaboe, S., and Girard-Ardhuin, F.: A new tracking algorithm for sea ice age distribution estimation, The Cryosphere, 12, 2073–2085, https://doi.org/10.5194/tc-12-2073-2018, 2018.\n4. Regan, H., Rampal, P., Ólason, E., Boutin, G., Korosov, A.: Modelling the evolution of Arctic multiyear sea ice over 2000-2018, The Cryosphere, 17, 1873-1893, https://doi.org/10.5194/tc-17-1873-2023, 2023."} {"chunk_id": "satellite_satellite-sea-ice-thickness_trend-assessment_q01__992e73b804e5", "report_id": "satellite_satellite-sea-ice-thickness_trend-assessment_q01", "dataset_id": "satellite-sea-ice-thickness", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q01", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Suitability of satellite sea ice thickness data for studying climate change > Quality assessment statement", "title": "Suitability of satellite sea ice thickness data for studying climate change", "chunk_index": 0, "token_count": 221, "text_raw": "These are the key outcomes of this assessment\n\n- We find that the usefullness of satellite sea ice thickness data for the purpose of studying climate change is limited, since the time series are relatively short and the sea ice thickness variable has quite a lot of interannual variability. Note that this conclusion does not apply to other applications of this data like validation of models (bearing in mind biases and uncertainties in the data).\n\n- In limited regions, linear trends can explain a significant amount (greater than 0.5) of the variability in the satellite data, but even in those small regions the fits are not overly representative of the satellite estimates. That is, with the small number of years with data, how good the fits are is difficult to assess visually.\n```", "text_with_prefix": "EQC Quality Assessment: \"Suitability of satellite sea ice thickness data for studying climate change\"\nDataset: satellite-sea-ice-thickness [CDS]\nAspect: trend-assessment_q01 | Category: Satellite_ECVs\nSection: Suitability of satellite sea ice thickness data for studying climate change > Quality assessment statement\n---\nThese are the key outcomes of this assessment\n\n- We find that the usefullness of satellite sea ice thickness data for the purpose of studying climate change is limited, since the time series are relatively short and the sea ice thickness variable has quite a lot of interannual variability. Note that this conclusion does not apply to other applications of this data like validation of models (bearing in mind biases and uncertainties in the data).\n\n- In limited regions, linear trends can explain a significant amount (greater than 0.5) of the variability in the satellite data, but even in those small regions the fits are not overly representative of the satellite estimates. That is, with the small number of years with data, how good the fits are is difficult to assess visually.\n```"} {"chunk_id": "satellite_satellite-sea-ice-thickness_trend-assessment_q01__0c7d842936a8", "report_id": "satellite_satellite-sea-ice-thickness_trend-assessment_q01", "dataset_id": "satellite-sea-ice-thickness", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q01", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Suitability of satellite sea ice thickness data for studying climate change > Methodology", "title": "Suitability of satellite sea ice thickness data for studying climate change", "chunk_index": 1, "token_count": 442, "text_raw": "We consider the Level-3 [sea ice thickness dataset](https://cds.climate.copernicus.eu/datasets/satellite-sea-ice-thickness?tab=overview), which has two CDR (Climate Data Record) products:\n- ENVISAT CDR (2002-2010)\n- CryoSat-2 CDR (2010-2020)\nCryoSat-2 also has an interim CDR (ICDR, covering 2020-2024) but we did not use this product in this notebook.\n\nWe fit linear trends to these two satellite products, and assess the goodness-of-fit using the $r^2$ and $p$-value statistics.\nWe do separate fits for each winter month (October-April), to all pixels with at least five years of data.\nMaps of the mean, standard deviation and RMS (root mean sqare) uncertainty of the thickness, the fitted linear trend and the goodness-of-fit statistics are shown.\nNote that the product uncertainty does not include biases and other errors. In the evaluations done ENVISAT was found to have a significant low bias while CryoSat-2 had a small positive bias (see the [Product User Guide, p5)](https://dast.copernicus-climate.eu/documents/satellite-sea-ice-thickness/level-3/v3-0/WP2-FDDP-2022-09_C3S2-Lot3_PUGS-of-v3.0-SeaIceThickness-products_v3.1_Final.pdf) for more details).\n\nThe \"Analysis and results\" section is structured as follows:\n\n**[](section-1)**\n\n**[](section-2)**\n\n**[](section-3)**", "text_with_prefix": "EQC Quality Assessment: \"Suitability of satellite sea ice thickness data for studying climate change\"\nDataset: satellite-sea-ice-thickness [CDS]\nAspect: trend-assessment_q01 | Category: Satellite_ECVs\nSection: Suitability of satellite sea ice thickness data for studying climate change > Methodology\n---\nWe consider the Level-3 [sea ice thickness dataset](https://cds.climate.copernicus.eu/datasets/satellite-sea-ice-thickness?tab=overview), which has two CDR (Climate Data Record) products:\n- ENVISAT CDR (2002-2010)\n- CryoSat-2 CDR (2010-2020)\nCryoSat-2 also has an interim CDR (ICDR, covering 2020-2024) but we did not use this product in this notebook.\n\nWe fit linear trends to these two satellite products, and assess the goodness-of-fit using the $r^2$ and $p$-value statistics.\nWe do separate fits for each winter month (October-April), to all pixels with at least five years of data.\nMaps of the mean, standard deviation and RMS (root mean sqare) uncertainty of the thickness, the fitted linear trend and the goodness-of-fit statistics are shown.\nNote that the product uncertainty does not include biases and other errors. In the evaluations done ENVISAT was found to have a significant low bias while CryoSat-2 had a small positive bias (see the [Product User Guide, p5)](https://dast.copernicus-climate.eu/documents/satellite-sea-ice-thickness/level-3/v3-0/WP2-FDDP-2022-09_C3S2-Lot3_PUGS-of-v3.0-SeaIceThickness-products_v3.1_Final.pdf) for more details).\n\nThe \"Analysis and results\" section is structured as follows:\n\n**[](section-1)**\n\n**[](section-2)**\n\n**[](section-3)**"} {"chunk_id": "satellite_satellite-sea-ice-thickness_trend-assessment_q01__0d6b2a73f60c", "report_id": "satellite_satellite-sea-ice-thickness_trend-assessment_q01", "dataset_id": "satellite-sea-ice-thickness", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q01", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Suitability of satellite sea ice thickness data for studying climate change > Analysis and results > 1. Parameters, requests and functions definition", "title": "Suitability of satellite sea ice thickness data for studying climate change", "chunk_index": 2, "token_count": 121, "text_raw": "- Define parameters and formulate requests for downloading with the EQC toolbox.\n- Define functions to be applied to process and reduce the size of the downloaded data.\n- Define functions to post-process and visualize the data.", "text_with_prefix": "EQC Quality Assessment: \"Suitability of satellite sea ice thickness data for studying climate change\"\nDataset: satellite-sea-ice-thickness [CDS]\nAspect: trend-assessment_q01 | Category: Satellite_ECVs\nSection: Suitability of satellite sea ice thickness data for studying climate change > Analysis and results > 1. Parameters, requests and functions definition\n---\n- Define parameters and formulate requests for downloading with the EQC toolbox.\n- Define functions to be applied to process and reduce the size of the downloaded data.\n- Define functions to post-process and visualize the data."} {"chunk_id": "satellite_satellite-sea-ice-thickness_trend-assessment_q01__165df4df8c14", "report_id": "satellite_satellite-sea-ice-thickness_trend-assessment_q01", "dataset_id": "satellite-sea-ice-thickness", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q01", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Suitability of satellite sea ice thickness data for studying climate change > Analysis and results > 1. Parameters, requests and functions definition > 1.2 Set parameters", "title": "Suitability of satellite sea ice thickness data for studying climate change", "chunk_index": 3, "token_count": 141, "text_raw": "- Set the time range for ENVISAT data with `year_start_envisat` and `year_stop_envisat`\n- Set the time range for CRYOSAT data with `year_start_cryosat` and `year_stop_cryosat`", "text_with_prefix": "EQC Quality Assessment: \"Suitability of satellite sea ice thickness data for studying climate change\"\nDataset: satellite-sea-ice-thickness [CDS]\nAspect: trend-assessment_q01 | Category: Satellite_ECVs\nSection: Suitability of satellite sea ice thickness data for studying climate change > Analysis and results > 1. Parameters, requests and functions definition > 1.2 Set parameters\n---\n- Set the time range for ENVISAT data with `year_start_envisat` and `year_stop_envisat`\n- Set the time range for CRYOSAT data with `year_start_cryosat` and `year_stop_cryosat`"} {"chunk_id": "satellite_satellite-sea-ice-thickness_trend-assessment_q01__dec5673b2aa1", "report_id": "satellite_satellite-sea-ice-thickness_trend-assessment_q01", "dataset_id": "satellite-sea-ice-thickness", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q01", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Suitability of satellite sea ice thickness data for studying climate change > Analysis and results > 1. Parameters, requests and functions definition > 1.4 Post-processing functions", "title": "Suitability of satellite sea ice thickness data for studying climate change", "chunk_index": 4, "token_count": 251, "text_raw": "- `sort_time` sorts the dataset by month and year.\n- `linregress_1d` fits a straight line to some training data (output and input).\n- `get_stats` calculates some statistics over the year dimension (mean, standard deviation, RMS uncertainty) and also the coefficients of a linear fit (input is year) and goodness of fit statistics ($r^2$ value and $p$-value). It outputs a reduced dataset with no year dimension.\n- `add_differences` adds the differences between the CryoSat-2 and ENVISAT statistics to the reduced dataset, and also reduces the spatial extent for better plotting.\n\nget SIT statistics and create dataset\nget linear fit of SIT vs year\nremove areas with no data ever\nensure grid is contiguous after reducing area", "text_with_prefix": "EQC Quality Assessment: \"Suitability of satellite sea ice thickness data for studying climate change\"\nDataset: satellite-sea-ice-thickness [CDS]\nAspect: trend-assessment_q01 | Category: Satellite_ECVs\nSection: Suitability of satellite sea ice thickness data for studying climate change > Analysis and results > 1. Parameters, requests and functions definition > 1.4 Post-processing functions\n---\n- `sort_time` sorts the dataset by month and year.\n- `linregress_1d` fits a straight line to some training data (output and input).\n- `get_stats` calculates some statistics over the year dimension (mean, standard deviation, RMS uncertainty) and also the coefficients of a linear fit (input is year) and goodness of fit statistics ($r^2$ value and $p$-value). It outputs a reduced dataset with no year dimension.\n- `add_differences` adds the differences between the CryoSat-2 and ENVISAT statistics to the reduced dataset, and also reduces the spatial extent for better plotting.\n\nget SIT statistics and create dataset\nget linear fit of SIT vs year\nremove areas with no data ever\nensure grid is contiguous after reducing area"} {"chunk_id": "satellite_satellite-sea-ice-thickness_trend-assessment_q01__d8075a72ee8e", "report_id": "satellite_satellite-sea-ice-thickness_trend-assessment_q01", "dataset_id": "satellite-sea-ice-thickness", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q01", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Suitability of satellite sea ice thickness data for studying climate change > Analysis and results > 1. Parameters, requests and functions definition > 1.5 Plotting functions", "title": "Suitability of satellite sea ice thickness data for studying climate change", "chunk_index": 5, "token_count": 260, "text_raw": "- `plot_maps` plots monthly maps for a given variable.\n- `plot_maps_trends` plots maps of the trend with the goodness-of-fit statistices, the $r^2$ and $p$ values.\n- `plot_maps_valid_years` plots monthly maps of the number of valid years for each satellite.\n- `check_fit` plots time series of the satellite thickness and error bars to compare to the linear fits for selected points.\n- `check_fit_envisat` selects the example points to check the fits to the ENVISAT data.\n- `check_fit_cryosat2` selects the example points to check the fits to the CryoSat-2 data.\n\nCreate plots\nLoop over variables\nplot a map showing the point locations\nget the trend line\nplot SIT time series and linear fit\n\n(section-2)=", "text_with_prefix": "EQC Quality Assessment: \"Suitability of satellite sea ice thickness data for studying climate change\"\nDataset: satellite-sea-ice-thickness [CDS]\nAspect: trend-assessment_q01 | Category: Satellite_ECVs\nSection: Suitability of satellite sea ice thickness data for studying climate change > Analysis and results > 1. Parameters, requests and functions definition > 1.5 Plotting functions\n---\n- `plot_maps` plots monthly maps for a given variable.\n- `plot_maps_trends` plots maps of the trend with the goodness-of-fit statistices, the $r^2$ and $p$ values.\n- `plot_maps_valid_years` plots monthly maps of the number of valid years for each satellite.\n- `check_fit` plots time series of the satellite thickness and error bars to compare to the linear fits for selected points.\n- `check_fit_envisat` selects the example points to check the fits to the ENVISAT data.\n- `check_fit_cryosat2` selects the example points to check the fits to the CryoSat-2 data.\n\nCreate plots\nLoop over variables\nplot a map showing the point locations\nget the trend line\nplot SIT time series and linear fit\n\n(section-2)="} {"chunk_id": "satellite_satellite-sea-ice-thickness_trend-assessment_q01__34f08a427eb9", "report_id": "satellite_satellite-sea-ice-thickness_trend-assessment_q01", "dataset_id": "satellite-sea-ice-thickness", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q01", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Suitability of satellite sea ice thickness data for studying climate change > Analysis and results > 2. Download and transform the data", "title": "Suitability of satellite sea ice thickness data for studying climate change", "chunk_index": 6, "token_count": 144, "text_raw": "This is where the data is downloaded, transformed using `monthly_weighted_linear_trend` and saved to disk by the EQC toolbox. If the code is rerun the transformed data is loaded from the disk.\nThe transformed data is then post-processed by `postprocess_dataset` after downloading or loading.\n\n(section-3)=", "text_with_prefix": "EQC Quality Assessment: \"Suitability of satellite sea ice thickness data for studying climate change\"\nDataset: satellite-sea-ice-thickness [CDS]\nAspect: trend-assessment_q01 | Category: Satellite_ECVs\nSection: Suitability of satellite sea ice thickness data for studying climate change > Analysis and results > 2. Download and transform the data\n---\nThis is where the data is downloaded, transformed using `monthly_weighted_linear_trend` and saved to disk by the EQC toolbox. If the code is rerun the transformed data is loaded from the disk.\nThe transformed data is then post-processed by `postprocess_dataset` after downloading or loading.\n\n(section-3)="} {"chunk_id": "satellite_satellite-sea-ice-thickness_trend-assessment_q01__6d52efa2951f", "report_id": "satellite_satellite-sea-ice-thickness_trend-assessment_q01", "dataset_id": "satellite-sea-ice-thickness", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q01", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Suitability of satellite sea ice thickness data for studying climate change > Analysis and results > 3. Results", "title": "Suitability of satellite sea ice thickness data for studying climate change", "chunk_index": 7, "token_count": 168, "text_raw": "In section [](section-3.1) we plot monthly maps showing the mean, variability and RMS uncertainty of sea ice thickness for the two satellites, and also plot the difference between the means for the two satellites to get an idea of the change in thickness going from the earlier to the later period.\n\nIn section [](section-3.2), we compute linear trends for the two satellites and test the goodness-of-fit of the linear fit.\n\n(section-3.1)=", "text_with_prefix": "EQC Quality Assessment: \"Suitability of satellite sea ice thickness data for studying climate change\"\nDataset: satellite-sea-ice-thickness [CDS]\nAspect: trend-assessment_q01 | Category: Satellite_ECVs\nSection: Suitability of satellite sea ice thickness data for studying climate change > Analysis and results > 3. Results\n---\nIn section [](section-3.1) we plot monthly maps showing the mean, variability and RMS uncertainty of sea ice thickness for the two satellites, and also plot the difference between the means for the two satellites to get an idea of the change in thickness going from the earlier to the later period.\n\nIn section [](section-3.2), we compute linear trends for the two satellites and test the goodness-of-fit of the linear fit.\n\n(section-3.1)="} {"chunk_id": "satellite_satellite-sea-ice-thickness_trend-assessment_q01__9466c4e82864", "report_id": "satellite_satellite-sea-ice-thickness_trend-assessment_q01", "dataset_id": "satellite-sea-ice-thickness", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q01", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Suitability of satellite sea ice thickness data for studying climate change > Analysis and results > 3. Results > 3.1 Temporal mean and variability, and difference between satellite periods", "title": "Suitability of satellite sea ice thickness data for studying climate change", "chunk_index": 8, "token_count": 971, "text_raw": "Below we show the mean sea ice thickness for the two satellites ENVISAT (2002-2010) and CryoSat-2 (2010-2020), along with the difference between them.\nThe thickness in the CryoSat-2 product is higher at the coast north of Greenland and Canada and this extends to around the North Pole in earlier winter and a bit further later in the winter.\nThis pattern is consistent with the pattern in the ENVISAT thickness, but much of the area of interest is obscured by the large polar hole in that product.\nHowever, looking at the difference in the means, there is a slight drop in thickness in the overlapping areas of thicker ice.\nIn general the mean thickness has dropped, with the exception of the area north of the Canadian archipelago (adjacent to the overlapping area of thicker ice) and in the Greenland Sea in the later winter months.\n\nThe two satellites have different biases (CryoSat-2 has a slight positive bias, while ENVISAT has a more significant negative bias) and accuracies so we should be be a little careful when comparing and trying to deduce temporal changes between them.\nGiven the satellite biases we can trust the reductions in thickness more than the increases in thickness. Moreover the increases in thickness also occur in regions of higher variability (plotted after the means and the difference in the means). Having said this we have used an 8-10 year averaging period so some of the variability will have been smoothed out.\n\nBelow we plot the standard deviations of the two satellite periods, and also the differences between them. For both the ENVISAT and the CryoSat-2 satellites, the variability is greatest in the areas with thicker and old ice, possibly since these areas are usually quite deformed due to ridging.\nThere is also high variability in the Greenland Sea, which is probably due in part to the highly dynamic nature of that area - the ice flows quite fast once it is exported through the Fram Strait. This exported ice can also be quite old and deformed with ridges which would also add to the variability. Looking at the differences, the variability has generally increased, with the exception of the area between the Beaufort Sea and the ENVISAT polar hole. This could be due to the higher resolution of the CryoSat-2 satellite, picking up smaller scale changes that ENVISAT cannot resolve, especially in regions of multi-year ice. CryoSat-2 also has less variability in the Labrador Sea Hudson bay from January to April, possibly since the ice there is generally more level, first-year ice.\nIncreased drift and deformation of the ice in latter years ([Rampal et al, 2009](https://doi.org/10.1029/2008JC005066)) may also be a factor.\n\nBelow we plot the RMS uncertainties of the two satellite periods, and also the differences between them. As with the variability, for both the ENVISAT and the CryoSat-2 satellites, the variability is greatest in the areas with thicker and old ice. It is also higher in coastal areas, possibly from a combination of more deformed ice, more uncertain snowfall and reference sea surface height, and radar signal contamination from land. Again like the variability, there is quite high uncertainty in the Greenland Sea. The uncertainty increases markedly as the winter progresses, possibly due to the ice becoming more deformed, and maybe higher errors in snow depth as more snow accumulates. In addition, once there starts to be surface melt melt ponds also start to interfere with the radar signal, which could also be increasing the April uncertainty.\n\nNote that the uncertainties above do not include biases and other errors. In the evaluations done ENVISAT was found to have a significant low bias while CryoSat-2 had a small positive bias (see the [Product User Guide, p5)](https://dast.copernicus-climate.eu/documents/satellite-sea-ice-thickness/level-3/v3-0/WP2-FDDP-2022-09_C3S2-Lot3_PUGS-of-v3.0-SeaIceThickness-products_v3.1_Final.pdf) for more details).\n\n(section-3.2)=", "text_with_prefix": "EQC Quality Assessment: \"Suitability of satellite sea ice thickness data for studying climate change\"\nDataset: satellite-sea-ice-thickness [CDS]\nAspect: trend-assessment_q01 | Category: Satellite_ECVs\nSection: Suitability of satellite sea ice thickness data for studying climate change > Analysis and results > 3. Results > 3.1 Temporal mean and variability, and difference between satellite periods\n---\nBelow we show the mean sea ice thickness for the two satellites ENVISAT (2002-2010) and CryoSat-2 (2010-2020), along with the difference between them.\nThe thickness in the CryoSat-2 product is higher at the coast north of Greenland and Canada and this extends to around the North Pole in earlier winter and a bit further later in the winter.\nThis pattern is consistent with the pattern in the ENVISAT thickness, but much of the area of interest is obscured by the large polar hole in that product.\nHowever, looking at the difference in the means, there is a slight drop in thickness in the overlapping areas of thicker ice.\nIn general the mean thickness has dropped, with the exception of the area north of the Canadian archipelago (adjacent to the overlapping area of thicker ice) and in the Greenland Sea in the later winter months.\n\nThe two satellites have different biases (CryoSat-2 has a slight positive bias, while ENVISAT has a more significant negative bias) and accuracies so we should be be a little careful when comparing and trying to deduce temporal changes between them.\nGiven the satellite biases we can trust the reductions in thickness more than the increases in thickness. Moreover the increases in thickness also occur in regions of higher variability (plotted after the means and the difference in the means). Having said this we have used an 8-10 year averaging period so some of the variability will have been smoothed out.\n\nBelow we plot the standard deviations of the two satellite periods, and also the differences between them. For both the ENVISAT and the CryoSat-2 satellites, the variability is greatest in the areas with thicker and old ice, possibly since these areas are usually quite deformed due to ridging.\nThere is also high variability in the Greenland Sea, which is probably due in part to the highly dynamic nature of that area - the ice flows quite fast once it is exported through the Fram Strait. This exported ice can also be quite old and deformed with ridges which would also add to the variability. Looking at the differences, the variability has generally increased, with the exception of the area between the Beaufort Sea and the ENVISAT polar hole. This could be due to the higher resolution of the CryoSat-2 satellite, picking up smaller scale changes that ENVISAT cannot resolve, especially in regions of multi-year ice. CryoSat-2 also has less variability in the Labrador Sea Hudson bay from January to April, possibly since the ice there is generally more level, first-year ice.\nIncreased drift and deformation of the ice in latter years ([Rampal et al, 2009](https://doi.org/10.1029/2008JC005066)) may also be a factor.\n\nBelow we plot the RMS uncertainties of the two satellite periods, and also the differences between them. As with the variability, for both the ENVISAT and the CryoSat-2 satellites, the variability is greatest in the areas with thicker and old ice. It is also higher in coastal areas, possibly from a combination of more deformed ice, more uncertain snowfall and reference sea surface height, and radar signal contamination from land. Again like the variability, there is quite high uncertainty in the Greenland Sea. The uncertainty increases markedly as the winter progresses, possibly due to the ice becoming more deformed, and maybe higher errors in snow depth as more snow accumulates. In addition, once there starts to be surface melt melt ponds also start to interfere with the radar signal, which could also be increasing the April uncertainty.\n\nNote that the uncertainties above do not include biases and other errors. In the evaluations done ENVISAT was found to have a significant low bias while CryoSat-2 had a small positive bias (see the [Product User Guide, p5)](https://dast.copernicus-climate.eu/documents/satellite-sea-ice-thickness/level-3/v3-0/WP2-FDDP-2022-09_C3S2-Lot3_PUGS-of-v3.0-SeaIceThickness-products_v3.1_Final.pdf) for more details).\n\n(section-3.2)="} {"chunk_id": "satellite_satellite-sea-ice-thickness_trend-assessment_q01__6210916479e6", "report_id": "satellite_satellite-sea-ice-thickness_trend-assessment_q01", "dataset_id": "satellite-sea-ice-thickness", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q01", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Suitability of satellite sea ice thickness data for studying climate change > Analysis and results > 3. Results > 3.2 Assessment of linear model for temporal variability", "title": "Suitability of satellite sea ice thickness data for studying climate change", "chunk_index": 9, "token_count": 973, "text_raw": "Before we show the trends and their goodness-of-fit statistics, we first plot maps showing the number of valid years for each pixel. The ENVISAT satellite generally has 8 valid years of data, with the exception of the marginal seas. CryoSat-2 has much less, possibly due to the smaller satellite footprint making it easier for the tracks to miss a given pixel. November/February/April mostly have 5/7/6 valid years, while other months have less than our minimum number of years (5). Hence we do not try to fit linear trends to October, December, January or March for the CryoSat-2 data.\n\nFor both satellite periods, we calcuated linear trends in sea ice thickness along with the goodness-of-fit statistices, the $r^2$ and $p$ values. The $r^2$ statistic measures the fraction of the variability explained by the linear trends and it is generally quite low, with the exception of a region between the Beaufort Sea and the polar hole in the ENVISAT data. This region also had the most significant reduction in sea ice variability. The $p$-value gives the probability of the data being explained by random variability around a zero trend. It tells a similar story to the $r^2$-value, in that it is generally quite high, with the exception of the same region in the ENVISAT data that had the higher r$^2$ value.\n\nBelow we plot the trend in the ENVISAT data for the month of February, along with the $r^2$ and $p$ values. The trend is about -10 cm/year in the region where the fit is acceptable. Other months are similar, in both the values and the region where the fit is acceptable. However, in general the linear fit to this data is explaining very little of the temporal variability.\n\nBelow we show three examples of the linear fits, at the positions shown on the map.\n\nThe first point has a high value of $r^2$ (0.79), and the fit is passing through all the error bars. It has a low $p$-value, so it is not well explained by random variations, but at the same time the behaviour is more of a step function than a linear decrease.\n\nThe second point has a borderline value of $r^2$ (0.46) and $p$-value (6.6%), so it is starting to be more explained by random variations (a common threshold is 5%). Visually the fit is not very representative of the data, and maybe a step function would be a better fit.\n\nThe third point has a very low value of $r^2$ and a high $p$-value or a high probability of being explained by random variability. With the exception of the first and last points, this series looks more like an increase followed by a decrease.\nThis example also shows the limitation of having such a short time series, in that removing possible outliers would leave very little data to work with.\n\nIn summary, for the ENVISAT data the region where linear fits are explaining a large fraction of the variability (as measured by $r^2$) is quite small, and even in those places a linear fit is still not overly convincing visually. The low number of data points also makes it hard to judge the fits visually.\n\nBelow we plot the trend in the CryoSat-2 data for the months with enough data, along with the $r^2$ and $p$ values. The regions where $r^2$ is reaching 0.5 and the $p$ value is below 0.1 are quite small and less coherent than for the ENVISAT data.\nIn general we can again say that the linear fits to the thickness data are explaining very little of the temporal variability.\nNovember has the largest area of higher $r^2$ but the thickness in that month is more uncertain for physical reasons (the ice is still quite dynamic; leads and melt ponds may still be refreezing; new ice has lower freeboard so it is harder to measure) so we do not concentrate on that one.\nSince melting is starting in April its thickness is also very uncertain, so we will now focus on the month of February, which also has the highest number of valid years (7).", "text_with_prefix": "EQC Quality Assessment: \"Suitability of satellite sea ice thickness data for studying climate change\"\nDataset: satellite-sea-ice-thickness [CDS]\nAspect: trend-assessment_q01 | Category: Satellite_ECVs\nSection: Suitability of satellite sea ice thickness data for studying climate change > Analysis and results > 3. Results > 3.2 Assessment of linear model for temporal variability\n---\nBefore we show the trends and their goodness-of-fit statistics, we first plot maps showing the number of valid years for each pixel. The ENVISAT satellite generally has 8 valid years of data, with the exception of the marginal seas. CryoSat-2 has much less, possibly due to the smaller satellite footprint making it easier for the tracks to miss a given pixel. November/February/April mostly have 5/7/6 valid years, while other months have less than our minimum number of years (5). Hence we do not try to fit linear trends to October, December, January or March for the CryoSat-2 data.\n\nFor both satellite periods, we calcuated linear trends in sea ice thickness along with the goodness-of-fit statistices, the $r^2$ and $p$ values. The $r^2$ statistic measures the fraction of the variability explained by the linear trends and it is generally quite low, with the exception of a region between the Beaufort Sea and the polar hole in the ENVISAT data. This region also had the most significant reduction in sea ice variability. The $p$-value gives the probability of the data being explained by random variability around a zero trend. It tells a similar story to the $r^2$-value, in that it is generally quite high, with the exception of the same region in the ENVISAT data that had the higher r$^2$ value.\n\nBelow we plot the trend in the ENVISAT data for the month of February, along with the $r^2$ and $p$ values. The trend is about -10 cm/year in the region where the fit is acceptable. Other months are similar, in both the values and the region where the fit is acceptable. However, in general the linear fit to this data is explaining very little of the temporal variability.\n\nBelow we show three examples of the linear fits, at the positions shown on the map.\n\nThe first point has a high value of $r^2$ (0.79), and the fit is passing through all the error bars. It has a low $p$-value, so it is not well explained by random variations, but at the same time the behaviour is more of a step function than a linear decrease.\n\nThe second point has a borderline value of $r^2$ (0.46) and $p$-value (6.6%), so it is starting to be more explained by random variations (a common threshold is 5%). Visually the fit is not very representative of the data, and maybe a step function would be a better fit.\n\nThe third point has a very low value of $r^2$ and a high $p$-value or a high probability of being explained by random variability. With the exception of the first and last points, this series looks more like an increase followed by a decrease.\nThis example also shows the limitation of having such a short time series, in that removing possible outliers would leave very little data to work with.\n\nIn summary, for the ENVISAT data the region where linear fits are explaining a large fraction of the variability (as measured by $r^2$) is quite small, and even in those places a linear fit is still not overly convincing visually. The low number of data points also makes it hard to judge the fits visually.\n\nBelow we plot the trend in the CryoSat-2 data for the months with enough data, along with the $r^2$ and $p$ values. The regions where $r^2$ is reaching 0.5 and the $p$ value is below 0.1 are quite small and less coherent than for the ENVISAT data.\nIn general we can again say that the linear fits to the thickness data are explaining very little of the temporal variability.\nNovember has the largest area of higher $r^2$ but the thickness in that month is more uncertain for physical reasons (the ice is still quite dynamic; leads and melt ponds may still be refreezing; new ice has lower freeboard so it is harder to measure) so we do not concentrate on that one.\nSince melting is starting in April its thickness is also very uncertain, so we will now focus on the month of February, which also has the highest number of valid years (7)."} {"chunk_id": "satellite_satellite-sea-ice-thickness_trend-assessment_q01__a07a2bdfbf05", "report_id": "satellite_satellite-sea-ice-thickness_trend-assessment_q01", "dataset_id": "satellite-sea-ice-thickness", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q01", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Suitability of satellite sea ice thickness data for studying climate change > Analysis and results > 3. Results > 3.2 Assessment of linear model for temporal variability", "title": "Suitability of satellite sea ice thickness data for studying climate change", "chunk_index": 10, "token_count": 311, "text_raw": "we do not concentrate on that one.\nSince melting is starting in April its thickness is also very uncertain, so we will now focus on the month of February, which also has the highest number of valid years (7).\n\nBelow we show three examples of the linear fits to the CryoSat-2 data for the month of February, at the positions shown on the map.\nThe first point has a higher value of $r^2$ (0.58), and a borderline $p$-value of 4.6%, so it is starting to have a significant chance of being explained by random variations.\nThe second point has a lower value of $r^2$ (0.37), so it is explaining less than half of the variability, and the $p$-value has increased to 15%. \nThe third point has a very low value of $r^2$ and a high $p$-value of 65%. While all of the fits are passing through all the error bars, it is difficult to decide visually if the fits are good ones due to the low number of points.", "text_with_prefix": "EQC Quality Assessment: \"Suitability of satellite sea ice thickness data for studying climate change\"\nDataset: satellite-sea-ice-thickness [CDS]\nAspect: trend-assessment_q01 | Category: Satellite_ECVs\nSection: Suitability of satellite sea ice thickness data for studying climate change > Analysis and results > 3. Results > 3.2 Assessment of linear model for temporal variability\n---\nwe do not concentrate on that one.\nSince melting is starting in April its thickness is also very uncertain, so we will now focus on the month of February, which also has the highest number of valid years (7).\n\nBelow we show three examples of the linear fits to the CryoSat-2 data for the month of February, at the positions shown on the map.\nThe first point has a higher value of $r^2$ (0.58), and a borderline $p$-value of 4.6%, so it is starting to have a significant chance of being explained by random variations.\nThe second point has a lower value of $r^2$ (0.37), so it is explaining less than half of the variability, and the $p$-value has increased to 15%. \nThe third point has a very low value of $r^2$ and a high $p$-value of 65%. While all of the fits are passing through all the error bars, it is difficult to decide visually if the fits are good ones due to the low number of points."} {"chunk_id": "satellite_satellite-sea-ice-thickness_trend-assessment_q01__4c234608d7a5", "report_id": "satellite_satellite-sea-ice-thickness_trend-assessment_q01", "dataset_id": "satellite-sea-ice-thickness", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q01", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Suitability of satellite sea ice thickness data for studying climate change > ℹ️ If you want to know more > Key resources", "title": "Suitability of satellite sea ice thickness data for studying climate change", "chunk_index": 11, "token_count": 292, "text_raw": "Introductory sea ice materials:\n- [Role of sea ice in the climate](https://marine.copernicus.eu/explainers/why-ocean-important/sea-ice)\n- [Sea ice as an indicator of climate change](https://climate.copernicus.eu/climate-indicators/sea-ice#)\n- [Observing sea ice with satellites](https://www.metoffice.gov.uk/research/climate/cryosphere-oceans/sea-ice/measure)\n\nCode libraries used:\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)", "text_with_prefix": "EQC Quality Assessment: \"Suitability of satellite sea ice thickness data for studying climate change\"\nDataset: satellite-sea-ice-thickness [CDS]\nAspect: trend-assessment_q01 | Category: Satellite_ECVs\nSection: Suitability of satellite sea ice thickness data for studying climate change > ℹ️ If you want to know more > Key resources\n---\nIntroductory sea ice materials:\n- [Role of sea ice in the climate](https://marine.copernicus.eu/explainers/why-ocean-important/sea-ice)\n- [Sea ice as an indicator of climate change](https://climate.copernicus.eu/climate-indicators/sea-ice#)\n- [Observing sea ice with satellites](https://www.metoffice.gov.uk/research/climate/cryosphere-oceans/sea-ice/measure)\n\nCode libraries used:\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)"} {"chunk_id": "satellite_satellite-sea-ice-thickness_trend-assessment_q01__b549b81afec5", "report_id": "satellite_satellite-sea-ice-thickness_trend-assessment_q01", "dataset_id": "satellite-sea-ice-thickness", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q01", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Suitability of satellite sea ice thickness data for studying climate change > ℹ️ If you want to know more > References", "title": "Suitability of satellite sea ice thickness data for studying climate change", "chunk_index": 12, "token_count": 401, "text_raw": "1. Tilling, R. L., Ridout, A., & Shepherd, A. (2018). Estimating Arctic sea ice thickness and volume using CryoSat-2 radar altimeter data. Advances in Space Research, 62(6), 1203-1225, [ https://doi.org/10.1016/j.asr.2017.10.051 ](https://doi.org/10.1016/j.asr.2017.10.051).\n\n2. Müller, F. L., Paul, S., Hendricks, S., and Dettmering, D. (2023). Monitoring Arctic thin ice: a comparison between CryoSat-2 SAR altimetry data and MODIS thermal-infrared imagery. The Cryosphere, 17, 809–825, [ https://doi.org/10.5194/tc-17-809-2023 ](https://doi.org/10.5194/tc-17-809-2023).\n\n3. Rampal, P., J. Weiss, and D. Marsan (2009). Positive trend in the mean speed and deformation rate of Arctic sea ice, 1979–2007, J. Geophys. Res., 114, C05013, [ https://doi.org/10.1029/2008JC005066 ](https://doi.org/10.1029/2008JC005066).", "text_with_prefix": "EQC Quality Assessment: \"Suitability of satellite sea ice thickness data for studying climate change\"\nDataset: satellite-sea-ice-thickness [CDS]\nAspect: trend-assessment_q01 | Category: Satellite_ECVs\nSection: Suitability of satellite sea ice thickness data for studying climate change > ℹ️ If you want to know more > References\n---\n1. Tilling, R. L., Ridout, A., & Shepherd, A. (2018). Estimating Arctic sea ice thickness and volume using CryoSat-2 radar altimeter data. Advances in Space Research, 62(6), 1203-1225, [ https://doi.org/10.1016/j.asr.2017.10.051 ](https://doi.org/10.1016/j.asr.2017.10.051).\n\n2. Müller, F. L., Paul, S., Hendricks, S., and Dettmering, D. (2023). Monitoring Arctic thin ice: a comparison between CryoSat-2 SAR altimetry data and MODIS thermal-infrared imagery. The Cryosphere, 17, 809–825, [ https://doi.org/10.5194/tc-17-809-2023 ](https://doi.org/10.5194/tc-17-809-2023).\n\n3. Rampal, P., J. Weiss, and D. Marsan (2009). Positive trend in the mean speed and deformation rate of Arctic sea ice, 1979–2007, J. Geophys. Res., 114, C05013, [ https://doi.org/10.1029/2008JC005066 ](https://doi.org/10.1029/2008JC005066)."} {"chunk_id": "satellite_satellite-sea-level-global_consistency_q01__a0b1beeee33d", "report_id": "satellite_satellite-sea-level-global_consistency_q01", "dataset_id": "satellite-sea-level-global", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q01", "aspect_base": "consistency", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Long-term global Eddy Kinetic Energy trends from satellite observations", "title": "Long-term global Eddy Kinetic Energy trends from satellite observations", "chunk_index": 0, "token_count": 116, "text_raw": "Production date: 4-11-2025\n\nProduced by: Joan Armajach and Blanca Fernández-Álvarez - IMEDEA (CSIC-UIB) [[webpage]](https://imedea.uib-csic.es/)", "text_with_prefix": "EQC Quality Assessment: \"Long-term global Eddy Kinetic Energy trends from satellite observations\"\nDataset: satellite-sea-level-global [CDS]\nAspect: consistency_q01 | Category: Satellite_ECVs\nSection: Long-term global Eddy Kinetic Energy trends from satellite observations\n---\nProduction date: 4-11-2025\n\nProduced by: Joan Armajach and Blanca Fernández-Álvarez - IMEDEA (CSIC-UIB) [[webpage]](https://imedea.uib-csic.es/)"} {"chunk_id": "satellite_satellite-sea-level-global_consistency_q01__9a96c5ac1e37", "report_id": "satellite_satellite-sea-level-global_consistency_q01", "dataset_id": "satellite-sea-level-global", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q01", "aspect_base": "consistency", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Long-term global Eddy Kinetic Energy trends from satellite observations > Quality assessment questions", "title": "Long-term global Eddy Kinetic Energy trends from satellite observations", "chunk_index": 1, "token_count": 887, "text_raw": "* **Can we assess the long-term variability of global ocean Eddy Kinetic Energy using satellite altimetry?**\n\nGlobal ocean circulation is both a cause and consequence of fluid interactions occurring across a wide range of spatial scales, from millimetres to more than 10,000 km [[1]](https://doi.org/10.1146/annurev.fluid.40.111406.102139). For ocean circulation, most of the available energy is expressed as kinetic energy (KE), which is a suitable index for measuring the intensity of ocean currents [[2]](https://doi.org/10.1126/sciadv.aax7727). Mesoscale variability, which ranges from tens to hundreds of kilometers depending on latitude [[3]](https://doi.org/10.1175/1520-0485%281998%29028<0433:GVOTFB>2.0.CO;2), contains nearly all of the ocean’s KE and represents the dominant signal in ocean circulation [[4]](https://doi.org/10.1029/2005GL024633). About 90% of the total KE is associated with the mesoscale activity, called Eddy Kinetic Energy (EKE) [[1](https://doi.org/10.1146/annurev.fluid.40.111406.102139), [5](https://doi.org/10.1038/s41598-025-06149-9)]. Meanders, eddies, fronts and jets are examples of mesoscale features that contribute a large part of the transport of heat, mass and chemical constituents in seawater [[6]](https://doi.org/10.1029/2007GL030812). Consequently, the dynamics of mesoscale features has a notable impact on the surface circulation and oceanic biogeochemistry, spanning local and global scales [[7]](https://doi.org/10.1175/JTECH-D-17-0010.1). To study the evolution of the EKE, gridded satellite sea level anomaly (SLA) products are widely used [[8]](https://doi.org/10.3389/fmars.2019.00703). Over the last two decades, satellite altimetry has significantly improved our ability to observe and understand ocean circulation at the mesoscale, providing global, high-resolution, and regular monitoring capabilities [[9](https://doi.org/10.1016/j.asr.2021.01.022), [10](https://doi.org/10.1016/j.asr.2011.09.033)]. At present, over 30 years of altimetric EKE time series are available, and several studies have reported a strengthening trend per decade in some energetic regions (e.g., [[5](https://doi.org/10.1038/s41598-025-06149-9), [11](https://doi.org/10.5194/egusphere-2025-4651), [12](https://doi.org/10.1038/s41558-021-01006-9)]). In long-term studies from satellite observations, we need to consider that gridded energy levels can vary depending on the data processing and the number of the altimeter missions [[13]](https://doi.org/10.1007/s10712-023-09778-9). As highlighted in [[13]](https://doi.org/10.1007/s10712-023-09778-9), a consistent sampling over time from the 2-satellite configuration is recommended for the evaluation of long-term EKE.\n\nThis notebook aims to assess the performance of sea level data from satellite observations to estimate the global EKE trends over 1993-2022, following [Barceló-Llull et al. (2025)](https://doi.org/10.1038/s41598-025-06149-9).", "text_with_prefix": "EQC Quality Assessment: \"Long-term global Eddy Kinetic Energy trends from satellite observations\"\nDataset: satellite-sea-level-global [CDS]\nAspect: consistency_q01 | Category: Satellite_ECVs\nSection: Long-term global Eddy Kinetic Energy trends from satellite observations > Quality assessment questions\n---\n* **Can we assess the long-term variability of global ocean Eddy Kinetic Energy using satellite altimetry?**\n\nGlobal ocean circulation is both a cause and consequence of fluid interactions occurring across a wide range of spatial scales, from millimetres to more than 10,000 km [[1]](https://doi.org/10.1146/annurev.fluid.40.111406.102139). For ocean circulation, most of the available energy is expressed as kinetic energy (KE), which is a suitable index for measuring the intensity of ocean currents [[2]](https://doi.org/10.1126/sciadv.aax7727). Mesoscale variability, which ranges from tens to hundreds of kilometers depending on latitude [[3]](https://doi.org/10.1175/1520-0485%281998%29028<0433:GVOTFB>2.0.CO;2), contains nearly all of the ocean’s KE and represents the dominant signal in ocean circulation [[4]](https://doi.org/10.1029/2005GL024633). About 90% of the total KE is associated with the mesoscale activity, called Eddy Kinetic Energy (EKE) [[1](https://doi.org/10.1146/annurev.fluid.40.111406.102139), [5](https://doi.org/10.1038/s41598-025-06149-9)]. Meanders, eddies, fronts and jets are examples of mesoscale features that contribute a large part of the transport of heat, mass and chemical constituents in seawater [[6]](https://doi.org/10.1029/2007GL030812). Consequently, the dynamics of mesoscale features has a notable impact on the surface circulation and oceanic biogeochemistry, spanning local and global scales [[7]](https://doi.org/10.1175/JTECH-D-17-0010.1). To study the evolution of the EKE, gridded satellite sea level anomaly (SLA) products are widely used [[8]](https://doi.org/10.3389/fmars.2019.00703). Over the last two decades, satellite altimetry has significantly improved our ability to observe and understand ocean circulation at the mesoscale, providing global, high-resolution, and regular monitoring capabilities [[9](https://doi.org/10.1016/j.asr.2021.01.022), [10](https://doi.org/10.1016/j.asr.2011.09.033)]. At present, over 30 years of altimetric EKE time series are available, and several studies have reported a strengthening trend per decade in some energetic regions (e.g., [[5](https://doi.org/10.1038/s41598-025-06149-9), [11](https://doi.org/10.5194/egusphere-2025-4651), [12](https://doi.org/10.1038/s41558-021-01006-9)]). In long-term studies from satellite observations, we need to consider that gridded energy levels can vary depending on the data processing and the number of the altimeter missions [[13]](https://doi.org/10.1007/s10712-023-09778-9). As highlighted in [[13]](https://doi.org/10.1007/s10712-023-09778-9), a consistent sampling over time from the 2-satellite configuration is recommended for the evaluation of long-term EKE.\n\nThis notebook aims to assess the performance of sea level data from satellite observations to estimate the global EKE trends over 1993-2022, following [Barceló-Llull et al. (2025)](https://doi.org/10.1038/s41598-025-06149-9)."} {"chunk_id": "satellite_satellite-sea-level-global_consistency_q01__7750f3ea0077", "report_id": "satellite_satellite-sea-level-global_consistency_q01", "dataset_id": "satellite-sea-level-global", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q01", "aspect_base": "consistency", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Long-term global Eddy Kinetic Energy trends from satellite observations > Quality assessment statement", "title": "Long-term global Eddy Kinetic Energy trends from satellite observations", "chunk_index": 2, "token_count": 343, "text_raw": "These are the key outcomes of this assessment\n\n* At the global level, mesoscale activity has intensified over the last decade. Most of this strengthening is consistently concentrated in regions characterized by high Eddy Kinetic Energy (EKE) in both the vDT2021 and vDT2024 datasets.\n\n* When comparing the sea level products distributed through the Climate Data Store (CDS), the vDT2021 dataset produces higher EKE both in the global ocean and in high-EKE regions throughout the studied period. Nevertheless, in regions such as the Kuroshio Extension (KExt), these differences between versions are much smaller. This demonstrates that a two-satellite constellation can properly capture mesoscale activity. A constellation with a stable number of satellites ensures the long-term stability of the data record, as is the case for this product, making it a suitable tool for this type of study.\n\n* In general, we recommend using the vDT2024 version, as it incorporates new input L2P (Level-2 Plus) altimeter standards compared to the older version (vDT2021). The L2P products serve as the main input for the altimeter-derived sea level production system, providing the necessary information to compute the along-track Sea Level Anomalies (SLA), while applying the same updated corrections and models for the altimeter missions. \n```", "text_with_prefix": "EQC Quality Assessment: \"Long-term global Eddy Kinetic Energy trends from satellite observations\"\nDataset: satellite-sea-level-global [CDS]\nAspect: consistency_q01 | Category: Satellite_ECVs\nSection: Long-term global Eddy Kinetic Energy trends from satellite observations > Quality assessment statement\n---\nThese are the key outcomes of this assessment\n\n* At the global level, mesoscale activity has intensified over the last decade. Most of this strengthening is consistently concentrated in regions characterized by high Eddy Kinetic Energy (EKE) in both the vDT2021 and vDT2024 datasets.\n\n* When comparing the sea level products distributed through the Climate Data Store (CDS), the vDT2021 dataset produces higher EKE both in the global ocean and in high-EKE regions throughout the studied period. Nevertheless, in regions such as the Kuroshio Extension (KExt), these differences between versions are much smaller. This demonstrates that a two-satellite constellation can properly capture mesoscale activity. A constellation with a stable number of satellites ensures the long-term stability of the data record, as is the case for this product, making it a suitable tool for this type of study.\n\n* In general, we recommend using the vDT2024 version, as it incorporates new input L2P (Level-2 Plus) altimeter standards compared to the older version (vDT2021). The L2P products serve as the main input for the altimeter-derived sea level production system, providing the necessary information to compute the along-track Sea Level Anomalies (SLA), while applying the same updated corrections and models for the altimeter missions. \n```"} {"chunk_id": "satellite_satellite-sea-level-global_consistency_q01__c92d77e50d07", "report_id": "satellite_satellite-sea-level-global_consistency_q01", "dataset_id": "satellite-sea-level-global", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q01", "aspect_base": "consistency", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Long-term global Eddy Kinetic Energy trends from satellite observations > Methodology", "title": "Long-term global Eddy Kinetic Energy trends from satellite observations", "chunk_index": 3, "token_count": 763, "text_raw": "Satellite Altimerty based sea level datasets are distributed by the [Copernicus Marine Environment Monitoring Service (CMEMS)](https://marine.copernicus.eu/) and the [Copernicus Climate Change Service (C3S)](https://climate.copernicus.eu/). In this assessment, we use [the Sea level gridded data from satellite observations](https://cds.climate.copernicus.eu/datasets/satellite-sea-level-global?tab=overview) provided by the [CDS](https://cds.climate.copernicus.eu/). This product is designed for climate applications and is based on a consistent and stable altimeter constellation to ensure the long-term stability of the ocean observing system. The constellation of the C3S product comprises more than 10 satellites, configured to ensure that two altimeters are always active. Of the satellites, one is always the active reference altimetry mission. Although there are changes over time, each replacement mission follows the same orbit, ensuring consistency. The other, known as the secondary mission, is a complementary satellite that provides significant information for estimating mesoscale signals. All the necessary information can be found in the [Product User Guide and Specification (PUGS)](https://confluence.ecmwf.int/display/CKB/Sea+Level+version+DT2024%3A+Product+User+Guide+and+Specification#SeaLevelversionDT2024:ProductUserGuideandSpecification-section4.2.2).\n\nAs previously mentioned, we follow the same methodology as in [[5]](https://doi.org/10.1038/s41598-025-06149-9) to estimate EKE time series and trends over the global ocean. In addition, we quantify EKE in regions of high mesoscale activity, as performed in that study. The authors used the vDT2018 and vDT2021 two-satellite products from C3S to compare them with the vDT2021 all-satellite product provided by CMEMS. The latter uses all available satellites, ranging from 2 to 7 over the altimetric period.\n\nIn this study, the vDT2024 version of the satellite gridded sea level observations available in the Climate Data Store is compared with the vDT2021, in order to evaluate how the new version resolves the EKE relative to the previous one. The TOPEX-A instrument drift correction for 1993-1998, which has been quantified in many studies (e.g., [[14]](https://doi.org/10.1038/NCLIMATE2635)), has not yet been computed in the vDT2024 data. In contrast, the vDT2021 product implements this correction as a separate variable (not directly included in the SLA estimate).\n\nThe analysis and results are organised in the following steps, which are detailed in the sections below:\n\n**[](section-1)**\n * Set up the python environment and import required libraries\n * Define the startup parameters\n * Define request for satellite altimetry, using both vDT2021 and vDT2024 datasets available from CDS\n\n**[](section-2)**\n\n**[](section-3)**\n\n**[](section-4)**\n\n**[](section-5)**\n * High EKE regions\n * Area-weighted mean EKE time series\n * EKE trends", "text_with_prefix": "EQC Quality Assessment: \"Long-term global Eddy Kinetic Energy trends from satellite observations\"\nDataset: satellite-sea-level-global [CDS]\nAspect: consistency_q01 | Category: Satellite_ECVs\nSection: Long-term global Eddy Kinetic Energy trends from satellite observations > Methodology\n---\nSatellite Altimerty based sea level datasets are distributed by the [Copernicus Marine Environment Monitoring Service (CMEMS)](https://marine.copernicus.eu/) and the [Copernicus Climate Change Service (C3S)](https://climate.copernicus.eu/). In this assessment, we use [the Sea level gridded data from satellite observations](https://cds.climate.copernicus.eu/datasets/satellite-sea-level-global?tab=overview) provided by the [CDS](https://cds.climate.copernicus.eu/). This product is designed for climate applications and is based on a consistent and stable altimeter constellation to ensure the long-term stability of the ocean observing system. The constellation of the C3S product comprises more than 10 satellites, configured to ensure that two altimeters are always active. Of the satellites, one is always the active reference altimetry mission. Although there are changes over time, each replacement mission follows the same orbit, ensuring consistency. The other, known as the secondary mission, is a complementary satellite that provides significant information for estimating mesoscale signals. All the necessary information can be found in the [Product User Guide and Specification (PUGS)](https://confluence.ecmwf.int/display/CKB/Sea+Level+version+DT2024%3A+Product+User+Guide+and+Specification#SeaLevelversionDT2024:ProductUserGuideandSpecification-section4.2.2).\n\nAs previously mentioned, we follow the same methodology as in [[5]](https://doi.org/10.1038/s41598-025-06149-9) to estimate EKE time series and trends over the global ocean. In addition, we quantify EKE in regions of high mesoscale activity, as performed in that study. The authors used the vDT2018 and vDT2021 two-satellite products from C3S to compare them with the vDT2021 all-satellite product provided by CMEMS. The latter uses all available satellites, ranging from 2 to 7 over the altimetric period.\n\nIn this study, the vDT2024 version of the satellite gridded sea level observations available in the Climate Data Store is compared with the vDT2021, in order to evaluate how the new version resolves the EKE relative to the previous one. The TOPEX-A instrument drift correction for 1993-1998, which has been quantified in many studies (e.g., [[14]](https://doi.org/10.1038/NCLIMATE2635)), has not yet been computed in the vDT2024 data. In contrast, the vDT2021 product implements this correction as a separate variable (not directly included in the SLA estimate).\n\nThe analysis and results are organised in the following steps, which are detailed in the sections below:\n\n**[](section-1)**\n * Set up the python environment and import required libraries\n * Define the startup parameters\n * Define request for satellite altimetry, using both vDT2021 and vDT2024 datasets available from CDS\n\n**[](section-2)**\n\n**[](section-3)**\n\n**[](section-4)**\n\n**[](section-5)**\n * High EKE regions\n * Area-weighted mean EKE time series\n * EKE trends"} {"chunk_id": "satellite_satellite-sea-level-global_consistency_q01__3579cd8197dc", "report_id": "satellite_satellite-sea-level-global_consistency_q01", "dataset_id": "satellite-sea-level-global", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q01", "aspect_base": "consistency", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Long-term global Eddy Kinetic Energy trends from satellite observations > Analysis and results > 1. Data selection and setup > Define startup parameters", "title": "Long-term global Eddy Kinetic Energy trends from satellite observations", "chunk_index": 4, "token_count": 132, "text_raw": "For this study, we select the period from 1993 to 2022. Moreover, we apply a mask to include only latitudes between 65°S and 65°N, ensuring that satellite observations remain consistent and are not affected by sea ice.\n\nGlobal region", "text_with_prefix": "EQC Quality Assessment: \"Long-term global Eddy Kinetic Energy trends from satellite observations\"\nDataset: satellite-sea-level-global [CDS]\nAspect: consistency_q01 | Category: Satellite_ECVs\nSection: Long-term global Eddy Kinetic Energy trends from satellite observations > Analysis and results > 1. Data selection and setup > Define startup parameters\n---\nFor this study, we select the period from 1993 to 2022. Moreover, we apply a mask to include only latitudes between 65°S and 65°N, ensuring that satellite observations remain consistent and are not affected by sea ice.\n\nGlobal region"} {"chunk_id": "satellite_satellite-sea-level-global_consistency_q01__060189d3619a", "report_id": "satellite_satellite-sea-level-global_consistency_q01", "dataset_id": "satellite-sea-level-global", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q01", "aspect_base": "consistency", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Long-term global Eddy Kinetic Energy trends from satellite observations > Analysis and results > 1. Data selection and setup > Define data request for satellite altimetry (vDT2021 and vDT2024)", "title": "Long-term global Eddy Kinetic Energy trends from satellite observations", "chunk_index": 5, "token_count": 150, "text_raw": "Here, we define the data request for the global gridded daily sea level products for each version (vDT2021 and vDT2024) using the [CDS API client](https://cds.climate.copernicus.eu/how-to-api).\n\n(section-2)=", "text_with_prefix": "EQC Quality Assessment: \"Long-term global Eddy Kinetic Energy trends from satellite observations\"\nDataset: satellite-sea-level-global [CDS]\nAspect: consistency_q01 | Category: Satellite_ECVs\nSection: Long-term global Eddy Kinetic Energy trends from satellite observations > Analysis and results > 1. Data selection and setup > Define data request for satellite altimetry (vDT2021 and vDT2024)\n---\nHere, we define the data request for the global gridded daily sea level products for each version (vDT2021 and vDT2024) using the [CDS API client](https://cds.climate.copernicus.eu/how-to-api).\n\n(section-2)="} {"chunk_id": "satellite_satellite-sea-level-global_consistency_q01__cada6eb3094d", "report_id": "satellite_satellite-sea-level-global_consistency_q01", "dataset_id": "satellite-sea-level-global", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q01", "aspect_base": "consistency", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Long-term global Eddy Kinetic Energy trends from satellite observations > Analysis and results > 2. Eddy Kinetic Energy (EKE) computation", "title": "Long-term global Eddy Kinetic Energy trends from satellite observations", "chunk_index": 6, "token_count": 402, "text_raw": "To compute EKE, we apply the specific formula:\n\n$$\nEKE = \\frac{1}{2} (u_a^2 + v_a^2) \\quad \\text{(cm² s⁻²)}\n$$\n\nwhere $u_a^2$ and $v_a^2$ refer to the zonal and meridional components of the geostrophic velocity anomalies, respectively. These variables are derived from the gridded SLA field. The computation is based on a 9-point stencil width [[15]](https://doi.org/10.1029/2011JC007367) for latitudes outside the ±5°N band. In the equatorial band, the Lagerloef method [[16]](https://doi.org/10.1029/1999JC900197) is applied, using the β plane approximation.\n\nIn this part, we download the datasets according to the request, integrating the function that computes EKE once all the data are available. Chunks are necessary to manage large volumes of information. Finally, we concatenate the EKE data along the dimension \"version\".\n\n```text\nversion='vdt2021'\n```\n\n```text\n100%|██████████| 360/360 [02:10<00:00, 2.77it/s]\n```\n\n```text\nversion='vdt2024'\n```\n\n```text\n100%|██████████| 360/360 [01:55<00:00, 3.13it/s]\n```\n\n(section-3)=", "text_with_prefix": "EQC Quality Assessment: \"Long-term global Eddy Kinetic Energy trends from satellite observations\"\nDataset: satellite-sea-level-global [CDS]\nAspect: consistency_q01 | Category: Satellite_ECVs\nSection: Long-term global Eddy Kinetic Energy trends from satellite observations > Analysis and results > 2. Eddy Kinetic Energy (EKE) computation\n---\nTo compute EKE, we apply the specific formula:\n\n$$\nEKE = \\frac{1}{2} (u_a^2 + v_a^2) \\quad \\text{(cm² s⁻²)}\n$$\n\nwhere $u_a^2$ and $v_a^2$ refer to the zonal and meridional components of the geostrophic velocity anomalies, respectively. These variables are derived from the gridded SLA field. The computation is based on a 9-point stencil width [[15]](https://doi.org/10.1029/2011JC007367) for latitudes outside the ±5°N band. In the equatorial band, the Lagerloef method [[16]](https://doi.org/10.1029/1999JC900197) is applied, using the β plane approximation.\n\nIn this part, we download the datasets according to the request, integrating the function that computes EKE once all the data are available. Chunks are necessary to manage large volumes of information. Finally, we concatenate the EKE data along the dimension \"version\".\n\n```text\nversion='vdt2021'\n```\n\n```text\n100%|██████████| 360/360 [02:10<00:00, 2.77it/s]\n```\n\n```text\nversion='vdt2024'\n```\n\n```text\n100%|██████████| 360/360 [01:55<00:00, 3.13it/s]\n```\n\n(section-3)="} {"chunk_id": "satellite_satellite-sea-level-global_consistency_q01__5cd851085932", "report_id": "satellite_satellite-sea-level-global_consistency_q01", "dataset_id": "satellite-sea-level-global", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q01", "aspect_base": "consistency", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Long-term global Eddy Kinetic Energy trends from satellite observations > Analysis and results > 3. Definition of high EKE regions", "title": "Long-term global Eddy Kinetic Energy trends from satellite observations", "chunk_index": 7, "token_count": 518, "text_raw": "As described in [[5]](https://doi.org/10.1038/s41598-025-06149-9), specific areas with high mesoscale activity due to their ocean dynamics are referred to as high EKE regions. To identify these high-intensity regions, the 90th spatial percentile of the mean EKE over the full period (1993-2022) is computed.\n\nTemporal mean of EKE\nDefine high EKE regions (90th percentile)\n\nOnce we have identified the areas with high EKE activity, we define these regions using the latitude and longitude bounds for each region. These regions are: Brazil-Malvinas Confluence (BMC), Loop Current (LC), Gulf Stream (GS), Great Whirl and Socotra Eddy (GWSE), Agulhas Current (AC), Kuroshio Extension (KExt), and East Australian Current (EAC).\n\nFor the regions with high EKE, we apply a function that fills holes, removes small regions using a threshold, and smooths contours. This results in well-defined regions and the minimisation of the spurious inclusion of small patches.\n\nWe create a dataset with the masks of all individual high EKE regions, previously defined as \"regions\". First, we generate a binary global mask based on the mean EKE, where values of 1 mean that the threshold (90th percentile) is exceeded, and values of 0 represent otherwise. Once done, we extract the high EKE regions and apply the smoothing function. Then, we concatenate all the regional masks along the dimension named \"region\" to differentiate each one.\n\nAdditionally, we integrate new masks, one composed of all high EKE regions combined (named \"high EKE\"), another representing the global mask without NaN values of EKE (named \"no ice\"), and finally a mask representing the tropical band (named \"tropical\"), excluding the regions that are already included in \"high EKE\" mask.\n\nGlobal mask (1 for high EKE, 0)\nAll individual high EKE regions\nAll high EKE regions combined ('high EKE') and global ('no ice')\nTropical mask\n\n(section-4)=", "text_with_prefix": "EQC Quality Assessment: \"Long-term global Eddy Kinetic Energy trends from satellite observations\"\nDataset: satellite-sea-level-global [CDS]\nAspect: consistency_q01 | Category: Satellite_ECVs\nSection: Long-term global Eddy Kinetic Energy trends from satellite observations > Analysis and results > 3. Definition of high EKE regions\n---\nAs described in [[5]](https://doi.org/10.1038/s41598-025-06149-9), specific areas with high mesoscale activity due to their ocean dynamics are referred to as high EKE regions. To identify these high-intensity regions, the 90th spatial percentile of the mean EKE over the full period (1993-2022) is computed.\n\nTemporal mean of EKE\nDefine high EKE regions (90th percentile)\n\nOnce we have identified the areas with high EKE activity, we define these regions using the latitude and longitude bounds for each region. These regions are: Brazil-Malvinas Confluence (BMC), Loop Current (LC), Gulf Stream (GS), Great Whirl and Socotra Eddy (GWSE), Agulhas Current (AC), Kuroshio Extension (KExt), and East Australian Current (EAC).\n\nFor the regions with high EKE, we apply a function that fills holes, removes small regions using a threshold, and smooths contours. This results in well-defined regions and the minimisation of the spurious inclusion of small patches.\n\nWe create a dataset with the masks of all individual high EKE regions, previously defined as \"regions\". First, we generate a binary global mask based on the mean EKE, where values of 1 mean that the threshold (90th percentile) is exceeded, and values of 0 represent otherwise. Once done, we extract the high EKE regions and apply the smoothing function. Then, we concatenate all the regional masks along the dimension named \"region\" to differentiate each one.\n\nAdditionally, we integrate new masks, one composed of all high EKE regions combined (named \"high EKE\"), another representing the global mask without NaN values of EKE (named \"no ice\"), and finally a mask representing the tropical band (named \"tropical\"), excluding the regions that are already included in \"high EKE\" mask.\n\nGlobal mask (1 for high EKE, 0)\nAll individual high EKE regions\nAll high EKE regions combined ('high EKE') and global ('no ice')\nTropical mask\n\n(section-4)="} {"chunk_id": "satellite_satellite-sea-level-global_consistency_q01__84cf5acfd70e", "report_id": "satellite_satellite-sea-level-global_consistency_q01", "dataset_id": "satellite-sea-level-global", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q01", "aspect_base": "consistency", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Long-term global Eddy Kinetic Energy trends from satellite observations > Analysis and results > 4. EKE time series and trends computation", "title": "Long-term global Eddy Kinetic Energy trends from satellite observations", "chunk_index": 8, "token_count": 1093, "text_raw": "In this section, we calculate the area-weighted mean for the spatial averages using latitude weighting. This process is applied to all the masks.\n\nCompute time series of EKE for all masks\n\nHere, we define the regions used in the test (no ice, high EKE, KExt, and GS) and the dictionaries to store the original time series and a 12-month rolling mean over a continuous period to smooth fluctuations in the data.\n\nRegions for the test\n12-month rolling mean\nStore time series in dictionaries\n\n```text\n----------------\nno ice\n----------------\n\n----------------\nhigh EKE\n----------------\n\n----------------\nKE\n----------------\n\n----------------\nGS\n----------------\n```\n\nFor the computation of EKE trends, we implement the Theil-Sen estimator, a non-parametric method for fitting a line to sample points. This regression algorithm is more robust to outliers compared to ordinary least-squares regression, which is sensitive to them. To assess the significance of trends, we use the Mann-Kendall test. However, the presence of serial correlation in time series can affect the results [[17]](https://doi.org/10.1023/B:WARM.0000043140.61082.60). Therefore, we employ the modified Mann-Kendall test proposed by [[17]](https://doi.org/10.1023/B:WARM.0000043140.61082.60), which accounts for this autocorrelation.\n\nFurthermore, we convert the trends, initially in cm² s⁻² month⁻¹ to cm² s⁻² year⁻¹ using a factor to facilitate the comparison of the results.\n\nFunction to compute trends (Mann-Kendall test)\nMultply by the factor 12 months to convert trends per month to trends per year\n\nIn this case, we calculate the trends over different periods of the EKE time series, starting each trend in a different year between 1993 and 2013 (inclusive) and ending in the last year of the study (2022). The MK test is applied to each EKE time series for every region and version. As before, we create new dictionaries to store all trends and p-values.\n\nYears for the test\nUpdated dictionary keys\nLists to save results for high EKE regions, global ocean, Kuroshio Extension (KExt) and Gulf Stream (GS)\nGlobal ocean\nHigh EKE regions\nKuroshio Extension (KExt)\n\n```text\n1993\n\n----------------\nno ice\n----------------\n----------------\nhigh EKE\n----------------\n----------------\nKE\n----------------\n----------------\nGS\n----------------\n\n1994\n\n----------------\nno ice\n----------------\n----------------\nhigh EKE\n----------------\n----------------\nKE\n----------------\n----------------\nGS\n----------------\n\n1995\n\n----------------\nno ice\n----------------\n----------------\nhigh EKE\n----------------\n----------------\nKE\n----------------\n----------------\nGS\n----------------\n\n1996\n\n----------------\nno ice\n----------------\n----------------\nhigh EKE\n----------------\n----------------\nKE\n----------------\n----------------\nGS\n----------------\n\n1997\n\n----------------\nno ice\n----------------\n----------------\nhigh EKE\n----------------\n----------------\nKE\n----------------\n----------------\nGS\n----------------\n\n1998\n\n----------------\nno ice\n----------------\n----------------\nhigh EKE\n----------------\n----------------\nKE\n----------------\n----------------\nGS\n----------------\n\n1999\n\n----------------\nno ice\n----------------\n----------------\nhigh EKE\n----------------\n----------------\nKE\n----------------\n----------------\nGS\n----------------\n\n2000\n\n----------------\nno ice\n----------------\n----------------\nhigh EKE\n----------------\n----------------\nKE\n----------------\n----------------\nGS\n----------------\n\n2001\n\n----------------\nno ice\n----------------\n----------------\nhigh EKE\n----------------\n----------------\nKE\n----------------\n----------------\nGS\n----------------\n\n2002\n\n----------------\nno ice\n----------------\n----------------\nhigh EKE\n----------------\n----------------\nKE\n----------------\n----------------\nGS\n----------------\n\n2003\n\n----------------\nno ice\n----------------\n----------------\nhigh EKE\n----------------\n----------------\nKE\n----------------\n----------------\nGS\n----------------\n\n2004\n\n----------------\nno ice\n----------------\n----------------\nhigh EKE\n----------------\n----------------\nKE\n----------------\n----------------\nGS\n----------------\n\n2005\n\n----------------\nno ice\n----------------\n----------------\nhigh EKE\n----------------\n----------------\nKE\n----------------\n----------------\nGS\n----------------\n\n2006\n\n----------------\nno ice\n----------------\n----------------\nhigh EKE\n----------------\n----------------\nKE\n----------------\n----------------\nGS\n----------------\n\n2007\n\n----------------\nno ice\n----------------\n----------------\nhigh EKE\n----------------\n----------------\nKE\n----------------\n----------------\nGS\n----------------\n\n2008\n\n----------------\nno ice\n----------------\n----------------\nhigh EKE\n----------------\n----------------\nKE\n----------------\n----------------\nGS\n----------------\n\n2009\n\n----------------\nno ice\n----------------\n----------------\nhigh EKE\n----------------\n----------------\nKE\n----------------\n----------------\nGS\n----------------\n\n2010", "text_with_prefix": "EQC Quality Assessment: \"Long-term global Eddy Kinetic Energy trends from satellite observations\"\nDataset: satellite-sea-level-global [CDS]\nAspect: consistency_q01 | Category: Satellite_ECVs\nSection: Long-term global Eddy Kinetic Energy trends from satellite observations > Analysis and results > 4. EKE time series and trends computation\n---\nIn this section, we calculate the area-weighted mean for the spatial averages using latitude weighting. This process is applied to all the masks.\n\nCompute time series of EKE for all masks\n\nHere, we define the regions used in the test (no ice, high EKE, KExt, and GS) and the dictionaries to store the original time series and a 12-month rolling mean over a continuous period to smooth fluctuations in the data.\n\nRegions for the test\n12-month rolling mean\nStore time series in dictionaries\n\n```text\n----------------\nno ice\n----------------\n\n----------------\nhigh EKE\n----------------\n\n----------------\nKE\n----------------\n\n----------------\nGS\n----------------\n```\n\nFor the computation of EKE trends, we implement the Theil-Sen estimator, a non-parametric method for fitting a line to sample points. This regression algorithm is more robust to outliers compared to ordinary least-squares regression, which is sensitive to them. To assess the significance of trends, we use the Mann-Kendall test. However, the presence of serial correlation in time series can affect the results [[17]](https://doi.org/10.1023/B:WARM.0000043140.61082.60). Therefore, we employ the modified Mann-Kendall test proposed by [[17]](https://doi.org/10.1023/B:WARM.0000043140.61082.60), which accounts for this autocorrelation.\n\nFurthermore, we convert the trends, initially in cm² s⁻² month⁻¹ to cm² s⁻² year⁻¹ using a factor to facilitate the comparison of the results.\n\nFunction to compute trends (Mann-Kendall test)\nMultply by the factor 12 months to convert trends per month to trends per year\n\nIn this case, we calculate the trends over different periods of the EKE time series, starting each trend in a different year between 1993 and 2013 (inclusive) and ending in the last year of the study (2022). The MK test is applied to each EKE time series for every region and version. As before, we create new dictionaries to store all trends and p-values.\n\nYears for the test\nUpdated dictionary keys\nLists to save results for high EKE regions, global ocean, Kuroshio Extension (KExt) and Gulf Stream (GS)\nGlobal ocean\nHigh EKE regions\nKuroshio Extension (KExt)\n\n```text\n1993\n\n----------------\nno ice\n----------------\n----------------\nhigh EKE\n----------------\n----------------\nKE\n----------------\n----------------\nGS\n----------------\n\n1994\n\n----------------\nno ice\n----------------\n----------------\nhigh EKE\n----------------\n----------------\nKE\n----------------\n----------------\nGS\n----------------\n\n1995\n\n----------------\nno ice\n----------------\n----------------\nhigh EKE\n----------------\n----------------\nKE\n----------------\n----------------\nGS\n----------------\n\n1996\n\n----------------\nno ice\n----------------\n----------------\nhigh EKE\n----------------\n----------------\nKE\n----------------\n----------------\nGS\n----------------\n\n1997\n\n----------------\nno ice\n----------------\n----------------\nhigh EKE\n----------------\n----------------\nKE\n----------------\n----------------\nGS\n----------------\n\n1998\n\n----------------\nno ice\n----------------\n----------------\nhigh EKE\n----------------\n----------------\nKE\n----------------\n----------------\nGS\n----------------\n\n1999\n\n----------------\nno ice\n----------------\n----------------\nhigh EKE\n----------------\n----------------\nKE\n----------------\n----------------\nGS\n----------------\n\n2000\n\n----------------\nno ice\n----------------\n----------------\nhigh EKE\n----------------\n----------------\nKE\n----------------\n----------------\nGS\n----------------\n\n2001\n\n----------------\nno ice\n----------------\n----------------\nhigh EKE\n----------------\n----------------\nKE\n----------------\n----------------\nGS\n----------------\n\n2002\n\n----------------\nno ice\n----------------\n----------------\nhigh EKE\n----------------\n----------------\nKE\n----------------\n----------------\nGS\n----------------\n\n2003\n\n----------------\nno ice\n----------------\n----------------\nhigh EKE\n----------------\n----------------\nKE\n----------------\n----------------\nGS\n----------------\n\n2004\n\n----------------\nno ice\n----------------\n----------------\nhigh EKE\n----------------\n----------------\nKE\n----------------\n----------------\nGS\n----------------\n\n2005\n\n----------------\nno ice\n----------------\n----------------\nhigh EKE\n----------------\n----------------\nKE\n----------------\n----------------\nGS\n----------------\n\n2006\n\n----------------\nno ice\n----------------\n----------------\nhigh EKE\n----------------\n----------------\nKE\n----------------\n----------------\nGS\n----------------\n\n2007\n\n----------------\nno ice\n----------------\n----------------\nhigh EKE\n----------------\n----------------\nKE\n----------------\n----------------\nGS\n----------------\n\n2008\n\n----------------\nno ice\n----------------\n----------------\nhigh EKE\n----------------\n----------------\nKE\n----------------\n----------------\nGS\n----------------\n\n2009\n\n----------------\nno ice\n----------------\n----------------\nhigh EKE\n----------------\n----------------\nKE\n----------------\n----------------\nGS\n----------------\n\n2010"} {"chunk_id": "satellite_satellite-sea-level-global_consistency_q01__cfbdfebb1b2c", "report_id": "satellite_satellite-sea-level-global_consistency_q01", "dataset_id": "satellite-sea-level-global", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q01", "aspect_base": "consistency", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Long-term global Eddy Kinetic Energy trends from satellite observations > Analysis and results > 4. EKE time series and trends computation", "title": "Long-term global Eddy Kinetic Energy trends from satellite observations", "chunk_index": 9, "token_count": 250, "text_raw": "high EKE\n----------------\n----------------\nKE\n----------------\n----------------\nGS\n----------------\n\n2009\n\n----------------\nno ice\n----------------\n----------------\nhigh EKE\n----------------\n----------------\nKE\n----------------\n----------------\nGS\n----------------\n\n2010\n\n----------------\nno ice\n----------------\n----------------\nhigh EKE\n----------------\n----------------\nKE\n----------------\n----------------\nGS\n----------------\n\n2011\n\n----------------\nno ice\n----------------\n----------------\nhigh EKE\n----------------\n----------------\nKE\n----------------\n----------------\nGS\n----------------\n\n2012\n\n----------------\nno ice\n----------------\n----------------\nhigh EKE\n----------------\n----------------\nKE\n----------------\n----------------\nGS\n----------------\n\n2013\n\n----------------\nno ice\n----------------\n----------------\nhigh EKE\n----------------\n----------------\nKE\n----------------\n----------------\nGS\n----------------\n```\n\n(section-5)=", "text_with_prefix": "EQC Quality Assessment: \"Long-term global Eddy Kinetic Energy trends from satellite observations\"\nDataset: satellite-sea-level-global [CDS]\nAspect: consistency_q01 | Category: Satellite_ECVs\nSection: Long-term global Eddy Kinetic Energy trends from satellite observations > Analysis and results > 4. EKE time series and trends computation\n---\nhigh EKE\n----------------\n----------------\nKE\n----------------\n----------------\nGS\n----------------\n\n2009\n\n----------------\nno ice\n----------------\n----------------\nhigh EKE\n----------------\n----------------\nKE\n----------------\n----------------\nGS\n----------------\n\n2010\n\n----------------\nno ice\n----------------\n----------------\nhigh EKE\n----------------\n----------------\nKE\n----------------\n----------------\nGS\n----------------\n\n2011\n\n----------------\nno ice\n----------------\n----------------\nhigh EKE\n----------------\n----------------\nKE\n----------------\n----------------\nGS\n----------------\n\n2012\n\n----------------\nno ice\n----------------\n----------------\nhigh EKE\n----------------\n----------------\nKE\n----------------\n----------------\nGS\n----------------\n\n2013\n\n----------------\nno ice\n----------------\n----------------\nhigh EKE\n----------------\n----------------\nKE\n----------------\n----------------\nGS\n----------------\n```\n\n(section-5)="} {"chunk_id": "satellite_satellite-sea-level-global_consistency_q01__9ad283767b0c", "report_id": "satellite_satellite-sea-level-global_consistency_q01", "dataset_id": "satellite-sea-level-global", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q01", "aspect_base": "consistency", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Long-term global Eddy Kinetic Energy trends from satellite observations > Analysis and results > 5. Plotting and discussion of results > High EKE regions", "title": "Long-term global Eddy Kinetic Energy trends from satellite observations", "chunk_index": 10, "token_count": 145, "text_raw": "Here, we prepare the high EKE and tropical masks as binary masks for proper visualisation. For the tropical mask, we apply \"add_cyclic_point\" to avoid discontinuity at 180° longitude. We also exclude these masks when graphing to represent only the individual high EKE regions.\n\nDefine high EKE regions", "text_with_prefix": "EQC Quality Assessment: \"Long-term global Eddy Kinetic Energy trends from satellite observations\"\nDataset: satellite-sea-level-global [CDS]\nAspect: consistency_q01 | Category: Satellite_ECVs\nSection: Long-term global Eddy Kinetic Energy trends from satellite observations > Analysis and results > 5. Plotting and discussion of results > High EKE regions\n---\nHere, we prepare the high EKE and tropical masks as binary masks for proper visualisation. For the tropical mask, we apply \"add_cyclic_point\" to avoid discontinuity at 180° longitude. We also exclude these masks when graphing to represent only the individual high EKE regions.\n\nDefine high EKE regions"} {"chunk_id": "satellite_satellite-sea-level-global_consistency_q01__d349db886803", "report_id": "satellite_satellite-sea-level-global_consistency_q01", "dataset_id": "satellite-sea-level-global", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q01", "aspect_base": "consistency", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Define tropical region (reindexing to avoid discontinuity at 180° when plotting)", "title": "Long-term global Eddy Kinetic Energy trends from satellite observations", "chunk_index": 11, "token_count": 290, "text_raw": "Define a mask with only the individual high EKE regions\n\nOnce the masks have been treated, we display the corresponding graphs in a layout of two subplots.\n\nGeneral parameters\nMean eke (Subplot A)\nTropical band\nHigh EKE regions\nMap mean EKE\nHigh EKE regions (Subplot B)\n\nThis figure shows a global map of the temporal mean EKE at each grid point over 1993-2022, using only the vDT2024 dataset. The green contours delimit the regions with high EKE activity (over the 90th percentile), and the blue line marks the tropical band. These areas contain most of the mesoscale activity, which averages zero in the remainder part of the globe.\n\nIn the other global map, we represent the exact high EKE regions (covering 5% of the global ocean) along with their corresponding acronyms. As a first observation, the largest regions (in terms of area) with high mesoscale activity occur in AC, KExt, and GS. In contrast, LC and EAC are the smallest regions.", "text_with_prefix": "EQC Quality Assessment: \"Long-term global Eddy Kinetic Energy trends from satellite observations\"\nDataset: satellite-sea-level-global [CDS]\nAspect: consistency_q01 | Category: Satellite_ECVs\nSection: Define tropical region (reindexing to avoid discontinuity at 180° when plotting)\n---\nDefine a mask with only the individual high EKE regions\n\nOnce the masks have been treated, we display the corresponding graphs in a layout of two subplots.\n\nGeneral parameters\nMean eke (Subplot A)\nTropical band\nHigh EKE regions\nMap mean EKE\nHigh EKE regions (Subplot B)\n\nThis figure shows a global map of the temporal mean EKE at each grid point over 1993-2022, using only the vDT2024 dataset. The green contours delimit the regions with high EKE activity (over the 90th percentile), and the blue line marks the tropical band. These areas contain most of the mesoscale activity, which averages zero in the remainder part of the globe.\n\nIn the other global map, we represent the exact high EKE regions (covering 5% of the global ocean) along with their corresponding acronyms. As a first observation, the largest regions (in terms of area) with high mesoscale activity occur in AC, KExt, and GS. In contrast, LC and EAC are the smallest regions."} {"chunk_id": "satellite_satellite-sea-level-global_consistency_q01__d39d13f7c19e", "report_id": "satellite_satellite-sea-level-global_consistency_q01", "dataset_id": "satellite-sea-level-global", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q01", "aspect_base": "consistency", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Define tropical region (reindexing to avoid discontinuity at 180° when plotting) > Area-weighted mean EKE time series", "title": "Long-term global Eddy Kinetic Energy trends from satellite observations", "chunk_index": 12, "token_count": 579, "text_raw": "In this section, we plot the EKE time series (in cm² s⁻²) over the global ocean and the high EKE regions, including the Kuroshio Extension and the Gulf Stream, covering the altimeter period 1993 to 2022.\n\nGeneral parameters\n\n##### Global region\n\nGlobal ocean and high EKE\n\nThe two subplots show the evolution of EKE in the global and high EKE regions. The color blue represents the vDT2024 version and the color red the vDT2021 version. The thinner lines show the raw (untreated) data, whilst the thicker lines indicate the data after applying an annual rolling mean. Regarding the results, notable differences are observed when comparing the two versions. For both the global ocean and high EKE regions, vDT2021 exhibits higher values throughout the entire period than vDT2024. One possible reason is the methodology used in producing each version, including the metrics, procedures, and equations implemented, resulting in more accurate data in the current version (vDT2024), as mentioned in the [Algorithm Theoretical Basis Document](https://dast.copernicus-climate.eu/documents/satellite-sea-level/vDT2024/C3S2-D312b-WP2-FDDP-SL-v3.0-202411-ATBD-of-vDT2024-v1_Final.pdf). For the global region, the rolling mean values are around 190-250 cm² s⁻² approximately. However, focusing on the high EKE regions, the values increase markedly, reaching up to 1150 cm² s⁻². These regions, as observed, demonstrate high mesoscale activity.\n\n##### Mesoscale High-Activity Regions\n\nKuroshio Extension and Gulf Stream\n\nFollowing [[5]](https://doi.org/10.1038/s41598-025-06149-9), two specific areas were selected and compared: the KExt and the GS. Similarly to the previous plot, we have two timeseries for each version (red for the vDT2021 and blue for the vDT2024), with the older version showing higher energy than the newer one. The difference is larger in the GS than in the KExt. When we compare the energy in both areas, both versions point towards an increase in EKE in the KExt. Both versions also capture the variability in the timeseries, with higher seasonal variability in the GS.", "text_with_prefix": "EQC Quality Assessment: \"Long-term global Eddy Kinetic Energy trends from satellite observations\"\nDataset: satellite-sea-level-global [CDS]\nAspect: consistency_q01 | Category: Satellite_ECVs\nSection: Define tropical region (reindexing to avoid discontinuity at 180° when plotting) > Area-weighted mean EKE time series\n---\nIn this section, we plot the EKE time series (in cm² s⁻²) over the global ocean and the high EKE regions, including the Kuroshio Extension and the Gulf Stream, covering the altimeter period 1993 to 2022.\n\nGeneral parameters\n\n##### Global region\n\nGlobal ocean and high EKE\n\nThe two subplots show the evolution of EKE in the global and high EKE regions. The color blue represents the vDT2024 version and the color red the vDT2021 version. The thinner lines show the raw (untreated) data, whilst the thicker lines indicate the data after applying an annual rolling mean. Regarding the results, notable differences are observed when comparing the two versions. For both the global ocean and high EKE regions, vDT2021 exhibits higher values throughout the entire period than vDT2024. One possible reason is the methodology used in producing each version, including the metrics, procedures, and equations implemented, resulting in more accurate data in the current version (vDT2024), as mentioned in the [Algorithm Theoretical Basis Document](https://dast.copernicus-climate.eu/documents/satellite-sea-level/vDT2024/C3S2-D312b-WP2-FDDP-SL-v3.0-202411-ATBD-of-vDT2024-v1_Final.pdf). For the global region, the rolling mean values are around 190-250 cm² s⁻² approximately. However, focusing on the high EKE regions, the values increase markedly, reaching up to 1150 cm² s⁻². These regions, as observed, demonstrate high mesoscale activity.\n\n##### Mesoscale High-Activity Regions\n\nKuroshio Extension and Gulf Stream\n\nFollowing [[5]](https://doi.org/10.1038/s41598-025-06149-9), two specific areas were selected and compared: the KExt and the GS. Similarly to the previous plot, we have two timeseries for each version (red for the vDT2021 and blue for the vDT2024), with the older version showing higher energy than the newer one. The difference is larger in the GS than in the KExt. When we compare the energy in both areas, both versions point towards an increase in EKE in the KExt. Both versions also capture the variability in the timeseries, with higher seasonal variability in the GS."} {"chunk_id": "satellite_satellite-sea-level-global_consistency_q01__046b99e85592", "report_id": "satellite_satellite-sea-level-global_consistency_q01", "dataset_id": "satellite-sea-level-global", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q01", "aspect_base": "consistency", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Define tropical region (reindexing to avoid discontinuity at 180° when plotting) > EKE trends", "title": "Long-term global Eddy Kinetic Energy trends from satellite observations", "chunk_index": 13, "token_count": 964, "text_raw": "For these plots, we first defined a specific marker and text style to distinguish non-significant trends in the bar plots. Non-significant trends are shown in italics, with oblique grey hatching.\n\nMarker (hatching) and text style (italic) for non-significant trends\nGlobal ocean\nHigh EKE regions\nKuroshio Extension (KExt)\n\n##### Global region\n\nGlobal ocean\nHigh EKE regions\n\nIn this case, we can observe the EKE trends over different periods, progressively shortening the intervals to study the trends. For the global ocean, the trends are larger over shorter timescales, reaching a maximum trend of 0.87 cm² s⁻² year⁻¹ (vDT2024) and 1.14 cm² s⁻² year⁻¹ (vDT2021) from 2012 to 2022. Looking in more detail, the initial trends computed with a starting year prior to 2000 show a slight decrease, resulting in non-significant trends for both datasets and even negative trends in the case of vDT2024, for which, if we consider the whole period, the EKE trend is not significant (whereas it is for vDT2021). This behaviour changes drastically after 2000, with a constant increase, revealing an intensification of the mesoscale activity globally.\n\nFor the high EKE regions, we can clearly see a substantial increase throughout the full period for both datasets. In this case, all trends are significant, although vDT2021 displays larger values than vDT2024. For the older version, the trend over 2012-2022 is 6.1 times larger than the trend during 1993-2022, increasing from 2.58 cm² s⁻² year⁻¹ to 15.75 cm² s⁻² year⁻¹. However, for the latest version, the trend is 6.5 times larger, from 2.29 cm² s⁻² year⁻¹ to 14.89 cm² s⁻² year⁻¹. This suggests that over the last two decades, EKE activity has become stronger and more energetic, especially in the high EKE regions, with a particularly high rate of increase in the most recent periods.\n\nIn [[5]](https://doi.org/10.1038/s41598-025-06149-9), the vDT2021 all-sat product from CMEMS is used to compare with the vDT2021 two-sat product from C3S. That study suggests that differences in trends between the two products are partly related to the number of satellites included in the all-sat product.\n\n##### Mesoscale High-Activity Regions\n\nKuroshio Extension\n\nHere, we focus on the regions where the mesoscale activity exhibits more significant positive EKE trends. This intensification is concentrated mainly in the Kuroshio Extension (KExt) and the Gulf Stream (GS). Regarding the KExt region, the computed trends show a progressive increase for both datasets. The vDT2021 presents higher values than vDT2024, but these differences are small compared with the other regions. Only the trend over 1999-2022 is not statistically significant for either version. As mentioned before, this fast increase occurs primarily in the last decade. For the period 2013-2022, the EKE trend for vDT2024 is 61.65 cm² s⁻² year⁻¹ and for vDT2021 is 61.99 cm² s⁻² year⁻¹, while over the entire 30-year period the trends are 8.23 cm² s⁻² year⁻¹ and 8.69 cm² s⁻² year⁻¹, respectively. This region is crucial for ocean circulation and mid-latitudes air-sea interactions and is one of the major western boundary currents [[18]](https://doi.org/10.1016/j.dynatmoce.2024.101497). As reported by [[19]](https://doi.org/10.1007/s10236-023-01565-9), the decadal variability of EKE in this region is linked to the Pacific Decadal Oscillation (PDO).", "text_with_prefix": "EQC Quality Assessment: \"Long-term global Eddy Kinetic Energy trends from satellite observations\"\nDataset: satellite-sea-level-global [CDS]\nAspect: consistency_q01 | Category: Satellite_ECVs\nSection: Define tropical region (reindexing to avoid discontinuity at 180° when plotting) > EKE trends\n---\nFor these plots, we first defined a specific marker and text style to distinguish non-significant trends in the bar plots. Non-significant trends are shown in italics, with oblique grey hatching.\n\nMarker (hatching) and text style (italic) for non-significant trends\nGlobal ocean\nHigh EKE regions\nKuroshio Extension (KExt)\n\n##### Global region\n\nGlobal ocean\nHigh EKE regions\n\nIn this case, we can observe the EKE trends over different periods, progressively shortening the intervals to study the trends. For the global ocean, the trends are larger over shorter timescales, reaching a maximum trend of 0.87 cm² s⁻² year⁻¹ (vDT2024) and 1.14 cm² s⁻² year⁻¹ (vDT2021) from 2012 to 2022. Looking in more detail, the initial trends computed with a starting year prior to 2000 show a slight decrease, resulting in non-significant trends for both datasets and even negative trends in the case of vDT2024, for which, if we consider the whole period, the EKE trend is not significant (whereas it is for vDT2021). This behaviour changes drastically after 2000, with a constant increase, revealing an intensification of the mesoscale activity globally.\n\nFor the high EKE regions, we can clearly see a substantial increase throughout the full period for both datasets. In this case, all trends are significant, although vDT2021 displays larger values than vDT2024. For the older version, the trend over 2012-2022 is 6.1 times larger than the trend during 1993-2022, increasing from 2.58 cm² s⁻² year⁻¹ to 15.75 cm² s⁻² year⁻¹. However, for the latest version, the trend is 6.5 times larger, from 2.29 cm² s⁻² year⁻¹ to 14.89 cm² s⁻² year⁻¹. This suggests that over the last two decades, EKE activity has become stronger and more energetic, especially in the high EKE regions, with a particularly high rate of increase in the most recent periods.\n\nIn [[5]](https://doi.org/10.1038/s41598-025-06149-9), the vDT2021 all-sat product from CMEMS is used to compare with the vDT2021 two-sat product from C3S. That study suggests that differences in trends between the two products are partly related to the number of satellites included in the all-sat product.\n\n##### Mesoscale High-Activity Regions\n\nKuroshio Extension\n\nHere, we focus on the regions where the mesoscale activity exhibits more significant positive EKE trends. This intensification is concentrated mainly in the Kuroshio Extension (KExt) and the Gulf Stream (GS). Regarding the KExt region, the computed trends show a progressive increase for both datasets. The vDT2021 presents higher values than vDT2024, but these differences are small compared with the other regions. Only the trend over 1999-2022 is not statistically significant for either version. As mentioned before, this fast increase occurs primarily in the last decade. For the period 2013-2022, the EKE trend for vDT2024 is 61.65 cm² s⁻² year⁻¹ and for vDT2021 is 61.99 cm² s⁻² year⁻¹, while over the entire 30-year period the trends are 8.23 cm² s⁻² year⁻¹ and 8.69 cm² s⁻² year⁻¹, respectively. This region is crucial for ocean circulation and mid-latitudes air-sea interactions and is one of the major western boundary currents [[18]](https://doi.org/10.1016/j.dynatmoce.2024.101497). As reported by [[19]](https://doi.org/10.1007/s10236-023-01565-9), the decadal variability of EKE in this region is linked to the Pacific Decadal Oscillation (PDO)."} {"chunk_id": "satellite_satellite-sea-level-global_consistency_q01__1b228600c9e4", "report_id": "satellite_satellite-sea-level-global_consistency_q01", "dataset_id": "satellite-sea-level-global", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q01", "aspect_base": "consistency", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Define tropical region (reindexing to avoid discontinuity at 180° when plotting) > EKE trends", "title": "Long-term global Eddy Kinetic Energy trends from satellite observations", "chunk_index": 14, "token_count": 351, "text_raw": "]](https://doi.org/10.1007/s10236-023-01565-9), the decadal variability of EKE in this region is linked to the Pacific Decadal Oscillation (PDO).\n\nIn the Gulf Stream (GS), the general trend shows a gradual increase. Over the 30 years, the EKE trends are 2.12 cm² s⁻² year⁻¹ for vDT2024 and 2.51 cm² s⁻² year⁻¹ for vDT2021, both of which are statistically non-significant. In the most recent decade, the trends have increased notably, reaching 18.09 cm² s⁻² year⁻¹ and 20.33 cm² s⁻² year⁻¹, respectively. This corresponds to an increase in the EKE trend of 8.5 and 8.1 times faster than over the entire 30-year period. The GS plays an important role due to its contribution to the Atlantic Meridional Overturning Circulation (AMOC) [[20]](https://doi.org/10.1002/2015RG000493). [[21]](https://doi.org/10.1175/2010JCLI3310.1) reported that the path of the GS is associated with the strength of the AMOC.", "text_with_prefix": "EQC Quality Assessment: \"Long-term global Eddy Kinetic Energy trends from satellite observations\"\nDataset: satellite-sea-level-global [CDS]\nAspect: consistency_q01 | Category: Satellite_ECVs\nSection: Define tropical region (reindexing to avoid discontinuity at 180° when plotting) > EKE trends\n---\n]](https://doi.org/10.1007/s10236-023-01565-9), the decadal variability of EKE in this region is linked to the Pacific Decadal Oscillation (PDO).\n\nIn the Gulf Stream (GS), the general trend shows a gradual increase. Over the 30 years, the EKE trends are 2.12 cm² s⁻² year⁻¹ for vDT2024 and 2.51 cm² s⁻² year⁻¹ for vDT2021, both of which are statistically non-significant. In the most recent decade, the trends have increased notably, reaching 18.09 cm² s⁻² year⁻¹ and 20.33 cm² s⁻² year⁻¹, respectively. This corresponds to an increase in the EKE trend of 8.5 and 8.1 times faster than over the entire 30-year period. The GS plays an important role due to its contribution to the Atlantic Meridional Overturning Circulation (AMOC) [[20]](https://doi.org/10.1002/2015RG000493). [[21]](https://doi.org/10.1175/2010JCLI3310.1) reported that the path of the GS is associated with the strength of the AMOC."} {"chunk_id": "satellite_satellite-sea-level-global_consistency_q01__97cfd8137562", "report_id": "satellite_satellite-sea-level-global_consistency_q01", "dataset_id": "satellite-sea-level-global", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q01", "aspect_base": "consistency", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Define tropical region (reindexing to avoid discontinuity at 180° when plotting) > ℹ️ If you want to know more > Key resources", "title": "Long-term global Eddy Kinetic Energy trends from satellite observations", "chunk_index": 15, "token_count": 361, "text_raw": "* [[5]](https://doi.org/10.1038/s41598-025-06149-9) Barceló-Llull, B., Rosselló, P., Combes, V., Sánchez-Román, A., Pujol, M. I., and Pascual, A. (2025). Kuroshio Extension and Gulf Stream dominate the Eddy Kinetic Energy intensification observed in the global ocean. Scientific Reports, 15(1), 21754. doi: 10.1038/s41598-025-06149-9 \n* [CDS entry](https://cds.climate.copernicus.eu/datasets/satellite-sea-level-global?tab=overview): Sea level gridded data from satellite observations for the global ocean from 1993 to present. \n* All the documentation for each version is available [here](https://cds.climate.copernicus.eu/datasets/satellite-sea-level-global?tab=documentation).\n\nFurther information: \n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/).", "text_with_prefix": "EQC Quality Assessment: \"Long-term global Eddy Kinetic Energy trends from satellite observations\"\nDataset: satellite-sea-level-global [CDS]\nAspect: consistency_q01 | Category: Satellite_ECVs\nSection: Define tropical region (reindexing to avoid discontinuity at 180° when plotting) > ℹ️ If you want to know more > Key resources\n---\n* [[5]](https://doi.org/10.1038/s41598-025-06149-9) Barceló-Llull, B., Rosselló, P., Combes, V., Sánchez-Román, A., Pujol, M. I., and Pascual, A. (2025). Kuroshio Extension and Gulf Stream dominate the Eddy Kinetic Energy intensification observed in the global ocean. Scientific Reports, 15(1), 21754. doi: 10.1038/s41598-025-06149-9 \n* [CDS entry](https://cds.climate.copernicus.eu/datasets/satellite-sea-level-global?tab=overview): Sea level gridded data from satellite observations for the global ocean from 1993 to present. \n* All the documentation for each version is available [here](https://cds.climate.copernicus.eu/datasets/satellite-sea-level-global?tab=documentation).\n\nFurther information: \n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)."} {"chunk_id": "satellite_satellite-sea-level-global_consistency_q01__f3510283fa56", "report_id": "satellite_satellite-sea-level-global_consistency_q01", "dataset_id": "satellite-sea-level-global", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q01", "aspect_base": "consistency", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Define tropical region (reindexing to avoid discontinuity at 180° when plotting) > ℹ️ If you want to know more > References", "title": "Long-term global Eddy Kinetic Energy trends from satellite observations", "chunk_index": 16, "token_count": 1047, "text_raw": "[[1]](https://doi.org/10.1146/annurev.fluid.40.111406.102139) Ferrari, R., & Wunsch, C. (2009). Ocean circulation kinetic energy: Reservoirs, sources, and sinks. Annual Review of Fluid Mechanics, 41(1), 253-282. doi: 10.1146/annurev.fluid.40.111406.102139\n\n[[2]](https://doi.org/10.1126/sciadv.aax7727) Hu, S., Sprintall, J., Guan, C., McPhaden, M. J., Wang, F., Hu, D., & Cai, W. (2020). Deep-reaching acceleration of global mean ocean circulation over the past two decades. Science advances, 6(6), eaax7727. doi: 10.1126/sciadv.aax7727\n\n[[3]](https://doi.org/10.1175/1520-0485%281998%29028<0433:GVOTFB>2.0.CO;2) Chelton, D. B., DeSzoeke, R. A., Schlax, M. G., El Naggar, K., & Siwertz, N. (1998). Geographical variability of the first baroclinic Rossby radius of deformation. Journal of Physical Oceanography, 28(3), 433-460. doi: 10.1175/1520-0485(1998)028<0433:GVOTFB>2.0.CO;2\n\n[[4]](https://doi.org/10.1029/2005GL024633) Pascual, A., Faugère, Y., Larnicol, G., & Le Traon, P. Y. (2006). Improved description of the ocean mesoscale variability by combining four satellite altimeters. Geophysical Research Letters, 33(2). doi: 10.1029/2005GL024633\n\n[[5]](https://doi.org/10.1038/s41598-025-06149-9) Barceló-Llull, B., Rosselló, P., Combes, V., Sánchez-Román, A., Pujol, M. I., & Pascual, A. (2025). Kuroshio Extension and Gulf Stream dominate the Eddy Kinetic Energy intensification observed in the global ocean. Scientific Reports, 15(1), 21754. doi: 10.1038/s41598-025-06149-9\n\n[[6]](https://doi.org/10.1029/2007GL030812) Chelton, D. B., Schlax, M. G., Samelson, R. M., & de Szoeke, R. A. (2007). Global observations of large oceanic eddies. Geophysical Research Letters, 34(15). doi: 10.1029/2007GL030812\n\n[[7]](https://doi.org/10.1175/JTECH-D-17-0010.1) Le Vu, B., Stegner, A., & Arsouze, T. (2018). Angular momentum eddy detection and tracking algorithm (AMEDA) and its application to coastal eddy formation. Journal of Atmospheric and Oceanic Technology, 35(4), 739-762. doi: 10.1175/JTECH-D-17-0010.1\n\n[[8]](https://doi.org/10.3389/fmars.2019.00703) Amores, A., Jordà, G., & Monserrat, S. (2019). Ocean eddies in the Mediterranean Sea from satellite altimetry: Sensitivity to satellite track location. Frontiers in Marine Science, 6, 703. doi: 10.3389/fmars.2019.00703\n\n[[9]](https://doi.org/10.1016/j.asr.2021.01.022) Abdalla, S., Kolahchi, A. A., Ablain, M., Adusumilli, S., Bhowmick, S. A., Alou-Font, E., ... & Hamon, M. (2021). Altimetry for the future: Building on 25 years of progress. Advances in Space Research, 68(2), 319-363. doi: 10.1016/j.asr.2021.01.022", "text_with_prefix": "EQC Quality Assessment: \"Long-term global Eddy Kinetic Energy trends from satellite observations\"\nDataset: satellite-sea-level-global [CDS]\nAspect: consistency_q01 | Category: Satellite_ECVs\nSection: Define tropical region (reindexing to avoid discontinuity at 180° when plotting) > ℹ️ If you want to know more > References\n---\n[[1]](https://doi.org/10.1146/annurev.fluid.40.111406.102139) Ferrari, R., & Wunsch, C. (2009). Ocean circulation kinetic energy: Reservoirs, sources, and sinks. Annual Review of Fluid Mechanics, 41(1), 253-282. doi: 10.1146/annurev.fluid.40.111406.102139\n\n[[2]](https://doi.org/10.1126/sciadv.aax7727) Hu, S., Sprintall, J., Guan, C., McPhaden, M. J., Wang, F., Hu, D., & Cai, W. (2020). Deep-reaching acceleration of global mean ocean circulation over the past two decades. Science advances, 6(6), eaax7727. doi: 10.1126/sciadv.aax7727\n\n[[3]](https://doi.org/10.1175/1520-0485%281998%29028<0433:GVOTFB>2.0.CO;2) Chelton, D. B., DeSzoeke, R. A., Schlax, M. G., El Naggar, K., & Siwertz, N. (1998). Geographical variability of the first baroclinic Rossby radius of deformation. Journal of Physical Oceanography, 28(3), 433-460. doi: 10.1175/1520-0485(1998)028<0433:GVOTFB>2.0.CO;2\n\n[[4]](https://doi.org/10.1029/2005GL024633) Pascual, A., Faugère, Y., Larnicol, G., & Le Traon, P. Y. (2006). Improved description of the ocean mesoscale variability by combining four satellite altimeters. Geophysical Research Letters, 33(2). doi: 10.1029/2005GL024633\n\n[[5]](https://doi.org/10.1038/s41598-025-06149-9) Barceló-Llull, B., Rosselló, P., Combes, V., Sánchez-Román, A., Pujol, M. I., & Pascual, A. (2025). Kuroshio Extension and Gulf Stream dominate the Eddy Kinetic Energy intensification observed in the global ocean. Scientific Reports, 15(1), 21754. doi: 10.1038/s41598-025-06149-9\n\n[[6]](https://doi.org/10.1029/2007GL030812) Chelton, D. B., Schlax, M. G., Samelson, R. M., & de Szoeke, R. A. (2007). Global observations of large oceanic eddies. Geophysical Research Letters, 34(15). doi: 10.1029/2007GL030812\n\n[[7]](https://doi.org/10.1175/JTECH-D-17-0010.1) Le Vu, B., Stegner, A., & Arsouze, T. (2018). Angular momentum eddy detection and tracking algorithm (AMEDA) and its application to coastal eddy formation. Journal of Atmospheric and Oceanic Technology, 35(4), 739-762. doi: 10.1175/JTECH-D-17-0010.1\n\n[[8]](https://doi.org/10.3389/fmars.2019.00703) Amores, A., Jordà, G., & Monserrat, S. (2019). Ocean eddies in the Mediterranean Sea from satellite altimetry: Sensitivity to satellite track location. Frontiers in Marine Science, 6, 703. doi: 10.3389/fmars.2019.00703\n\n[[9]](https://doi.org/10.1016/j.asr.2021.01.022) Abdalla, S., Kolahchi, A. A., Ablain, M., Adusumilli, S., Bhowmick, S. A., Alou-Font, E., ... & Hamon, M. (2021). Altimetry for the future: Building on 25 years of progress. Advances in Space Research, 68(2), 319-363. doi: 10.1016/j.asr.2021.01.022"} {"chunk_id": "satellite_satellite-sea-level-global_consistency_q01__fe5eecefb753", "report_id": "satellite_satellite-sea-level-global_consistency_q01", "dataset_id": "satellite-sea-level-global", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q01", "aspect_base": "consistency", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Define tropical region (reindexing to avoid discontinuity at 180° when plotting) > ℹ️ If you want to know more > References", "title": "Long-term global Eddy Kinetic Energy trends from satellite observations", "chunk_index": 17, "token_count": 1089, "text_raw": "). Altimetry for the future: Building on 25 years of progress. Advances in Space Research, 68(2), 319-363. doi: 10.1016/j.asr.2021.01.022\n\n[[10]](https://doi.org/10.1016/j.asr.2011.09.033) Morrow, R., & Le Traon, P. Y. (2012). Recent advances in observing mesoscale ocean dynamics with satellite altimetry. Advances in Space Research, 50(8), 1062-1076. doi: 10.1016/j.asr.2011.09.033\n\n[[11]](https://doi.org/10.5194/egusphere-2025-4651) Hargous, P., Combes, V., Barceló-Llull, B., & Pascual, A. (2025). Eddy Kinetic Energy Variability From 30 Years of Altimetry in the Mediterranean Sea. EGUsphere, 2025, 1-22. doi: 10.5194/egusphere-2025-4651\n\n[[12]](https://doi.org/10.1038/s41558-021-01006-9) Martínez-Moreno, J., Hogg, A. M., England, M. H., Constantinou, N. C., Kiss, A. E., & Morrison, A. K. (2021). Global changes in oceanic mesoscale currents over the satellite altimetry record. Nature Climate Change, 11(5), 397-403. doi: 10.1038/s41558-021-01006-9\n\n[[13]](https://doi.org/10.1007/s10712-023-09778-9) Morrow, R., Fu, L. L., Rio, M. H., Ray, R., Prandi, P., Le Traon, P. Y., & Benveniste, J. (2023). Ocean circulation from space. Surveys in Geophysics, 44(5), 1243-1286. doi: 10.1007/s10712-023-09778-9\n\n[[14]](https://doi.org/10.1038/NCLIMATE2635) Watson, C. S., White, N. J., Church, J. A., King, M. A., Burgette, R. J., & Legresy, B. (2015). Unabated global mean sea-level rise over the satellite altimeter era. Nature Climate Change, 5(6), 565-568. doi: 10.1038/NCLIMATE2635\n\n[[15]](https://doi.org/10.1029/2011JC007367) Arbic, B. K., Scott, R. B., Chelton, D. B., Richman, J. G., & Shriver, J. F. (2012). Effects of stencil width on surface ocean geostrophic velocity and vorticity estimation from gridded satellite altimeter data. Journal of Geophysical Research: Oceans, 117(C3). doi: 10.1029/2011JC007367\n\n[[16]](https://doi.org/10.1029/1999JC900197) Lagerloef, G. S., Mitchum, G. T., Lukas, R. B., & Niiler, P. P. (1999). Tropical Pacific near‐surface currents estimated from altimeter, wind, and drifter data. Journal of Geophysical Research: Oceans, 104(C10), 23313-23326. doi: 10.1029/1999JC900197\n\n[[17]](https://doi.org/10.1023/B:WARM.0000043140.61082.60) Yue, S., & Wang, C. (2004). The Mann-Kendall test modified by effective sample size to detect trend in serially correlated hydrological series. Water resources management, 18(3), 201-218. doi: 10.1023/B:WARM.0000043140.61082.60\n\n[[18]](https://doi.org/10.1016/j.dynatmoce.2024.101497) Xiaodong, M., Lei, Z., Weishuai, X., Qinghong, L., & Maolin, L. (2024). Analysis and prediction of mesoscale eddy kinetic energy variations in the Kuroshio extension. Dynamics of Atmospheres and Oceans, 108, 101497. doi: 10.1016/j.dynatmoce.2024.101497", "text_with_prefix": "EQC Quality Assessment: \"Long-term global Eddy Kinetic Energy trends from satellite observations\"\nDataset: satellite-sea-level-global [CDS]\nAspect: consistency_q01 | Category: Satellite_ECVs\nSection: Define tropical region (reindexing to avoid discontinuity at 180° when plotting) > ℹ️ If you want to know more > References\n---\n). Altimetry for the future: Building on 25 years of progress. Advances in Space Research, 68(2), 319-363. doi: 10.1016/j.asr.2021.01.022\n\n[[10]](https://doi.org/10.1016/j.asr.2011.09.033) Morrow, R., & Le Traon, P. Y. (2012). Recent advances in observing mesoscale ocean dynamics with satellite altimetry. Advances in Space Research, 50(8), 1062-1076. doi: 10.1016/j.asr.2011.09.033\n\n[[11]](https://doi.org/10.5194/egusphere-2025-4651) Hargous, P., Combes, V., Barceló-Llull, B., & Pascual, A. (2025). Eddy Kinetic Energy Variability From 30 Years of Altimetry in the Mediterranean Sea. EGUsphere, 2025, 1-22. doi: 10.5194/egusphere-2025-4651\n\n[[12]](https://doi.org/10.1038/s41558-021-01006-9) Martínez-Moreno, J., Hogg, A. M., England, M. H., Constantinou, N. C., Kiss, A. E., & Morrison, A. K. (2021). Global changes in oceanic mesoscale currents over the satellite altimetry record. Nature Climate Change, 11(5), 397-403. doi: 10.1038/s41558-021-01006-9\n\n[[13]](https://doi.org/10.1007/s10712-023-09778-9) Morrow, R., Fu, L. L., Rio, M. H., Ray, R., Prandi, P., Le Traon, P. Y., & Benveniste, J. (2023). Ocean circulation from space. Surveys in Geophysics, 44(5), 1243-1286. doi: 10.1007/s10712-023-09778-9\n\n[[14]](https://doi.org/10.1038/NCLIMATE2635) Watson, C. S., White, N. J., Church, J. A., King, M. A., Burgette, R. J., & Legresy, B. (2015). Unabated global mean sea-level rise over the satellite altimeter era. Nature Climate Change, 5(6), 565-568. doi: 10.1038/NCLIMATE2635\n\n[[15]](https://doi.org/10.1029/2011JC007367) Arbic, B. K., Scott, R. B., Chelton, D. B., Richman, J. G., & Shriver, J. F. (2012). Effects of stencil width on surface ocean geostrophic velocity and vorticity estimation from gridded satellite altimeter data. Journal of Geophysical Research: Oceans, 117(C3). doi: 10.1029/2011JC007367\n\n[[16]](https://doi.org/10.1029/1999JC900197) Lagerloef, G. S., Mitchum, G. T., Lukas, R. B., & Niiler, P. P. (1999). Tropical Pacific near‐surface currents estimated from altimeter, wind, and drifter data. Journal of Geophysical Research: Oceans, 104(C10), 23313-23326. doi: 10.1029/1999JC900197\n\n[[17]](https://doi.org/10.1023/B:WARM.0000043140.61082.60) Yue, S., & Wang, C. (2004). The Mann-Kendall test modified by effective sample size to detect trend in serially correlated hydrological series. Water resources management, 18(3), 201-218. doi: 10.1023/B:WARM.0000043140.61082.60\n\n[[18]](https://doi.org/10.1016/j.dynatmoce.2024.101497) Xiaodong, M., Lei, Z., Weishuai, X., Qinghong, L., & Maolin, L. (2024). Analysis and prediction of mesoscale eddy kinetic energy variations in the Kuroshio extension. Dynamics of Atmospheres and Oceans, 108, 101497. doi: 10.1016/j.dynatmoce.2024.101497"} {"chunk_id": "satellite_satellite-sea-level-global_consistency_q01__dfc6fd0aba97", "report_id": "satellite_satellite-sea-level-global_consistency_q01", "dataset_id": "satellite-sea-level-global", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q01", "aspect_base": "consistency", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Define tropical region (reindexing to avoid discontinuity at 180° when plotting) > ℹ️ If you want to know more > References", "title": "Long-term global Eddy Kinetic Energy trends from satellite observations", "chunk_index": 18, "token_count": 406, "text_raw": "cale eddy kinetic energy variations in the Kuroshio extension. Dynamics of Atmospheres and Oceans, 108, 101497. doi: 10.1016/j.dynatmoce.2024.101497\n\n[[19]](https://doi.org/10.1007/s10236-023-01565-9) Yang, C., Yang, H., Chen, Z., Gan, B., Liu, Y., & Wu, L. (2023). Seasonal variability of eddy characteristics and energetics in the Kuroshio Extension. Ocean Dynamics, 73(8), 531-544. doi: 10.1007/s10236-023-01565-9\n\n[[20]](https://doi.org/10.1002/2015RG000493) Buckley, M. W., & Marshall, J. (2016). Observations, inferences, and mechanisms of the Atlantic Meridional Overturning Circulation: A review. Reviews of Geophysics, 54(1), 5-63. doi: 10.1002/2015RG000493\n\n[[21]](https://doi.org/10.1175/2010JCLI3310.1) Joyce, T. M., & Zhang, R. (2010). On the path of the Gulf Stream and the Atlantic meridional overturning circulation. Journal of Climate, 23(11), 3146-3154. doi: 10.1175/2010JCLI3310.1", "text_with_prefix": "EQC Quality Assessment: \"Long-term global Eddy Kinetic Energy trends from satellite observations\"\nDataset: satellite-sea-level-global [CDS]\nAspect: consistency_q01 | Category: Satellite_ECVs\nSection: Define tropical region (reindexing to avoid discontinuity at 180° when plotting) > ℹ️ If you want to know more > References\n---\ncale eddy kinetic energy variations in the Kuroshio extension. Dynamics of Atmospheres and Oceans, 108, 101497. doi: 10.1016/j.dynatmoce.2024.101497\n\n[[19]](https://doi.org/10.1007/s10236-023-01565-9) Yang, C., Yang, H., Chen, Z., Gan, B., Liu, Y., & Wu, L. (2023). Seasonal variability of eddy characteristics and energetics in the Kuroshio Extension. Ocean Dynamics, 73(8), 531-544. doi: 10.1007/s10236-023-01565-9\n\n[[20]](https://doi.org/10.1002/2015RG000493) Buckley, M. W., & Marshall, J. (2016). Observations, inferences, and mechanisms of the Atlantic Meridional Overturning Circulation: A review. Reviews of Geophysics, 54(1), 5-63. doi: 10.1002/2015RG000493\n\n[[21]](https://doi.org/10.1175/2010JCLI3310.1) Joyce, T. M., & Zhang, R. (2010). On the path of the Gulf Stream and the Atlantic meridional overturning circulation. Journal of Climate, 23(11), 3146-3154. doi: 10.1175/2010JCLI3310.1"} {"chunk_id": "satellite_satellite-sea-level-global_trend-assessment_q01__cb6a15d482ee", "report_id": "satellite_satellite-sea-level-global_trend-assessment_q01", "dataset_id": "satellite-sea-level-global", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q01", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Regional sea level trend assessment in the Mediterranean Sea from Satellite (observations) at basin and sub-basin scale", "title": "Regional sea level trend assessment in the Mediterranean Sea from Satellite (observations) at basin and sub-basin scale", "chunk_index": 0, "token_count": 122, "text_raw": "Production date: 27-08-2024\n\nProduced by: IMEDEA (CSIC-UIB) [[webpage]](https://imedea.uib-csic.es/)", "text_with_prefix": "EQC Quality Assessment: \"Regional sea level trend assessment in the Mediterranean Sea from Satellite (observations) at basin and sub-basin scale\"\nDataset: satellite-sea-level-global [CDS]\nAspect: trend-assessment_q01 | Category: Satellite_ECVs\nSection: Regional sea level trend assessment in the Mediterranean Sea from Satellite (observations) at basin and sub-basin scale\n---\nProduction date: 27-08-2024\n\nProduced by: IMEDEA (CSIC-UIB) [[webpage]](https://imedea.uib-csic.es/)"} {"chunk_id": "satellite_satellite-sea-level-global_trend-assessment_q01__91760d1139e1", "report_id": "satellite_satellite-sea-level-global_trend-assessment_q01", "dataset_id": "satellite-sea-level-global", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q01", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Regional sea level trend assessment in the Mediterranean Sea from Satellite (observations) at basin and sub-basin scale > Quality assessment question", "title": "Regional sea level trend assessment in the Mediterranean Sea from Satellite (observations) at basin and sub-basin scale", "chunk_index": 1, "token_count": 628, "text_raw": "* **Can we use satellite-derived sea level anomalies to detect sea level trend and its spatial variability over the Mediterranean Sea?**\n\nSea level variation is an important indicator of climate change, as it integrates changes of almost all the components in the climate system [[1]](https://doi.org/10.5194/essd-10-281-2018). It specially impacts coastal societies and environments [[2]](https://doi.org/10.3389/fmars.2023.1150488). However, sea level change at the global scale is not uniform varying on a regional to local scale from global mean rates (e.g.,[[3]](https://doi.org/10.1016/j.gloplacha.2013.12.004), [[4]](https://doi.org/10.5194/esd-5-243-2014)). For instance, sea level trends in the Mediterranean differ from global mean sea level trends, mainly because of its semi-enclosed conditions [[5]](https://doi.org/10.1175/JCLI-D-13-00139.1). In addition, within the Mediterranean, sea level trends can differ from the basin mean [[6]](https://doi.org/10.1007/s00382-016-3001-2) due to non-linear local oceanographic processes [[7]](https://doi.org/10.1016/j.gloplacha.2009.04.002). The Mediterranean Sea has been classified as one of the most susceptible climate change zones worldwide [[8]](https://doi.org/10.1029/2006GL025734). Thus, sea level rise may significantly impact coastal populations and activities due, among others, to the local vertical land motion processes in the basin that can cause land subsidence amplifying the effects of rising sea levels in the region [[9]](https://doi.org/10.1029/2011JC007469), [[10]](https://doi.org/10.1016/j.jog.2019.05.007). This makes necessary the understanding of sea level change at the scale of the Mediterranean Sea [[2]](https://doi.org/10.3389/fmars.2023.1150488). In this context, this notebook aims to evaluate the long-term sea level variability in the Mediterranean at both regional and sub-basin scales over the period 1993-2023. This assessment is based on the analysis conducted in [[2]](https://doi.org/10.3389/fmars.2023.1150488) from satellite sea level data.", "text_with_prefix": "EQC Quality Assessment: \"Regional sea level trend assessment in the Mediterranean Sea from Satellite (observations) at basin and sub-basin scale\"\nDataset: satellite-sea-level-global [CDS]\nAspect: trend-assessment_q01 | Category: Satellite_ECVs\nSection: Regional sea level trend assessment in the Mediterranean Sea from Satellite (observations) at basin and sub-basin scale > Quality assessment question\n---\n* **Can we use satellite-derived sea level anomalies to detect sea level trend and its spatial variability over the Mediterranean Sea?**\n\nSea level variation is an important indicator of climate change, as it integrates changes of almost all the components in the climate system [[1]](https://doi.org/10.5194/essd-10-281-2018). It specially impacts coastal societies and environments [[2]](https://doi.org/10.3389/fmars.2023.1150488). However, sea level change at the global scale is not uniform varying on a regional to local scale from global mean rates (e.g.,[[3]](https://doi.org/10.1016/j.gloplacha.2013.12.004), [[4]](https://doi.org/10.5194/esd-5-243-2014)). For instance, sea level trends in the Mediterranean differ from global mean sea level trends, mainly because of its semi-enclosed conditions [[5]](https://doi.org/10.1175/JCLI-D-13-00139.1). In addition, within the Mediterranean, sea level trends can differ from the basin mean [[6]](https://doi.org/10.1007/s00382-016-3001-2) due to non-linear local oceanographic processes [[7]](https://doi.org/10.1016/j.gloplacha.2009.04.002). The Mediterranean Sea has been classified as one of the most susceptible climate change zones worldwide [[8]](https://doi.org/10.1029/2006GL025734). Thus, sea level rise may significantly impact coastal populations and activities due, among others, to the local vertical land motion processes in the basin that can cause land subsidence amplifying the effects of rising sea levels in the region [[9]](https://doi.org/10.1029/2011JC007469), [[10]](https://doi.org/10.1016/j.jog.2019.05.007). This makes necessary the understanding of sea level change at the scale of the Mediterranean Sea [[2]](https://doi.org/10.3389/fmars.2023.1150488). In this context, this notebook aims to evaluate the long-term sea level variability in the Mediterranean at both regional and sub-basin scales over the period 1993-2023. This assessment is based on the analysis conducted in [[2]](https://doi.org/10.3389/fmars.2023.1150488) from satellite sea level data."} {"chunk_id": "satellite_satellite-sea-level-global_trend-assessment_q01__5f3108f53a7c", "report_id": "satellite_satellite-sea-level-global_trend-assessment_q01", "dataset_id": "satellite-sea-level-global", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q01", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Regional sea level trend assessment in the Mediterranean Sea from Satellite (observations) at basin and sub-basin scale > Quality assessment statement", "title": "Regional sea level trend assessment in the Mediterranean Sea from Satellite (observations) at basin and sub-basin scale", "chunk_index": 2, "token_count": 789, "text_raw": "These are the key outcomes of this assessment\n\n* This dataset provides statistically significant long-term sea level trends at regional scale in the Mediterranean Sea. The average sea level trend in the basin is 2.1 ± 0.5 mm/yr (95% confidence interval) and presents a large spatial variability since 1993\n* At sub-basin scale, the highest positive rates are observed over the Aegean Sea (3.1 ± 1.0 mm/yr) and Levantine Basin (2.6 ± 0.9 mm/yr), especially in regions where recurring gyres and eddies in the ocean circulation are present\n* The sea level trends obtained and their variability at regional and sub-basin scales are consistent with previous studies based on both altimetry and in situ measurements [[6]](https://doi.org/10.1007/s00382-016-3001-2), [[11]](https://doi.org/10.1007/s00024-019-02156-w). Thus, this dataset can be used to estimate long-term sea level trends in the Mediterranean Sea\n* However, there are regions where the significance of the rate obtained from the trend analysis cannot be statistically confirmed due to local dynamics, such as the influence of gyres/eddies and/or currents. These zones are: the Balearic Islands area in the Western Mediterranean, in a very large portion of the Ionian Peninsula, and also over Southern Crete.\n```\n\n![MedSea_currents](https://www.frontiersin.org/files/Articles/1150488/fmars-10-1150488-HTML/image_m/fmars-10-1150488-g001.jpg)\n\nFigure 1 from [[2]](https://doi.org/10.3389/fmars.2023.1150488). Panel A shows the bathymetry (km) of the Mediterranean Sea with the sub-regions used to compute sea level trends according to [[17]](https://doi.org/10.1007/s00382-012-1369-1) and [[18]](https://doi.org/10.1016/j.gloplacha.2014.10.007). Panel B displays the mean velocity currents (m s⁻¹) in the basin at 15 m depth together with a schematic representation of the surface (black arrows) and intermediate (red arrows) circulation in the Mediterranean. The labels of currents are in italics white, while gyres are in bold cyan. Acronyms: AIS, Atlantic Ionian Stream; AMC, Asia Minor Current; EAG, Eastern Alboran Gyre; EIC, Eastern Ionian Current; IG, Ierapetra Gyre; LGG, Gulf of Lion Gyre; LPCC, Liguro-Provencal-Catalan Current; MIJ, Mid-Ionian Jet; MMGS, Mersa Matruh Gyre System; MMJ, Mid-Mediterranean Jet; NIG, Northern Ionian Gyre; PG, Pelops Gyre; RG, Rhodes Gyre; SAG, Southern Adriatic Gyre; SG, Sirte Gyre; SGS; Shikmona Gyre System; SSTC, Sicily Strait Tunisian Current; WAG, Western Alboran Gyre.", "text_with_prefix": "EQC Quality Assessment: \"Regional sea level trend assessment in the Mediterranean Sea from Satellite (observations) at basin and sub-basin scale\"\nDataset: satellite-sea-level-global [CDS]\nAspect: trend-assessment_q01 | Category: Satellite_ECVs\nSection: Regional sea level trend assessment in the Mediterranean Sea from Satellite (observations) at basin and sub-basin scale > Quality assessment statement\n---\nThese are the key outcomes of this assessment\n\n* This dataset provides statistically significant long-term sea level trends at regional scale in the Mediterranean Sea. The average sea level trend in the basin is 2.1 ± 0.5 mm/yr (95% confidence interval) and presents a large spatial variability since 1993\n* At sub-basin scale, the highest positive rates are observed over the Aegean Sea (3.1 ± 1.0 mm/yr) and Levantine Basin (2.6 ± 0.9 mm/yr), especially in regions where recurring gyres and eddies in the ocean circulation are present\n* The sea level trends obtained and their variability at regional and sub-basin scales are consistent with previous studies based on both altimetry and in situ measurements [[6]](https://doi.org/10.1007/s00382-016-3001-2), [[11]](https://doi.org/10.1007/s00024-019-02156-w). Thus, this dataset can be used to estimate long-term sea level trends in the Mediterranean Sea\n* However, there are regions where the significance of the rate obtained from the trend analysis cannot be statistically confirmed due to local dynamics, such as the influence of gyres/eddies and/or currents. These zones are: the Balearic Islands area in the Western Mediterranean, in a very large portion of the Ionian Peninsula, and also over Southern Crete.\n```\n\n![MedSea_currents](https://www.frontiersin.org/files/Articles/1150488/fmars-10-1150488-HTML/image_m/fmars-10-1150488-g001.jpg)\n\nFigure 1 from [[2]](https://doi.org/10.3389/fmars.2023.1150488). Panel A shows the bathymetry (km) of the Mediterranean Sea with the sub-regions used to compute sea level trends according to [[17]](https://doi.org/10.1007/s00382-012-1369-1) and [[18]](https://doi.org/10.1016/j.gloplacha.2014.10.007). Panel B displays the mean velocity currents (m s⁻¹) in the basin at 15 m depth together with a schematic representation of the surface (black arrows) and intermediate (red arrows) circulation in the Mediterranean. The labels of currents are in italics white, while gyres are in bold cyan. Acronyms: AIS, Atlantic Ionian Stream; AMC, Asia Minor Current; EAG, Eastern Alboran Gyre; EIC, Eastern Ionian Current; IG, Ierapetra Gyre; LGG, Gulf of Lion Gyre; LPCC, Liguro-Provencal-Catalan Current; MIJ, Mid-Ionian Jet; MMGS, Mersa Matruh Gyre System; MMJ, Mid-Mediterranean Jet; NIG, Northern Ionian Gyre; PG, Pelops Gyre; RG, Rhodes Gyre; SAG, Southern Adriatic Gyre; SG, Sirte Gyre; SGS; Shikmona Gyre System; SSTC, Sicily Strait Tunisian Current; WAG, Western Alboran Gyre."} {"chunk_id": "satellite_satellite-sea-level-global_trend-assessment_q01__36cd175d9497", "report_id": "satellite_satellite-sea-level-global_trend-assessment_q01", "dataset_id": "satellite-sea-level-global", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q01", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Regional sea level trend assessment in the Mediterranean Sea from Satellite (observations) at basin and sub-basin scale > Methodology", "title": "Regional sea level trend assessment in the Mediterranean Sea from Satellite (observations) at basin and sub-basin scale", "chunk_index": 3, "token_count": 657, "text_raw": "This notebook provides an assessment of long-term sea level trends in the Mediterranean Sea from satellite-derived sea level observations at both regional and sub-basin scales. The Climate Data Store (CDS) catalogue entry used is the following:\n\n* Sea level gridded data from satellite observations for the global ocean from 1993 to present [[CDS entry]](https://doi.org/10.24381/cds.4c328c78)\n\nIn a first step, these observations are corrected for glacial isostatic adjustment (GIA) using the geoid height estimates (dGeoid) from the ICE-6G_C (VM5a) model [[12]](https://doi.org/10.1002/2014JB011176) as a correction for the local vertical land movements. In a second step, a TOPEX-A instrumental drift correction, derived from altimetry and tide gauge global comparisons, is added to the sea level dataset to account for altimeter instrumental drift that influences the accuracy and uncertainty of the records between 1993-1998 (e.g., [[13]](https://doi.org/10.1038/nclimate2635)). Although this correction is computed for the global mean sea level, it can also be used at regional or local scales as a best available estimate. This is still preferable to not correcting at all, given that the regional variation of the instrumental drift is currently unknown [[2]](https://doi.org/10.3389/fmars.2023.1150488). In addition, the mean seasonal cycle is removed from the time series. Finally, a non-parametric Mann–Kendall test ([[14]](https://doi.org/10.2307/1907187), [[15]](https://psycnet.apa.org/record/1948-15040-000)) modified for auto-correlated data [[16]](https://doi.org/10.1016/S0022-1694%2897%2900125-X) is used to assess statistical significance of the trend at the 95% confidence interval. Sea level trends are computed over the period 1993-2023 at basin scale and for the following sub-regions: Western Mediterranean basin, Adriatic Sea, Aegean Sea, Ionian Sea, Levantine Basin, Southern Central Mediterranean, Southern Crete and Tyrrhenian Sea. They agree with the dynamical regions described in [[17]](https://doi.org/10.1007/s00382-012-1369-1), [[18]](https://doi.org/10.1016/j.gloplacha.2014.10.007)\n\nThe analysis and results are organised in the following steps, which are detailed in the sections below:\n\n**[](results)**", "text_with_prefix": "EQC Quality Assessment: \"Regional sea level trend assessment in the Mediterranean Sea from Satellite (observations) at basin and sub-basin scale\"\nDataset: satellite-sea-level-global [CDS]\nAspect: trend-assessment_q01 | Category: Satellite_ECVs\nSection: Regional sea level trend assessment in the Mediterranean Sea from Satellite (observations) at basin and sub-basin scale > Methodology\n---\nThis notebook provides an assessment of long-term sea level trends in the Mediterranean Sea from satellite-derived sea level observations at both regional and sub-basin scales. The Climate Data Store (CDS) catalogue entry used is the following:\n\n* Sea level gridded data from satellite observations for the global ocean from 1993 to present [[CDS entry]](https://doi.org/10.24381/cds.4c328c78)\n\nIn a first step, these observations are corrected for glacial isostatic adjustment (GIA) using the geoid height estimates (dGeoid) from the ICE-6G_C (VM5a) model [[12]](https://doi.org/10.1002/2014JB011176) as a correction for the local vertical land movements. In a second step, a TOPEX-A instrumental drift correction, derived from altimetry and tide gauge global comparisons, is added to the sea level dataset to account for altimeter instrumental drift that influences the accuracy and uncertainty of the records between 1993-1998 (e.g., [[13]](https://doi.org/10.1038/nclimate2635)). Although this correction is computed for the global mean sea level, it can also be used at regional or local scales as a best available estimate. This is still preferable to not correcting at all, given that the regional variation of the instrumental drift is currently unknown [[2]](https://doi.org/10.3389/fmars.2023.1150488). In addition, the mean seasonal cycle is removed from the time series. Finally, a non-parametric Mann–Kendall test ([[14]](https://doi.org/10.2307/1907187), [[15]](https://psycnet.apa.org/record/1948-15040-000)) modified for auto-correlated data [[16]](https://doi.org/10.1016/S0022-1694%2897%2900125-X) is used to assess statistical significance of the trend at the 95% confidence interval. Sea level trends are computed over the period 1993-2023 at basin scale and for the following sub-regions: Western Mediterranean basin, Adriatic Sea, Aegean Sea, Ionian Sea, Levantine Basin, Southern Central Mediterranean, Southern Crete and Tyrrhenian Sea. They agree with the dynamical regions described in [[17]](https://doi.org/10.1007/s00382-012-1369-1), [[18]](https://doi.org/10.1016/j.gloplacha.2014.10.007)\n\nThe analysis and results are organised in the following steps, which are detailed in the sections below:\n\n**[](results)**"} {"chunk_id": "satellite_satellite-sea-level-global_trend-assessment_q01__b2a91a469d73", "report_id": "satellite_satellite-sea-level-global_trend-assessment_q01", "dataset_id": "satellite-sea-level-global", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q01", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Regional sea level trend assessment in the Mediterranean Sea from Satellite (observations) at basin and sub-basin scale > Analysis and results", "title": "Regional sea level trend assessment in the Mediterranean Sea from Satellite (observations) at basin and sub-basin scale", "chunk_index": 4, "token_count": 128, "text_raw": "This notebook does not include code. It is a literature review based on the results reported in [[2]](https://doi.org/10.3389/fmars.2023.1150488)\n\n(results)=", "text_with_prefix": "EQC Quality Assessment: \"Regional sea level trend assessment in the Mediterranean Sea from Satellite (observations) at basin and sub-basin scale\"\nDataset: satellite-sea-level-global [CDS]\nAspect: trend-assessment_q01 | Category: Satellite_ECVs\nSection: Regional sea level trend assessment in the Mediterranean Sea from Satellite (observations) at basin and sub-basin scale > Analysis and results\n---\nThis notebook does not include code. It is a literature review based on the results reported in [[2]](https://doi.org/10.3389/fmars.2023.1150488)\n\n(results)="} {"chunk_id": "satellite_satellite-sea-level-global_trend-assessment_q01__a559f1ab65f9", "report_id": "satellite_satellite-sea-level-global_trend-assessment_q01", "dataset_id": "satellite-sea-level-global", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q01", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Regional sea level trend assessment in the Mediterranean Sea from Satellite (observations) at basin and sub-basin scale > Analysis and results > 1. Results", "title": "Regional sea level trend assessment in the Mediterranean Sea from Satellite (observations) at basin and sub-basin scale", "chunk_index": 5, "token_count": 943, "text_raw": "The average sea level trend in the basin is 2.1 ± 0.5 mm/yr (95% confidence interval) and presents a large spatial variability since 1993. The sea level trends for the different subregions investigated are the following: Adriatic Sea (2.6 ± 0.8 mm/yr); Aegean Sea (3.1 ± 1.0 mm/yr); Ionian Sea (1.6 ± 1.6 mm/yr); Levantine Basin (2.6 ± 0.9 mm/yr); Southern Central Mediterranean (2.1 ± 0.8 mm/yr); Southern Crete (0.3 ± 1.3 mm/yr); Tyrrhenian Sea (2.5 ± 0.6 mm/yr) and Western Mediterranean (1.8 ± 0.6 mm/yr). The following figure (Figure 5 from [[2]](https://doi.org/10.3389/fmars.2023.1150488)) shows the temporal evolution of the sea level in the Mediterranean Sea at basin and sub-basin scale. The time series of average annual total mean sea level from altimetry for the entire Mediterranean Sea (panel A, solid black line) highlights how the sea level is rising in the basin.\n\n![sub-basin trends](https://www.frontiersin.org/files/Articles/1150488/fmars-10-1150488-HTML/image_m/fmars-10-1150488-g005.jpg)\n\nFigure 5 from [[2]](https://doi.org/10.3389/fmars.2023.1150488). Annual total sea level time series from altimetry (black solid line) and steric (dashed yellow), thermosteric (dotted red), and halosteric (dash-dotted blue) components over different spatial scales (see also map at the top of figure): Mediterranean Sea (A), Adriatic Sea (B), Aegean Sea (C), Ionian Sea (D), Levantine Basin (E), Southern Central Mediterranean (F), Southern Crete (G), Tyrrhenian Sea (H) and Western Mediterranean Basin (I). Vertical lines denote significant changepoint in the related component (same color and line style) at a specific time. Trends refer to the period 1993–2019. Satellite observations provide total sea level, whereas the steric component of sea level due to the thermosteric and halosteric effects are computed in [[2]](https://doi.org/10.3389/fmars.2023.1150488) using ocean temperature and salinity profiles from a reanalysis product. This notebook focuses on the use of satellite observations to compute trends of total sea level. Thus, the steric component of sea level is not assessed.\n\nThe sea level trends obtained and their variability at regional and sub-basin scales are consistent with previous studies based on both satellite and in situ measurements ([[6]](https://doi.org/10.1007/s00382-016-3001-2), [[11]](https://doi.org/10.1007/s00024-019-02156-w)). Thus, this dataset can be used to estimate long-term sea level trends in the Mediterranean Sea.\nThe highest positive rates are observed over the Aegean (3.1 ± 1.0 mm/yr) and Levantine (2.6 ± 0.9 mm/yr), especially in regions where recurring gyres and eddies in the circulation are present (see Figure 1 from [[2]](https://doi.org/10.3389/fmars.2023.1150488)). However, there are zones of non-significance in the sea level trend that reflect areas of critical passages of different water masses with related complex circulation at surface and intermediate depths. It is the case of the Balearic Islands sector in the Southern Central Mediterranean and the Southern Crete region. Also the Ionian Sea lacks a significant trend over the whole period considered.", "text_with_prefix": "EQC Quality Assessment: \"Regional sea level trend assessment in the Mediterranean Sea from Satellite (observations) at basin and sub-basin scale\"\nDataset: satellite-sea-level-global [CDS]\nAspect: trend-assessment_q01 | Category: Satellite_ECVs\nSection: Regional sea level trend assessment in the Mediterranean Sea from Satellite (observations) at basin and sub-basin scale > Analysis and results > 1. Results\n---\nThe average sea level trend in the basin is 2.1 ± 0.5 mm/yr (95% confidence interval) and presents a large spatial variability since 1993. The sea level trends for the different subregions investigated are the following: Adriatic Sea (2.6 ± 0.8 mm/yr); Aegean Sea (3.1 ± 1.0 mm/yr); Ionian Sea (1.6 ± 1.6 mm/yr); Levantine Basin (2.6 ± 0.9 mm/yr); Southern Central Mediterranean (2.1 ± 0.8 mm/yr); Southern Crete (0.3 ± 1.3 mm/yr); Tyrrhenian Sea (2.5 ± 0.6 mm/yr) and Western Mediterranean (1.8 ± 0.6 mm/yr). The following figure (Figure 5 from [[2]](https://doi.org/10.3389/fmars.2023.1150488)) shows the temporal evolution of the sea level in the Mediterranean Sea at basin and sub-basin scale. The time series of average annual total mean sea level from altimetry for the entire Mediterranean Sea (panel A, solid black line) highlights how the sea level is rising in the basin.\n\n![sub-basin trends](https://www.frontiersin.org/files/Articles/1150488/fmars-10-1150488-HTML/image_m/fmars-10-1150488-g005.jpg)\n\nFigure 5 from [[2]](https://doi.org/10.3389/fmars.2023.1150488). Annual total sea level time series from altimetry (black solid line) and steric (dashed yellow), thermosteric (dotted red), and halosteric (dash-dotted blue) components over different spatial scales (see also map at the top of figure): Mediterranean Sea (A), Adriatic Sea (B), Aegean Sea (C), Ionian Sea (D), Levantine Basin (E), Southern Central Mediterranean (F), Southern Crete (G), Tyrrhenian Sea (H) and Western Mediterranean Basin (I). Vertical lines denote significant changepoint in the related component (same color and line style) at a specific time. Trends refer to the period 1993–2019. Satellite observations provide total sea level, whereas the steric component of sea level due to the thermosteric and halosteric effects are computed in [[2]](https://doi.org/10.3389/fmars.2023.1150488) using ocean temperature and salinity profiles from a reanalysis product. This notebook focuses on the use of satellite observations to compute trends of total sea level. Thus, the steric component of sea level is not assessed.\n\nThe sea level trends obtained and their variability at regional and sub-basin scales are consistent with previous studies based on both satellite and in situ measurements ([[6]](https://doi.org/10.1007/s00382-016-3001-2), [[11]](https://doi.org/10.1007/s00024-019-02156-w)). Thus, this dataset can be used to estimate long-term sea level trends in the Mediterranean Sea.\nThe highest positive rates are observed over the Aegean (3.1 ± 1.0 mm/yr) and Levantine (2.6 ± 0.9 mm/yr), especially in regions where recurring gyres and eddies in the circulation are present (see Figure 1 from [[2]](https://doi.org/10.3389/fmars.2023.1150488)). However, there are zones of non-significance in the sea level trend that reflect areas of critical passages of different water masses with related complex circulation at surface and intermediate depths. It is the case of the Balearic Islands sector in the Southern Central Mediterranean and the Southern Crete region. Also the Ionian Sea lacks a significant trend over the whole period considered."} {"chunk_id": "satellite_satellite-sea-level-global_trend-assessment_q01__e9c23759f5a3", "report_id": "satellite_satellite-sea-level-global_trend-assessment_q01", "dataset_id": "satellite-sea-level-global", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q01", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Regional sea level trend assessment in the Mediterranean Sea from Satellite (observations) at basin and sub-basin scale > ℹ️ If you want to know more > Key resources", "title": "Regional sea level trend assessment in the Mediterranean Sea from Satellite (observations) at basin and sub-basin scale", "chunk_index": 6, "token_count": 298, "text_raw": "* CDS entry: [Sea level gridded data from satellite observations for the global ocean from 1993 to present](https://cds.climate.copernicus.eu/datasets/satellite-sea-level-global?tab=overview)\n* Product User Guide and Specification (PUGS): [vDT2018](https://dast.copernicus-climate.eu/documents/satellite-sea-level/vDT2018/D3.SL.1-v1.2_PUGS_of_v1DT2018_SeaLevel_products_v2.4.pdf) and [vDT2021](https://confluence.ecmwf.int/pages/viewpage.action?pageId=333790908)\n\nFurther information:\n* Ocean Indicator: [Sea Level](https://climate.copernicus.eu/climate-indicators/sea-level)\n* Understanding Sea Level: [Sea Level Change by NASA](https://sealevel.nasa.gov/understanding-sea-level/global-sea-level/overview/)", "text_with_prefix": "EQC Quality Assessment: \"Regional sea level trend assessment in the Mediterranean Sea from Satellite (observations) at basin and sub-basin scale\"\nDataset: satellite-sea-level-global [CDS]\nAspect: trend-assessment_q01 | Category: Satellite_ECVs\nSection: Regional sea level trend assessment in the Mediterranean Sea from Satellite (observations) at basin and sub-basin scale > ℹ️ If you want to know more > Key resources\n---\n* CDS entry: [Sea level gridded data from satellite observations for the global ocean from 1993 to present](https://cds.climate.copernicus.eu/datasets/satellite-sea-level-global?tab=overview)\n* Product User Guide and Specification (PUGS): [vDT2018](https://dast.copernicus-climate.eu/documents/satellite-sea-level/vDT2018/D3.SL.1-v1.2_PUGS_of_v1DT2018_SeaLevel_products_v2.4.pdf) and [vDT2021](https://confluence.ecmwf.int/pages/viewpage.action?pageId=333790908)\n\nFurther information:\n* Ocean Indicator: [Sea Level](https://climate.copernicus.eu/climate-indicators/sea-level)\n* Understanding Sea Level: [Sea Level Change by NASA](https://sealevel.nasa.gov/understanding-sea-level/global-sea-level/overview/)"} {"chunk_id": "satellite_satellite-sea-level-global_trend-assessment_q01__d30fa5b1a570", "report_id": "satellite_satellite-sea-level-global_trend-assessment_q01", "dataset_id": "satellite-sea-level-global", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q01", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Regional sea level trend assessment in the Mediterranean Sea from Satellite (observations) at basin and sub-basin scale > ℹ️ If you want to know more > References", "title": "Regional sea level trend assessment in the Mediterranean Sea from Satellite (observations) at basin and sub-basin scale", "chunk_index": 7, "token_count": 1004, "text_raw": "[[1]](https://doi.org/10.5194/essd-10-281-2018) Legeais, J.-F., Ablain, M., Zawadzki, L., Zuo, H., Johannessen, J. A., Scharffenberg, M.G., et al. (2018). An improved and homogeneous altimeter sea level record from the ESA climate change initiative. Earth System Sci. Data 10 (1), 281–301. doi: 10.5194/essd-10-281-2018\n\n[[2]](https://doi.org/10.3389/fmars.2023.1150488) Meli M, Camargo CML, Olivieri M, Slangen ABA and Romagnoli C (2023) Sea-level trend variability in the Mediterranean during the 1993–2019 period. Front. Mar. Sci. 10:1150488. doi: 10.3389/fmars.2023.1150488\n\n[[3]](https://doi.org/10.1016/j.gloplacha.2013.12.004) Jevrejeva, S., Moore, J. C., Grinsted, A., Matthews, A. P., and Spada, G. (2014). Trends and acceleration in global and regional sea levels since 1807. Glob. Planet. Change 113, 11–22. doi: 10.1016/j.gloplacha.2013.12.004\n\n[[4]](https://doi.org/10.5194/esd-5-243-2014) Slangen, A. B. A., van de Wal, R. S. W., Wada, Y. , and Vermeersen, L. L. A. (2014). Comparing tide gauge observations to regional patterns of sea-level change, (1961-2003). Earth Syst. Dynam. 5, 243–255. doi: 10.5194/esd-5-243-2014\n\n[[5]](https://doi.org/10.1175/JCLI-D-13-00139.1) Pinardi, N., Bonaduce, A., Navarra, A., Dobricic, S., and Oddo, P. (2014). The mean Sea level equation and its application to the Mediterranean Sea. J. Clim. 27, 442–447. doi: 10.1175/JCLI-D-13-00139.1\n\n[[6]](https://doi.org/10.1007/s00382-016-3001-2) Bonaduce, A., Pinardi, N., Oddo, P., Spada, G., and Larnicol, G. (2016). Sea-Level variability in the Mediterranean Sea from altimetry and tide gauges. Clim. Dyn. 47, 2851–2866. doi: 10.1007/s00382-016-3001-2\n\n[[7]](https://doi.org/10.1016/j.gloplacha.2009.04.002) Vera, J. D. R., Criado-Aldeanueva, F., Garcı́a-Lafuente, J., and Soto-Navarro, F. J.(2009). A new insight on the decreasing sea level trend over the Ionian basin in the last decades. Global Planetary Change 68, 232–235. doi: 10.1016/j.gloplacha.2009.04.002\n\n[[8]](https://doi.org/10.1029/2006GL025734) Giorgi, F. (2006). Climate change hot-spots. Geophys. Res. Lett. 33, L08707. doi: 10.1029/2006GL025734\n\n[[9]](https://doi.org/10.1029/2011JC007469) Wöppelmann, G., and Marcos, M. (2012). Coastal sea level rise in southern Europe and the nonclimate contribution of vertical land motion. J. Geophys. Res. 117, C01007. doi: 10.1029/2011JC007469", "text_with_prefix": "EQC Quality Assessment: \"Regional sea level trend assessment in the Mediterranean Sea from Satellite (observations) at basin and sub-basin scale\"\nDataset: satellite-sea-level-global [CDS]\nAspect: trend-assessment_q01 | Category: Satellite_ECVs\nSection: Regional sea level trend assessment in the Mediterranean Sea from Satellite (observations) at basin and sub-basin scale > ℹ️ If you want to know more > References\n---\n[[1]](https://doi.org/10.5194/essd-10-281-2018) Legeais, J.-F., Ablain, M., Zawadzki, L., Zuo, H., Johannessen, J. A., Scharffenberg, M.G., et al. (2018). An improved and homogeneous altimeter sea level record from the ESA climate change initiative. Earth System Sci. Data 10 (1), 281–301. doi: 10.5194/essd-10-281-2018\n\n[[2]](https://doi.org/10.3389/fmars.2023.1150488) Meli M, Camargo CML, Olivieri M, Slangen ABA and Romagnoli C (2023) Sea-level trend variability in the Mediterranean during the 1993–2019 period. Front. Mar. Sci. 10:1150488. doi: 10.3389/fmars.2023.1150488\n\n[[3]](https://doi.org/10.1016/j.gloplacha.2013.12.004) Jevrejeva, S., Moore, J. C., Grinsted, A., Matthews, A. P., and Spada, G. (2014). Trends and acceleration in global and regional sea levels since 1807. Glob. Planet. Change 113, 11–22. doi: 10.1016/j.gloplacha.2013.12.004\n\n[[4]](https://doi.org/10.5194/esd-5-243-2014) Slangen, A. B. A., van de Wal, R. S. W., Wada, Y. , and Vermeersen, L. L. A. (2014). Comparing tide gauge observations to regional patterns of sea-level change, (1961-2003). Earth Syst. Dynam. 5, 243–255. doi: 10.5194/esd-5-243-2014\n\n[[5]](https://doi.org/10.1175/JCLI-D-13-00139.1) Pinardi, N., Bonaduce, A., Navarra, A., Dobricic, S., and Oddo, P. (2014). The mean Sea level equation and its application to the Mediterranean Sea. J. Clim. 27, 442–447. doi: 10.1175/JCLI-D-13-00139.1\n\n[[6]](https://doi.org/10.1007/s00382-016-3001-2) Bonaduce, A., Pinardi, N., Oddo, P., Spada, G., and Larnicol, G. (2016). Sea-Level variability in the Mediterranean Sea from altimetry and tide gauges. Clim. Dyn. 47, 2851–2866. doi: 10.1007/s00382-016-3001-2\n\n[[7]](https://doi.org/10.1016/j.gloplacha.2009.04.002) Vera, J. D. R., Criado-Aldeanueva, F., Garcı́a-Lafuente, J., and Soto-Navarro, F. J.(2009). A new insight on the decreasing sea level trend over the Ionian basin in the last decades. Global Planetary Change 68, 232–235. doi: 10.1016/j.gloplacha.2009.04.002\n\n[[8]](https://doi.org/10.1029/2006GL025734) Giorgi, F. (2006). Climate change hot-spots. Geophys. Res. Lett. 33, L08707. doi: 10.1029/2006GL025734\n\n[[9]](https://doi.org/10.1029/2011JC007469) Wöppelmann, G., and Marcos, M. (2012). Coastal sea level rise in southern Europe and the nonclimate contribution of vertical land motion. J. Geophys. Res. 117, C01007. doi: 10.1029/2011JC007469"} {"chunk_id": "satellite_satellite-sea-level-global_trend-assessment_q01__53201f843af1", "report_id": "satellite_satellite-sea-level-global_trend-assessment_q01", "dataset_id": "satellite-sea-level-global", "store": "CDS", "doc_type": "EQC_QA", "aspect": "trend-assessment_q01", "aspect_base": "trend-assessment", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Regional sea level trend assessment in the Mediterranean Sea from Satellite (observations) at basin and sub-basin scale > ℹ️ If you want to know more > References", "title": "Regional sea level trend assessment in the Mediterranean Sea from Satellite (observations) at basin and sub-basin scale", "chunk_index": 8, "token_count": 1000, "text_raw": "). Coastal sea level rise in southern Europe and the nonclimate contribution of vertical land motion. J. Geophys. Res. 117, C01007. doi: 10.1029/2011JC007469\n\n[[10]](https://doi.org/10.1016/j.jog.2019.05.007) Mohamed, B., Mohamed, A., Alam El-Din, K., Nagy, H., and Elsherbiny, A. (2019a). Sea Level changes and vertical land motion from altimetry and tide gauges in the southern levantine basin. J. Geodyn. 128, 1–10. doi: 10.1016/j.jog.2019.05.007\n\n[[11]](https://doi.org/10.1007/s00024-019-02156-w) Mohamed, B., Abdallah, A. M., Alam El-Din, K., Nagy, H., and Shaltout, M. (2019a). Inter-annual variability and trends of Sea level and Sea surface temperature in the Mediterranean Sea over the last 25 years. Pure Appl. Geophys. 176, 3787–3810. doi: 10.1007/s00024-019-02156-w\n\n[[12]](https://doi.org/10.1002/2014JB011176) Peltier, W. R., Argus, D. F., and Drummond, R. (2015). Space geodesy constrains ice age terminal deglaciation: The global ICE-6G-C (VM5a) model. J. Geophys. Res. Solid Earth 120, 450–487. doi: 10.1002/2014JB011176\n\n[[13]](https://doi.org/10.1038/nclimate2635) Watson, C. S., White, N. J., Church, J. A., King, M. A., Burgette, R. J., and Legresy, B.(2015). Unabated global mean sea-level rise over the satellite altimeter era. Nat. Clim. Change 5, 565–568. doi: 10.1038/nclimate2635\n\n[[14]](https://doi.org/10.2307/1907187) Mann, H. B. (1945). Non-parametric tests against trend. Econometrica 13, 163–171. doi: 10.2307/1907187\n\n[[15]](https://psycnet.apa.org/record/1948-15040-000) Kendall, M. G. (1975). Rank correlation methods. 4th ed. Ed. C. Griffin (London, UK:Charles Griffin)\n\n[[16]](https://doi.org/10.1016/S0022-1694%2897%2900125-X) Hamed, K. H., and Rao, A. R. (1998). A modified Mann-Kendall trend test for autocorrelated data. J. Hydrol. 204, 182–196. doi: 10.1016/S0022-1694(97)00125-X\n\n[[17]](https://doi.org/10.1007/s00382-012-1369-1) Carillo, A., Sannino, G., Artale, V., Ruti, P., Calmanti, S., and Dell'Aquila, A. (2012).\nSteric sea level rise over the Mediterranean Sea: present climate and scenario\nsimulations. Clim. Dyn. 39, 2167–2184. doi: 10.1007/s00382-012-1369-1\n\n[[18]](https://doi.org/10.1016/j.gloplacha.2014.10.007) Galassi, G., and Spada, G. (2014). Sea-Level rise in the Mediterranean Sea by 2050:\nRoles of terrestrial ice melt, steric effects and glacial isostatic adjustment. Glob. Planet.\nChange 123, 55–66. doi: 10.1016/j.gloplacha.2014.10.007", "text_with_prefix": "EQC Quality Assessment: \"Regional sea level trend assessment in the Mediterranean Sea from Satellite (observations) at basin and sub-basin scale\"\nDataset: satellite-sea-level-global [CDS]\nAspect: trend-assessment_q01 | Category: Satellite_ECVs\nSection: Regional sea level trend assessment in the Mediterranean Sea from Satellite (observations) at basin and sub-basin scale > ℹ️ If you want to know more > References\n---\n). Coastal sea level rise in southern Europe and the nonclimate contribution of vertical land motion. J. Geophys. Res. 117, C01007. doi: 10.1029/2011JC007469\n\n[[10]](https://doi.org/10.1016/j.jog.2019.05.007) Mohamed, B., Mohamed, A., Alam El-Din, K., Nagy, H., and Elsherbiny, A. (2019a). Sea Level changes and vertical land motion from altimetry and tide gauges in the southern levantine basin. J. Geodyn. 128, 1–10. doi: 10.1016/j.jog.2019.05.007\n\n[[11]](https://doi.org/10.1007/s00024-019-02156-w) Mohamed, B., Abdallah, A. M., Alam El-Din, K., Nagy, H., and Shaltout, M. (2019a). Inter-annual variability and trends of Sea level and Sea surface temperature in the Mediterranean Sea over the last 25 years. Pure Appl. Geophys. 176, 3787–3810. doi: 10.1007/s00024-019-02156-w\n\n[[12]](https://doi.org/10.1002/2014JB011176) Peltier, W. R., Argus, D. F., and Drummond, R. (2015). Space geodesy constrains ice age terminal deglaciation: The global ICE-6G-C (VM5a) model. J. Geophys. Res. Solid Earth 120, 450–487. doi: 10.1002/2014JB011176\n\n[[13]](https://doi.org/10.1038/nclimate2635) Watson, C. S., White, N. J., Church, J. A., King, M. A., Burgette, R. J., and Legresy, B.(2015). Unabated global mean sea-level rise over the satellite altimeter era. Nat. Clim. Change 5, 565–568. doi: 10.1038/nclimate2635\n\n[[14]](https://doi.org/10.2307/1907187) Mann, H. B. (1945). Non-parametric tests against trend. Econometrica 13, 163–171. doi: 10.2307/1907187\n\n[[15]](https://psycnet.apa.org/record/1948-15040-000) Kendall, M. G. (1975). Rank correlation methods. 4th ed. Ed. C. Griffin (London, UK:Charles Griffin)\n\n[[16]](https://doi.org/10.1016/S0022-1694%2897%2900125-X) Hamed, K. H., and Rao, A. R. (1998). A modified Mann-Kendall trend test for autocorrelated data. J. Hydrol. 204, 182–196. doi: 10.1016/S0022-1694(97)00125-X\n\n[[17]](https://doi.org/10.1007/s00382-012-1369-1) Carillo, A., Sannino, G., Artale, V., Ruti, P., Calmanti, S., and Dell'Aquila, A. (2012).\nSteric sea level rise over the Mediterranean Sea: present climate and scenario\nsimulations. Clim. Dyn. 39, 2167–2184. doi: 10.1007/s00382-012-1369-1\n\n[[18]](https://doi.org/10.1016/j.gloplacha.2014.10.007) Galassi, G., and Spada, G. (2014). Sea-Level rise in the Mediterranean Sea by 2050:\nRoles of terrestrial ice melt, steric effects and glacial isostatic adjustment. Glob. Planet.\nChange 123, 55–66. doi: 10.1016/j.gloplacha.2014.10.007"} {"chunk_id": "satellite_satellite-sea-level-global_validation_q01__af31989169da", "report_id": "satellite_satellite-sea-level-global_validation_q01", "dataset_id": "satellite-sea-level-global", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q01", "aspect_base": "validation", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Sea level measurements accuracy assessment from Satellite (observations) in the coastal zone", "title": "Sea level measurements accuracy assessment from Satellite (observations) in the coastal zone", "chunk_index": 0, "token_count": 108, "text_raw": "Production date: 29-08-2024\n\nProduced by: Antonio Sánchez-Román [IMEDEA (CSIC-UIB)](https://imedea.uib-csic.es/)", "text_with_prefix": "EQC Quality Assessment: \"Sea level measurements accuracy assessment from Satellite (observations) in the coastal zone\"\nDataset: satellite-sea-level-global [CDS]\nAspect: validation_q01 | Category: Satellite_ECVs\nSection: Sea level measurements accuracy assessment from Satellite (observations) in the coastal zone\n---\nProduction date: 29-08-2024\n\nProduced by: Antonio Sánchez-Román [IMEDEA (CSIC-UIB)](https://imedea.uib-csic.es/)"} {"chunk_id": "satellite_satellite-sea-level-global_validation_q01__13e1e13890e6", "report_id": "satellite_satellite-sea-level-global_validation_q01", "dataset_id": "satellite-sea-level-global", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q01", "aspect_base": "validation", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Sea level measurements accuracy assessment from Satellite (observations) in the coastal zone > Quality assessment questions", "title": "Sea level measurements accuracy assessment from Satellite (observations) in the coastal zone", "chunk_index": 1, "token_count": 748, "text_raw": "* **How accurate are satellite-derived sea level anomaly measurements in the coastal band of the European Seas?**\n* **Can we use satellite-derived sea level anomalies to monitor the long-term evolution of sea level in the coastal band of the European Seas?**\n\nTraditional altimetry has been often unable to produce meaningful signals of sea level change in the coastal zone due to the typically shallower water, bathymetric gradients and shoreline shapes, among others [[1]](https://doi.org/10.1007/s10712-019-09569-1), [[2]](https://doi.org/10.3390/rs12233970). Actually, satellite sea level products are not optimized for the coastal band promoting larger errors in the retrieval of sea level with regard to the open ocean. Nevertheless, the monitoring of sea level changes in coastal areas is an important societal issue [[3]](https://doi.org/10.3390/rs15030793). Thus, most of the efforts of the international community in the recent past have been focused on the research and development of techniques for coastal altimetry, with substantial support from space agencies such as the European Space Agency (ESA), the Centre National d’Études Spatiales (CNES) and other research institutions [[4]](https://doi.org/10.1007/s10712-016-9392-0). Efforts of the coastal altimetry community are aimed at extending the capabilities of current altimeters closer to the coastal zone. This includes the application of improved geophysical corrections, data recovery strategies near the coast using new editing criteria and high-frequency along-track sampling associated with updated quality control procedures [[1]](https://doi.org/10.1007/s10712-019-09569-1).\nAs a result, regional altimeter products focused on the coastal zone have been developed over the last years [[3]](https://doi.org/10.3390/rs15030793). These products are disseminated to both the international scientific community and society through regular specific coastal altimetry workshops.\nDifferent metrics are used to assess the quality of satellite sea level data. They mainly consist in the analysis of the sea level anomaly (SLA) field at different steps of the processing, checks of the consistency of the SLA along the tracks of different altimeters and between gridded and along-track products, and comparisons with external in situ measurements [[5]](https://documentation.marine.copernicus.eu/QUID/CMEMS-SL-QUID-008-032-068.pdf).\nIn situ and satellite observations are complementary and are often assumed to observe the same signals [[6]](https://doi.org/10.1002/2015RG000502). In coastal areas, tide gauge measurements are commonly used (e.g., [[7]](https://doi.org/10.5194/os-15-1207-2019)). \nIn this context, this notebook aims to evaluate the performance of satellite global sea level gridded products in the retrieval of sea level in the coastal band of the European Seas over the period 1993-2020. This assessment is based on the analysis conducted in [[8]](https://doi.org/10.5194/os-19-793-2023) from satellite sea level data.", "text_with_prefix": "EQC Quality Assessment: \"Sea level measurements accuracy assessment from Satellite (observations) in the coastal zone\"\nDataset: satellite-sea-level-global [CDS]\nAspect: validation_q01 | Category: Satellite_ECVs\nSection: Sea level measurements accuracy assessment from Satellite (observations) in the coastal zone > Quality assessment questions\n---\n* **How accurate are satellite-derived sea level anomaly measurements in the coastal band of the European Seas?**\n* **Can we use satellite-derived sea level anomalies to monitor the long-term evolution of sea level in the coastal band of the European Seas?**\n\nTraditional altimetry has been often unable to produce meaningful signals of sea level change in the coastal zone due to the typically shallower water, bathymetric gradients and shoreline shapes, among others [[1]](https://doi.org/10.1007/s10712-019-09569-1), [[2]](https://doi.org/10.3390/rs12233970). Actually, satellite sea level products are not optimized for the coastal band promoting larger errors in the retrieval of sea level with regard to the open ocean. Nevertheless, the monitoring of sea level changes in coastal areas is an important societal issue [[3]](https://doi.org/10.3390/rs15030793). Thus, most of the efforts of the international community in the recent past have been focused on the research and development of techniques for coastal altimetry, with substantial support from space agencies such as the European Space Agency (ESA), the Centre National d’Études Spatiales (CNES) and other research institutions [[4]](https://doi.org/10.1007/s10712-016-9392-0). Efforts of the coastal altimetry community are aimed at extending the capabilities of current altimeters closer to the coastal zone. This includes the application of improved geophysical corrections, data recovery strategies near the coast using new editing criteria and high-frequency along-track sampling associated with updated quality control procedures [[1]](https://doi.org/10.1007/s10712-019-09569-1).\nAs a result, regional altimeter products focused on the coastal zone have been developed over the last years [[3]](https://doi.org/10.3390/rs15030793). These products are disseminated to both the international scientific community and society through regular specific coastal altimetry workshops.\nDifferent metrics are used to assess the quality of satellite sea level data. They mainly consist in the analysis of the sea level anomaly (SLA) field at different steps of the processing, checks of the consistency of the SLA along the tracks of different altimeters and between gridded and along-track products, and comparisons with external in situ measurements [[5]](https://documentation.marine.copernicus.eu/QUID/CMEMS-SL-QUID-008-032-068.pdf).\nIn situ and satellite observations are complementary and are often assumed to observe the same signals [[6]](https://doi.org/10.1002/2015RG000502). In coastal areas, tide gauge measurements are commonly used (e.g., [[7]](https://doi.org/10.5194/os-15-1207-2019)). \nIn this context, this notebook aims to evaluate the performance of satellite global sea level gridded products in the retrieval of sea level in the coastal band of the European Seas over the period 1993-2020. This assessment is based on the analysis conducted in [[8]](https://doi.org/10.5194/os-19-793-2023) from satellite sea level data."} {"chunk_id": "satellite_satellite-sea-level-global_validation_q01__bef94f51f6fc", "report_id": "satellite_satellite-sea-level-global_validation_q01", "dataset_id": "satellite-sea-level-global", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q01", "aspect_base": "validation", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Sea level measurements accuracy assessment from Satellite (observations) in the coastal zone > Quality assessment statement", "title": "Sea level measurements accuracy assessment from Satellite (observations) in the coastal zone", "chunk_index": 2, "token_count": 414, "text_raw": "These are the key outcomes of this assessment\n\n* The satellite dataset shows good agreement with tide gauge sea level observations in the European Seas, with a mean RMS difference of about 4.35 cm for the vDT2021 version. This difference shows spatial variability, likely due to inaccuracies in corrections applied to the satellite data and greater variability in non-tidal sea level changes.\n* The satellite data indicates a consistent sea level rise across the study region, with a mean trend of about 3.13 mm/year. These trends appear more spatially consistent than those derived from tide gauges; however, differences between datasets suggest that regional estimates, particularly in coastal areas, should be interpreted with caution. \n* Differences between vDT2021 and vDT2018 trend estimates are small (4.35 cm and 4.41 cm RMS difference with the tide gauges, respectively) and fall within the expected uncertainty range [[20]](https://doi.org/10.1038/s41597-020-00786-7), meaning no clear improvement can be established between the two versions. This assessment also does not include the newer vDT2024 dataset.\n```\n\n![consistency](https://raw.githubusercontent.com/Blanca-Fdez/c3s2-JN-assets/refs/heads/main/Assets/compTG_2sat_DT2021_tot_%25var_JN_BIEN.png)\n\nFigure 1. Location of the tide gauges along the European coasts and the western Mediterranean Sea used for a comparison with satellite data. Colours indicate the mean square differences between the tide gauge and satellite sea level (vDT2021 version). Units are the percentage of the tide gauge variance.", "text_with_prefix": "EQC Quality Assessment: \"Sea level measurements accuracy assessment from Satellite (observations) in the coastal zone\"\nDataset: satellite-sea-level-global [CDS]\nAspect: validation_q01 | Category: Satellite_ECVs\nSection: Sea level measurements accuracy assessment from Satellite (observations) in the coastal zone > Quality assessment statement\n---\nThese are the key outcomes of this assessment\n\n* The satellite dataset shows good agreement with tide gauge sea level observations in the European Seas, with a mean RMS difference of about 4.35 cm for the vDT2021 version. This difference shows spatial variability, likely due to inaccuracies in corrections applied to the satellite data and greater variability in non-tidal sea level changes.\n* The satellite data indicates a consistent sea level rise across the study region, with a mean trend of about 3.13 mm/year. These trends appear more spatially consistent than those derived from tide gauges; however, differences between datasets suggest that regional estimates, particularly in coastal areas, should be interpreted with caution. \n* Differences between vDT2021 and vDT2018 trend estimates are small (4.35 cm and 4.41 cm RMS difference with the tide gauges, respectively) and fall within the expected uncertainty range [[20]](https://doi.org/10.1038/s41597-020-00786-7), meaning no clear improvement can be established between the two versions. This assessment also does not include the newer vDT2024 dataset.\n```\n\n![consistency](https://raw.githubusercontent.com/Blanca-Fdez/c3s2-JN-assets/refs/heads/main/Assets/compTG_2sat_DT2021_tot_%25var_JN_BIEN.png)\n\nFigure 1. Location of the tide gauges along the European coasts and the western Mediterranean Sea used for a comparison with satellite data. Colours indicate the mean square differences between the tide gauge and satellite sea level (vDT2021 version). Units are the percentage of the tide gauge variance."} {"chunk_id": "satellite_satellite-sea-level-global_validation_q01__84a0ac4cef72", "report_id": "satellite_satellite-sea-level-global_validation_q01", "dataset_id": "satellite-sea-level-global", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q01", "aspect_base": "validation", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Sea level measurements accuracy assessment from Satellite (observations) in the coastal zone > Methodology", "title": "Sea level measurements accuracy assessment from Satellite (observations) in the coastal zone", "chunk_index": 3, "token_count": 818, "text_raw": "We provide an assessment of the accuracy of satellite-derived sea level observations in the coastal zone of the European Seas over the period 1993-2020 through their comparison with in situ tide gauge records located along the European coasts. The Climate Data Store (CDS) catalogue entry used is the following:\n\n* [Sea level gridded data from satellite observations for the global ocean from 1993 to present](https://doi.org/10.24381/cds.4c328c78)\n\nOn the time of podruction, two data versions are available in the CDS: vDT2018 and vDT2021. The improvements in the latest (vDT2021) dataset in the retrieval of sea level in the coastal band of the European Seas respect to the previous version (vDT2018) is investigated. SLA L4 satellite-derived from the CDS are provided on regular grids with nominal spatial resolutions of 0.25 x 0.25 degrees.\n\nOther entries used for the assessment:\n\n* [Tide gauge entry](http://www.marineinsitu.eu)\n* [DAC entry](https://www.aviso.altimetry.fr/en/data/products/auxiliary-products/dynamic-atmospheric-correction.html)\n\nBefore they can be compared with satellite data, tide gauge measurements have to be processed to remove oceanographic signals whose temporal periods are not resolved by satellite sea level data, thus avoiding important aliasing errors [[1]](https://doi.org/10.1007/s10712-019-09569-1). In the following there are summarised the corrections applied to the tide gauge records:\n\n* Correction of oceanic tidal effects by filtering tidal components (mainly diurnal and semidiurnal tidal constituents). The u-tide software [[12]](https://www.po.gso.uri.edu/codiga/utide/2011Codiga-UTide-Report.pdf) is used. The annual and semiannual frequencies, mainly driven by steric effect, are kept in the tidal residuals since they are included in the satellite data.\n* Removal of the atmospheric induced sea level variability caused by the action of atmospheric pressure and wind [[13]](https://doi.org/10.1175/1520-0426%281999%29016<1279:EOGMAP>2.0.CO;2), [[14]](https://doi.org/10.1029/2002GL016473). The same dynamic atmospheric correction (DAC) as for satellite data is applied for the sake of consistency. The 6-hourly fields of this correction, available from the Archiving, Validation and Interpretation of Satellite Oceanographic Data (AVISO) website [[DAC entry]](https://www.aviso.altimetry.fr/en/data/products/auxiliary-products/dynamic-atmospheric-correction.html), are used. For each tide gauge site, the nearest grid point was selected and used to remove the atmospherically induced sea level from observations, previously converted into 6-hourly records [[15]](https://doi.org/10.1016/j.asr.2015.04.027).\n* Correction of vertical movements associated with glacial isostatic adjustment (GIA). GIA is regarded as the only source of vertical land motions. Its effects are removed from the sea surface height (SSH) records, previously averaged into daily data, by using the Peltier mantle viscosity model (VM2) [[16]](https://doi.org/10.1029/98RG02638), [[17]](https://doi.org/10.1146/annurev.earth.32.082503.144359).", "text_with_prefix": "EQC Quality Assessment: \"Sea level measurements accuracy assessment from Satellite (observations) in the coastal zone\"\nDataset: satellite-sea-level-global [CDS]\nAspect: validation_q01 | Category: Satellite_ECVs\nSection: Sea level measurements accuracy assessment from Satellite (observations) in the coastal zone > Methodology\n---\nWe provide an assessment of the accuracy of satellite-derived sea level observations in the coastal zone of the European Seas over the period 1993-2020 through their comparison with in situ tide gauge records located along the European coasts. The Climate Data Store (CDS) catalogue entry used is the following:\n\n* [Sea level gridded data from satellite observations for the global ocean from 1993 to present](https://doi.org/10.24381/cds.4c328c78)\n\nOn the time of podruction, two data versions are available in the CDS: vDT2018 and vDT2021. The improvements in the latest (vDT2021) dataset in the retrieval of sea level in the coastal band of the European Seas respect to the previous version (vDT2018) is investigated. SLA L4 satellite-derived from the CDS are provided on regular grids with nominal spatial resolutions of 0.25 x 0.25 degrees.\n\nOther entries used for the assessment:\n\n* [Tide gauge entry](http://www.marineinsitu.eu)\n* [DAC entry](https://www.aviso.altimetry.fr/en/data/products/auxiliary-products/dynamic-atmospheric-correction.html)\n\nBefore they can be compared with satellite data, tide gauge measurements have to be processed to remove oceanographic signals whose temporal periods are not resolved by satellite sea level data, thus avoiding important aliasing errors [[1]](https://doi.org/10.1007/s10712-019-09569-1). In the following there are summarised the corrections applied to the tide gauge records:\n\n* Correction of oceanic tidal effects by filtering tidal components (mainly diurnal and semidiurnal tidal constituents). The u-tide software [[12]](https://www.po.gso.uri.edu/codiga/utide/2011Codiga-UTide-Report.pdf) is used. The annual and semiannual frequencies, mainly driven by steric effect, are kept in the tidal residuals since they are included in the satellite data.\n* Removal of the atmospheric induced sea level variability caused by the action of atmospheric pressure and wind [[13]](https://doi.org/10.1175/1520-0426%281999%29016<1279:EOGMAP>2.0.CO;2), [[14]](https://doi.org/10.1029/2002GL016473). The same dynamic atmospheric correction (DAC) as for satellite data is applied for the sake of consistency. The 6-hourly fields of this correction, available from the Archiving, Validation and Interpretation of Satellite Oceanographic Data (AVISO) website [[DAC entry]](https://www.aviso.altimetry.fr/en/data/products/auxiliary-products/dynamic-atmospheric-correction.html), are used. For each tide gauge site, the nearest grid point was selected and used to remove the atmospherically induced sea level from observations, previously converted into 6-hourly records [[15]](https://doi.org/10.1016/j.asr.2015.04.027).\n* Correction of vertical movements associated with glacial isostatic adjustment (GIA). GIA is regarded as the only source of vertical land motions. Its effects are removed from the sea surface height (SSH) records, previously averaged into daily data, by using the Peltier mantle viscosity model (VM2) [[16]](https://doi.org/10.1029/98RG02638), [[17]](https://doi.org/10.1146/annurev.earth.32.082503.144359)."} {"chunk_id": "satellite_satellite-sea-level-global_validation_q01__448dd92aa906", "report_id": "satellite_satellite-sea-level-global_validation_q01", "dataset_id": "satellite-sea-level-global", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q01", "aspect_base": "validation", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Sea level measurements accuracy assessment from Satellite (observations) in the coastal zone > Methodology", "title": "Sea level measurements accuracy assessment from Satellite (observations) in the coastal zone", "chunk_index": 4, "token_count": 444, "text_raw": ".1029/98RG02638), [[17]](https://doi.org/10.1146/annurev.earth.32.082503.144359).\n\nThe comparison method of satellite data with tide gauges consists in collocating both datasets in time and space. As a first step, a 15 day low-pass LOESS filter is applied to satellite and tide gauge time series to remove the high frequencies that cannot be resolved by the satellite data [[2]](https://doi.org/10.3390/rs12233970), [[18]](https://doi.org/10.1175/2008JTECHO556.1), [[19]](https://doi.org/10.5194/os-15-1091-2019). Then, the correlations between each tide gauge record and SLA time series corresponding to grid points within a radius of 1 degree around the tide gauge site is computed and the most correlated altimetry point is chosen. Only long-term monitoring stations with a lifetime of more than 3 years are used to allow statistical significance. Statistical analyses are performed using all available data pairs (satellite – tide gauge) in terms of the root mean square (rms) difference and variance of the time series. Consistency between the satellite dataset and the tide gauge data is computed from Eq. (1) in [[2]](https://doi.org/10.3390/rs12233970). It is expressed as the mean square differences between both datasets, computed as the variance of the differences (satellite – tide gauge), in terms of percentage of the tide gauge variance.\n\nThe results are organised in the following steps, which are detailed in the sections below:\n\n**[](section-1)**\n\n**[](section-2)**\n\n**[](section-3)**\n\n**[](section-4)**", "text_with_prefix": "EQC Quality Assessment: \"Sea level measurements accuracy assessment from Satellite (observations) in the coastal zone\"\nDataset: satellite-sea-level-global [CDS]\nAspect: validation_q01 | Category: Satellite_ECVs\nSection: Sea level measurements accuracy assessment from Satellite (observations) in the coastal zone > Methodology\n---\n.1029/98RG02638), [[17]](https://doi.org/10.1146/annurev.earth.32.082503.144359).\n\nThe comparison method of satellite data with tide gauges consists in collocating both datasets in time and space. As a first step, a 15 day low-pass LOESS filter is applied to satellite and tide gauge time series to remove the high frequencies that cannot be resolved by the satellite data [[2]](https://doi.org/10.3390/rs12233970), [[18]](https://doi.org/10.1175/2008JTECHO556.1), [[19]](https://doi.org/10.5194/os-15-1091-2019). Then, the correlations between each tide gauge record and SLA time series corresponding to grid points within a radius of 1 degree around the tide gauge site is computed and the most correlated altimetry point is chosen. Only long-term monitoring stations with a lifetime of more than 3 years are used to allow statistical significance. Statistical analyses are performed using all available data pairs (satellite – tide gauge) in terms of the root mean square (rms) difference and variance of the time series. Consistency between the satellite dataset and the tide gauge data is computed from Eq. (1) in [[2]](https://doi.org/10.3390/rs12233970). It is expressed as the mean square differences between both datasets, computed as the variance of the differences (satellite – tide gauge), in terms of percentage of the tide gauge variance.\n\nThe results are organised in the following steps, which are detailed in the sections below:\n\n**[](section-1)**\n\n**[](section-2)**\n\n**[](section-3)**\n\n**[](section-4)**"} {"chunk_id": "satellite_satellite-sea-level-global_validation_q01__2a870ee335f4", "report_id": "satellite_satellite-sea-level-global_validation_q01", "dataset_id": "satellite-sea-level-global", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q01", "aspect_base": "validation", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Sea level measurements accuracy assessment from Satellite (observations) in the coastal zone > Analysis and results", "title": "Sea level measurements accuracy assessment from Satellite (observations) in the coastal zone", "chunk_index": 5, "token_count": 113, "text_raw": "This notebook does not include code. It is a literature review based on the results reported in [[8]](https://doi.org/10.5194/os-19-793-2023).\n\n(section-1)=", "text_with_prefix": "EQC Quality Assessment: \"Sea level measurements accuracy assessment from Satellite (observations) in the coastal zone\"\nDataset: satellite-sea-level-global [CDS]\nAspect: validation_q01 | Category: Satellite_ECVs\nSection: Sea level measurements accuracy assessment from Satellite (observations) in the coastal zone > Analysis and results\n---\nThis notebook does not include code. It is a literature review based on the results reported in [[8]](https://doi.org/10.5194/os-19-793-2023).\n\n(section-1)="} {"chunk_id": "satellite_satellite-sea-level-global_validation_q01__a67a661d82dd", "report_id": "satellite_satellite-sea-level-global_validation_q01", "dataset_id": "satellite-sea-level-global", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q01", "aspect_base": "validation", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Sea level measurements accuracy assessment from Satellite (observations) in the coastal zone > Analysis and results > 1. Performance of vDT2021 dataset in the retreival of sea level in the coastal band", "title": "Sea level measurements accuracy assessment from Satellite (observations) in the coastal zone", "chunk_index": 6, "token_count": 835, "text_raw": "The mean value (Table 1) of the rms difference between the vDT2021 satellite dataset and tide gauges is 4.35 cm, the variance of the differences (satellite – tide gauge) is 19 cm², whereas the variance of the satellite and tide gauge datasets are 79 cm² and 89 cm², respectively. Thus, this dataset properly captures the sea level in the coastal region and its variability when compared with in-situ tide gauge data. In addition, the mean distance between the location of the tide gauge and the corresponding satellite data with the highest correlation is 87 km. Overall, consistency (Figure 1) between the vDT2021 satellite dataset and the tide gauge data (expressed according to [[2]](https://doi.org/10.3390/rs12233970) as the mean square differences between both datasets in terms of percentage of the tide gauge variance) shows spatial variability with values lower than 5 % in the central and eastern parts of the Baltic Sea, emphasizing the precision of the correction applied to the satellite data in the basin that provides accurate sea level measurements when compared with tide gauges. However, they reach poorer values between 20 % and 30 % for stations located in the connection region with the North Atlantic Ocean. In addition, consistency values are between 20 % and 50 % for most of the stations located along the Atlantic shore; this includes the Strait of Gibraltar area. Such a large error could be related to imprecisions in the correction applied (i.e. ocean tide and DAC) to the satellite data [[2]](https://doi.org/10.3390/rs12233970), [[9]](https://doi.org/10.1029/2007jc004459), [[10]](https://doi.org/10.1016/j.jmarsys.2016.03.006) and to a larger non-tidal variance with respect to that found in the Baltic Sea [[11]](https://doi.org/10.1080/1755876X.2018.1489208). Finally, the Norwegian Sea shows consistency values better than those obtained for the Atlantic shore ranging between 15 % and 30 %, except for the south-western part of Norway where values between 5 % and 15% are found providing accurate sea level measurements when compared with tide gauges.\n\nTable 1. Intercomparison of vDT2021 and vDT2018 satellite dataset (SAT) and tide gauge (TG) data from the European coasts in terms of the rms differences (cm) and variance (cm²) of the differences between the datasets. The mean distance between tide gauges and the most correlated gridded satellite points used in the computation is displayed. The common tide gauge stations for the vDT2021 and vDT2018 datasets were used. Finally, the improvement (%) of the vDT2021 dataset with respect to the previous vDT2018 version is also displayed.\n\n| | vDT2021 | vDT2018 | vDT2021 improvement |\n| :- | :-: | :-: | :-: |\n| rms diff. (cm) | 4.35 | 4.41 | 1% |\n| var. TG (cm²) | 89 | 89 | - |\n| var. SAT (cm²) | 79 | 78 | no improvement |\n| var. TG-SAT (cm²) | 19 | 19 | no improvement |\n| distance TG (km) | 87 | 90 | 3% |\n\n(section-2)=", "text_with_prefix": "EQC Quality Assessment: \"Sea level measurements accuracy assessment from Satellite (observations) in the coastal zone\"\nDataset: satellite-sea-level-global [CDS]\nAspect: validation_q01 | Category: Satellite_ECVs\nSection: Sea level measurements accuracy assessment from Satellite (observations) in the coastal zone > Analysis and results > 1. Performance of vDT2021 dataset in the retreival of sea level in the coastal band\n---\nThe mean value (Table 1) of the rms difference between the vDT2021 satellite dataset and tide gauges is 4.35 cm, the variance of the differences (satellite – tide gauge) is 19 cm², whereas the variance of the satellite and tide gauge datasets are 79 cm² and 89 cm², respectively. Thus, this dataset properly captures the sea level in the coastal region and its variability when compared with in-situ tide gauge data. In addition, the mean distance between the location of the tide gauge and the corresponding satellite data with the highest correlation is 87 km. Overall, consistency (Figure 1) between the vDT2021 satellite dataset and the tide gauge data (expressed according to [[2]](https://doi.org/10.3390/rs12233970) as the mean square differences between both datasets in terms of percentage of the tide gauge variance) shows spatial variability with values lower than 5 % in the central and eastern parts of the Baltic Sea, emphasizing the precision of the correction applied to the satellite data in the basin that provides accurate sea level measurements when compared with tide gauges. However, they reach poorer values between 20 % and 30 % for stations located in the connection region with the North Atlantic Ocean. In addition, consistency values are between 20 % and 50 % for most of the stations located along the Atlantic shore; this includes the Strait of Gibraltar area. Such a large error could be related to imprecisions in the correction applied (i.e. ocean tide and DAC) to the satellite data [[2]](https://doi.org/10.3390/rs12233970), [[9]](https://doi.org/10.1029/2007jc004459), [[10]](https://doi.org/10.1016/j.jmarsys.2016.03.006) and to a larger non-tidal variance with respect to that found in the Baltic Sea [[11]](https://doi.org/10.1080/1755876X.2018.1489208). Finally, the Norwegian Sea shows consistency values better than those obtained for the Atlantic shore ranging between 15 % and 30 %, except for the south-western part of Norway where values between 5 % and 15% are found providing accurate sea level measurements when compared with tide gauges.\n\nTable 1. Intercomparison of vDT2021 and vDT2018 satellite dataset (SAT) and tide gauge (TG) data from the European coasts in terms of the rms differences (cm) and variance (cm²) of the differences between the datasets. The mean distance between tide gauges and the most correlated gridded satellite points used in the computation is displayed. The common tide gauge stations for the vDT2021 and vDT2018 datasets were used. Finally, the improvement (%) of the vDT2021 dataset with respect to the previous vDT2018 version is also displayed.\n\n| | vDT2021 | vDT2018 | vDT2021 improvement |\n| :- | :-: | :-: | :-: |\n| rms diff. (cm) | 4.35 | 4.41 | 1% |\n| var. TG (cm²) | 89 | 89 | - |\n| var. SAT (cm²) | 79 | 78 | no improvement |\n| var. TG-SAT (cm²) | 19 | 19 | no improvement |\n| distance TG (km) | 87 | 90 | 3% |\n\n(section-2)="} {"chunk_id": "satellite_satellite-sea-level-global_validation_q01__db6881842d51", "report_id": "satellite_satellite-sea-level-global_validation_q01", "dataset_id": "satellite-sea-level-global", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q01", "aspect_base": "validation", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Sea level measurements accuracy assessment from Satellite (observations) in the coastal zone > Analysis and results > 2. Improvement in vDT2021 version over vDT2018 dataset", "title": "Sea level measurements accuracy assessment from Satellite (observations) in the coastal zone", "chunk_index": 7, "token_count": 582, "text_raw": "The mean value of the rms difference between the vDT2018 satellite dataset and tide gauges (Table 1) is 4.41 cm, the variance of the differences (satellite – tide gauge) is 19 cm² , and the mean distance between the location of the tide gauge and the corresponding satellite data is 90 km. If these results are compared with those reported above for the comparison using the vDT2021 satellite dataset, it can be observed that the latter only improves the previous vDT2018 version in terms of the errors with tide gauges that are reduced by 1 % and the mean distance between the most correlated satellite point and tide gauges, reduced by 3 %, whereas the variance of the differences between the datasets is quite similar.\nThis fact is reflected in the spatial distribution of the differences between vDT2021 and vDT2018 consistency with tide gauges (Figure 2). A better performance of the vDT2021 dataset is obtained at the connection region between the Baltic Sea and the eastern North Atlantic Ocean and in parts of the Atlantic shore (coasts of United Kingdom and France). There is a degradation of the vDT2021 dataset in most of the stations located in the western Mediterranean Sea and the southern coasts of Spain; and in some stations located\non the coasts of France, England and Ireland. Also, negligible discrepancies between the two versions are found in the Baltic Sea. A degradation of the vDT2021 dataset is observed in most of the stations located in both the North-West Shelf (NWS) area (southern coasts of the North Sea, see Figure 2 in [[2]](https://doi.org/10.3390/rs12233970) for the location of this region) and the Norwegian Sea.\n\n![vardiff_new](https://raw.githubusercontent.com/Blanca-Fdez/c3s2-JN-assets/refs/heads/main/Assets/compTG_2sat_DT21_DT18_tot_%25vardiff_JN_BIEN-1.png?)\n\nFigure 2. Spatial distribution of the differences (vDT2021 minus vDT2018) for the mean square differences between the tide gauge and satellite sea level data. Units are the percentage of the tidal variance. Blue colours denote an improvement in the vDT2021 dataset, whilst red colours indicate its degradation with respect to the vDT2018 version.\n\n(section-3)=", "text_with_prefix": "EQC Quality Assessment: \"Sea level measurements accuracy assessment from Satellite (observations) in the coastal zone\"\nDataset: satellite-sea-level-global [CDS]\nAspect: validation_q01 | Category: Satellite_ECVs\nSection: Sea level measurements accuracy assessment from Satellite (observations) in the coastal zone > Analysis and results > 2. Improvement in vDT2021 version over vDT2018 dataset\n---\nThe mean value of the rms difference between the vDT2018 satellite dataset and tide gauges (Table 1) is 4.41 cm, the variance of the differences (satellite – tide gauge) is 19 cm² , and the mean distance between the location of the tide gauge and the corresponding satellite data is 90 km. If these results are compared with those reported above for the comparison using the vDT2021 satellite dataset, it can be observed that the latter only improves the previous vDT2018 version in terms of the errors with tide gauges that are reduced by 1 % and the mean distance between the most correlated satellite point and tide gauges, reduced by 3 %, whereas the variance of the differences between the datasets is quite similar.\nThis fact is reflected in the spatial distribution of the differences between vDT2021 and vDT2018 consistency with tide gauges (Figure 2). A better performance of the vDT2021 dataset is obtained at the connection region between the Baltic Sea and the eastern North Atlantic Ocean and in parts of the Atlantic shore (coasts of United Kingdom and France). There is a degradation of the vDT2021 dataset in most of the stations located in the western Mediterranean Sea and the southern coasts of Spain; and in some stations located\non the coasts of France, England and Ireland. Also, negligible discrepancies between the two versions are found in the Baltic Sea. A degradation of the vDT2021 dataset is observed in most of the stations located in both the North-West Shelf (NWS) area (southern coasts of the North Sea, see Figure 2 in [[2]](https://doi.org/10.3390/rs12233970) for the location of this region) and the Norwegian Sea.\n\n![vardiff_new](https://raw.githubusercontent.com/Blanca-Fdez/c3s2-JN-assets/refs/heads/main/Assets/compTG_2sat_DT21_DT18_tot_%25vardiff_JN_BIEN-1.png?)\n\nFigure 2. Spatial distribution of the differences (vDT2021 minus vDT2018) for the mean square differences between the tide gauge and satellite sea level data. Units are the percentage of the tidal variance. Blue colours denote an improvement in the vDT2021 dataset, whilst red colours indicate its degradation with respect to the vDT2018 version.\n\n(section-3)="} {"chunk_id": "satellite_satellite-sea-level-global_validation_q01__048ef66c0961", "report_id": "satellite_satellite-sea-level-global_validation_q01", "dataset_id": "satellite-sea-level-global", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q01", "aspect_base": "validation", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Sea level measurements accuracy assessment from Satellite (observations) in the coastal zone > Analysis and results > 3. Performance of vDT2021 dataset in monitoring the long-term evolution of sea level", "title": "Sea level measurements accuracy assessment from Satellite (observations) in the coastal zone", "chunk_index": 8, "token_count": 946, "text_raw": "The analyses described above were repeated for a specific time period spanning 20 years: from January 2000 to December 2019. Tide gauge time series with valid data within such a time interval were considered, allowing of the intercomparison satellite – tide gauges for long-term time series with the same length. Moreover, only tide gauge time series with at least 99 % of valid data were used in order to allow the analysis of linear trends. This reduced the original tide gauge dataset to a subset of 27 stations mainly located in the northern half of the Baltic Sea (70 % of stations) with sparse stations distributed along the coasts of France and Spain (Figure 3).\nLinear trends based on monthly observations at each tide gauge site computed from the vDT2021 satellite dataset (Figure 3, upper panel) show a homogeneous spatial pattern with overall values ranging from 2.60 to 3.80 mm yr⁻¹ in the Baltic and Mediterranean seas and between 2.40 and 3.40 mm yr⁻¹ along the North Atlantic European coasts. This promotes a mean trend for the whole domain of 3.13 mm yr⁻¹. Linear trends computed from tide gauges (Figure 3, lower panel) exhibit a more heterogeneous spatial pattern with values ranging between less than 1 mm yr⁻¹ for some stations located in the Baltic Sea and 5.06 mm yr⁻¹ for a station located in the western Mediterranean Sea. However, most of the tide gauge stations present trend values ranging from 1.30 to 3 mm yr⁻¹ with a mean trend for the whole domain of 1.96 mm yr⁻¹.\n\n![trend_TG_SAT_new](https://raw.githubusercontent.com/Blanca-Fdez/c3s2-JN-assets/refs/heads/main/Assets/trends_TG_vDT2021_JN_BIEN-2.png?)\n\nFigure 3. Spatial distribution of linear trends (mm yr⁻¹) for satellite (a) and tide gauge (b) data computed from monthly averaged data for the 20-year time period from January 2000 to December 2019. The vDT2021 dataset has been used.\n\nTrends computed from vDT2021 dataset are on average around 1.2 mm yr⁻¹ larger than those obtained from tide gauges. These discrepancies could be attributed to the heterogeneous distribution of both datasets and also the crustal land uplift due to postglacial rebound resulting from the last glacial age affecting the Baltic Basin, where most of the tide gauge stations are located. This translates into vDT2021 satellite measurements not being accurate enough in the coastal zone. These results provide further evidence, if needed, of the European seas coastal sea level rise, including the westernmost part of the Mediterranean Sea. \nOverall, linear trend differences (satellite – tide gauge) for the vDT2021 dataset varying, in absolute value, from 0.03 to 2.65 mm yr⁻¹, with an average of 1.40 mm yr⁻¹ are obtained (Figure 4). These discrepancies are lower than 1.5 mm yr⁻¹ on average and corroborate the agreement and complementarity of the two techniques to measure sea level variability in the coastal zone. However, Figure 4 displays an overall overestimation of trends from vDT2021 satellite data in the whole domain except for three tide gauge sites located on the Atlantic Spanish coast, the eastern side of the Baltic Sea, and the western Mediterranean Sea. The differences in trend could be attributed to the aforementioned factors rendering satellite measurements not accurate enough in the coastal zone.\n\n![trend_diff_DT21_TG](https://github.com/antonjes/C3S2/blob/main/trend_diff_vDT2021_TG_JN.png?raw=true)\n\nFigure 4. Spatial distribution of the differences in linear trends (mm yr⁻¹) between the vDT2021 satellite and tide gauge sea level computed for the 20-year time period from January 2000 to December 2019. Blue (red) colours denote a larger (lower) satellite linear trend.", "text_with_prefix": "EQC Quality Assessment: \"Sea level measurements accuracy assessment from Satellite (observations) in the coastal zone\"\nDataset: satellite-sea-level-global [CDS]\nAspect: validation_q01 | Category: Satellite_ECVs\nSection: Sea level measurements accuracy assessment from Satellite (observations) in the coastal zone > Analysis and results > 3. Performance of vDT2021 dataset in monitoring the long-term evolution of sea level\n---\nThe analyses described above were repeated for a specific time period spanning 20 years: from January 2000 to December 2019. Tide gauge time series with valid data within such a time interval were considered, allowing of the intercomparison satellite – tide gauges for long-term time series with the same length. Moreover, only tide gauge time series with at least 99 % of valid data were used in order to allow the analysis of linear trends. This reduced the original tide gauge dataset to a subset of 27 stations mainly located in the northern half of the Baltic Sea (70 % of stations) with sparse stations distributed along the coasts of France and Spain (Figure 3).\nLinear trends based on monthly observations at each tide gauge site computed from the vDT2021 satellite dataset (Figure 3, upper panel) show a homogeneous spatial pattern with overall values ranging from 2.60 to 3.80 mm yr⁻¹ in the Baltic and Mediterranean seas and between 2.40 and 3.40 mm yr⁻¹ along the North Atlantic European coasts. This promotes a mean trend for the whole domain of 3.13 mm yr⁻¹. Linear trends computed from tide gauges (Figure 3, lower panel) exhibit a more heterogeneous spatial pattern with values ranging between less than 1 mm yr⁻¹ for some stations located in the Baltic Sea and 5.06 mm yr⁻¹ for a station located in the western Mediterranean Sea. However, most of the tide gauge stations present trend values ranging from 1.30 to 3 mm yr⁻¹ with a mean trend for the whole domain of 1.96 mm yr⁻¹.\n\n![trend_TG_SAT_new](https://raw.githubusercontent.com/Blanca-Fdez/c3s2-JN-assets/refs/heads/main/Assets/trends_TG_vDT2021_JN_BIEN-2.png?)\n\nFigure 3. Spatial distribution of linear trends (mm yr⁻¹) for satellite (a) and tide gauge (b) data computed from monthly averaged data for the 20-year time period from January 2000 to December 2019. The vDT2021 dataset has been used.\n\nTrends computed from vDT2021 dataset are on average around 1.2 mm yr⁻¹ larger than those obtained from tide gauges. These discrepancies could be attributed to the heterogeneous distribution of both datasets and also the crustal land uplift due to postglacial rebound resulting from the last glacial age affecting the Baltic Basin, where most of the tide gauge stations are located. This translates into vDT2021 satellite measurements not being accurate enough in the coastal zone. These results provide further evidence, if needed, of the European seas coastal sea level rise, including the westernmost part of the Mediterranean Sea. \nOverall, linear trend differences (satellite – tide gauge) for the vDT2021 dataset varying, in absolute value, from 0.03 to 2.65 mm yr⁻¹, with an average of 1.40 mm yr⁻¹ are obtained (Figure 4). These discrepancies are lower than 1.5 mm yr⁻¹ on average and corroborate the agreement and complementarity of the two techniques to measure sea level variability in the coastal zone. However, Figure 4 displays an overall overestimation of trends from vDT2021 satellite data in the whole domain except for three tide gauge sites located on the Atlantic Spanish coast, the eastern side of the Baltic Sea, and the western Mediterranean Sea. The differences in trend could be attributed to the aforementioned factors rendering satellite measurements not accurate enough in the coastal zone.\n\n![trend_diff_DT21_TG](https://github.com/antonjes/C3S2/blob/main/trend_diff_vDT2021_TG_JN.png?raw=true)\n\nFigure 4. Spatial distribution of the differences in linear trends (mm yr⁻¹) between the vDT2021 satellite and tide gauge sea level computed for the 20-year time period from January 2000 to December 2019. Blue (red) colours denote a larger (lower) satellite linear trend."} {"chunk_id": "satellite_satellite-sea-level-global_validation_q01__3dbdff56d955", "report_id": "satellite_satellite-sea-level-global_validation_q01", "dataset_id": "satellite-sea-level-global", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q01", "aspect_base": "validation", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Sea level measurements accuracy assessment from Satellite (observations) in the coastal zone > Analysis and results > 3. Performance of vDT2021 dataset in monitoring the long-term evolution of sea level", "title": "Sea level measurements accuracy assessment from Satellite (observations) in the coastal zone", "chunk_index": 9, "token_count": 588, "text_raw": "2021 satellite and tide gauge sea level computed for the 20-year time period from January 2000 to December 2019. Blue (red) colours denote a larger (lower) satellite linear trend.\n\nLinear trends computed from the vDT2018 dataset (figure not shown) exhibit quite a similar spatial pattern to that reported for the vDT2021 version with overall values ranging from 2.40 to 3.60 mm yr⁻¹ in the Baltic and Mediterranean seas and between 2.10 and 2.85 mm yr⁻¹ along the North Atlantic European coasts. This promotes a mean trend for the whole domain of 2.85 mm yr⁻¹. Thus, some differences in range are observed between the two versions, with a lower variability observed for the vDT2018 dataset. \nThis fact has an impact on the spatial distribution of the differences between the vDT2021 and vDT2018 versions (Figure 5). It depicts a homogeneous spatial pattern with overall larger trends for the vDT2021 dataset except for a tide gauge station located in the connection region between the Baltic Sea and the eastern North Atlantic Ocean. Figure 5 also reveals, when making a comparison with linear trends from tide gauges, that the vDT2021 dataset presents larger differences with tide gauges with respect to the vDT2018 version in the whole domain. Nevertheless, conclusions should be drawn with caution, as these differences are not statistically significant and fall within the margin of 0.8 mm yr⁻¹ error of local sea level trends in Europe obtained by [[20]](https://doi.org/10.1038/s41597-020-00786-7).\n\n![trend_diff_DT21_DT18](https://github.com/antonjes/C3S2/blob/main/trend_diff_vDT2021_vDT2018.png?raw=true)\n\nFigure 5. Panel b of Figure 5 from [[8]](https://doi.org/10.5194/os-19-793-2023). Spatial distribution of the differences (vDT2021 minus vDT2018 satellite dataset) for linear trends (mm yr⁻¹) for satellite data. Monthly averaged data for the 20-year time period from January 2000 to December 2019 have been used. Blue (red) colours denote lower (larger) trends for the vDT2021 dataset.\n\n(section-4)=", "text_with_prefix": "EQC Quality Assessment: \"Sea level measurements accuracy assessment from Satellite (observations) in the coastal zone\"\nDataset: satellite-sea-level-global [CDS]\nAspect: validation_q01 | Category: Satellite_ECVs\nSection: Sea level measurements accuracy assessment from Satellite (observations) in the coastal zone > Analysis and results > 3. Performance of vDT2021 dataset in monitoring the long-term evolution of sea level\n---\n2021 satellite and tide gauge sea level computed for the 20-year time period from January 2000 to December 2019. Blue (red) colours denote a larger (lower) satellite linear trend.\n\nLinear trends computed from the vDT2018 dataset (figure not shown) exhibit quite a similar spatial pattern to that reported for the vDT2021 version with overall values ranging from 2.40 to 3.60 mm yr⁻¹ in the Baltic and Mediterranean seas and between 2.10 and 2.85 mm yr⁻¹ along the North Atlantic European coasts. This promotes a mean trend for the whole domain of 2.85 mm yr⁻¹. Thus, some differences in range are observed between the two versions, with a lower variability observed for the vDT2018 dataset. \nThis fact has an impact on the spatial distribution of the differences between the vDT2021 and vDT2018 versions (Figure 5). It depicts a homogeneous spatial pattern with overall larger trends for the vDT2021 dataset except for a tide gauge station located in the connection region between the Baltic Sea and the eastern North Atlantic Ocean. Figure 5 also reveals, when making a comparison with linear trends from tide gauges, that the vDT2021 dataset presents larger differences with tide gauges with respect to the vDT2018 version in the whole domain. Nevertheless, conclusions should be drawn with caution, as these differences are not statistically significant and fall within the margin of 0.8 mm yr⁻¹ error of local sea level trends in Europe obtained by [[20]](https://doi.org/10.1038/s41597-020-00786-7).\n\n![trend_diff_DT21_DT18](https://github.com/antonjes/C3S2/blob/main/trend_diff_vDT2021_vDT2018.png?raw=true)\n\nFigure 5. Panel b of Figure 5 from [[8]](https://doi.org/10.5194/os-19-793-2023). Spatial distribution of the differences (vDT2021 minus vDT2018 satellite dataset) for linear trends (mm yr⁻¹) for satellite data. Monthly averaged data for the 20-year time period from January 2000 to December 2019 have been used. Blue (red) colours denote lower (larger) trends for the vDT2021 dataset.\n\n(section-4)="} {"chunk_id": "satellite_satellite-sea-level-global_validation_q01__075e0e5de8aa", "report_id": "satellite_satellite-sea-level-global_validation_q01", "dataset_id": "satellite-sea-level-global", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q01", "aspect_base": "validation", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Sea level measurements accuracy assessment from Satellite (observations) in the coastal zone > Analysis and results > 4. Conclusions", "title": "Sea level measurements accuracy assessment from Satellite (observations) in the coastal zone", "chunk_index": 10, "token_count": 1021, "text_raw": "The mean value of the rms difference between the vDT2021 satellite dataset and tide gauges along the coastal band of the European Seas is 4.35 cm, the variance of the differences (satellite – tide gauge) is 19 cm² whilst the variance of the satellite and tide gauge datasets are 89 cm² and 79 cm², respectively. Thus, this dataset properly captures the sea level in the coastal region of the European Seas and its variability when compared with in-situ tide gauge data. In addition, the mean distance between the location of the tide gauges and the corresponding satellite data with the highest correlation is 87 km.\n\nThe mean value of the rms difference between the vDT2018 version and tide gauges is 4.41 cm, the variance of the differences (altimetry–tide gauge) is 19 cm², and the mean distance between the location of the tide gauge and the corresponding satellite data is 90 km. Thus, compared to its predecessor, vDT2018, the vDT201 dataset shows only slight improvements. The average difference with tide gauges decreased by just 1%, and the mean distance between tide gauges and the most correlated satellite data points was reduced by 3% (from 90 km to 87 km), whereas the variance of the differences between the datasets is quite similar.\n\nConsistency between the satellite dataset vDT2021 and the tide gauge data shows spatial variability. In the central and eastern Baltic Sea, differences are under 5%, confirming the accuracy of the satellite corrections applied in this area. However, in regions connected to the North Atlantic, such as the Strait of Gibraltar and along the Atlantic shore, discrepancies increase, ranging between 20% and 50%. This could be due to inaccuracies in corrections (i.e. ocean tide and DAC) applied to the satellite data, as well as greater variability in non-tidal sea level changes. The Norwegian Sea shows better results than those obtained for the Atlantic shore with consistency values ranging between 15 % and 30 %, except for the south-western part of Norway where even better values ranging between 5 % and 15% are obtained providing accurate sea level measurements when compared with tide gauges.\n\nLinear trends derived from the vDT2021 satellite dataset indicate a consistent sea level rise along the European coasts and the western Mediterranean, with values ranging from 2.60 to 3.80 mm yr⁻¹ and a mean trend for the whole domain of 3.13 mm yr⁻¹. In contrast, trends computed from tide gauge records display a more heterogeneous spatial pattern, with values ranging between 1.30 and 3 mm yr⁻¹ and a mean trend for the whole domain of 1.96 mm yr⁻¹. These results provide further evidence of the European seas coastal sea level rise, including the westernmost part of the Mediterranean Sea. Overall, linear trend differences (satellite – tide gauge) for the vDT2021 dataset vary, in absolute value, from 0.03 to 2.65 mm yr⁻¹, with an average of 1.40 mm yr⁻¹. These low discrepancies corroborate the agreement and complementarity of the two techniques to measure sea level variability in the coastal zone. For vDT2018 dataset, the mean trend for the whole domain (2.85 mm yr⁻¹) shows smaller differences with respect to the tide gauges than vDT2021. This suggests that the vDT2021 satellite measurements contain nearshore errors that may affect the accuracy of regional mean sea level estimates in coastal zones, leading to discrepancies with tide gauge trends. For studies focused on long-term trends, [[20]](https://doi.org/10.1038/s41597-020-00786-7) shows that over the altimetric period (1993–2019), the average uncertainty of local sea level trends in Europe was 0.8 mm yr⁻¹. In this study, the differences between the vDT2021 and vDT2018 trends are generally < 0.8 mm yr⁻¹. Therefore, conclusions should be drawn with caution, because these differences are not statistically significant and fall within the margin of error obtained by [[20]](https://doi.org/10.1038/s41597-020-00786-7). Consequently, we cannot say with complete certainty which version is better.\n\nThis assessment is mostly based on published material comparing versions vDT2018 and vDT2021, and it does not include the newest version vDT2024.", "text_with_prefix": "EQC Quality Assessment: \"Sea level measurements accuracy assessment from Satellite (observations) in the coastal zone\"\nDataset: satellite-sea-level-global [CDS]\nAspect: validation_q01 | Category: Satellite_ECVs\nSection: Sea level measurements accuracy assessment from Satellite (observations) in the coastal zone > Analysis and results > 4. Conclusions\n---\nThe mean value of the rms difference between the vDT2021 satellite dataset and tide gauges along the coastal band of the European Seas is 4.35 cm, the variance of the differences (satellite – tide gauge) is 19 cm² whilst the variance of the satellite and tide gauge datasets are 89 cm² and 79 cm², respectively. Thus, this dataset properly captures the sea level in the coastal region of the European Seas and its variability when compared with in-situ tide gauge data. In addition, the mean distance between the location of the tide gauges and the corresponding satellite data with the highest correlation is 87 km.\n\nThe mean value of the rms difference between the vDT2018 version and tide gauges is 4.41 cm, the variance of the differences (altimetry–tide gauge) is 19 cm², and the mean distance between the location of the tide gauge and the corresponding satellite data is 90 km. Thus, compared to its predecessor, vDT2018, the vDT201 dataset shows only slight improvements. The average difference with tide gauges decreased by just 1%, and the mean distance between tide gauges and the most correlated satellite data points was reduced by 3% (from 90 km to 87 km), whereas the variance of the differences between the datasets is quite similar.\n\nConsistency between the satellite dataset vDT2021 and the tide gauge data shows spatial variability. In the central and eastern Baltic Sea, differences are under 5%, confirming the accuracy of the satellite corrections applied in this area. However, in regions connected to the North Atlantic, such as the Strait of Gibraltar and along the Atlantic shore, discrepancies increase, ranging between 20% and 50%. This could be due to inaccuracies in corrections (i.e. ocean tide and DAC) applied to the satellite data, as well as greater variability in non-tidal sea level changes. The Norwegian Sea shows better results than those obtained for the Atlantic shore with consistency values ranging between 15 % and 30 %, except for the south-western part of Norway where even better values ranging between 5 % and 15% are obtained providing accurate sea level measurements when compared with tide gauges.\n\nLinear trends derived from the vDT2021 satellite dataset indicate a consistent sea level rise along the European coasts and the western Mediterranean, with values ranging from 2.60 to 3.80 mm yr⁻¹ and a mean trend for the whole domain of 3.13 mm yr⁻¹. In contrast, trends computed from tide gauge records display a more heterogeneous spatial pattern, with values ranging between 1.30 and 3 mm yr⁻¹ and a mean trend for the whole domain of 1.96 mm yr⁻¹. These results provide further evidence of the European seas coastal sea level rise, including the westernmost part of the Mediterranean Sea. Overall, linear trend differences (satellite – tide gauge) for the vDT2021 dataset vary, in absolute value, from 0.03 to 2.65 mm yr⁻¹, with an average of 1.40 mm yr⁻¹. These low discrepancies corroborate the agreement and complementarity of the two techniques to measure sea level variability in the coastal zone. For vDT2018 dataset, the mean trend for the whole domain (2.85 mm yr⁻¹) shows smaller differences with respect to the tide gauges than vDT2021. This suggests that the vDT2021 satellite measurements contain nearshore errors that may affect the accuracy of regional mean sea level estimates in coastal zones, leading to discrepancies with tide gauge trends. For studies focused on long-term trends, [[20]](https://doi.org/10.1038/s41597-020-00786-7) shows that over the altimetric period (1993–2019), the average uncertainty of local sea level trends in Europe was 0.8 mm yr⁻¹. In this study, the differences between the vDT2021 and vDT2018 trends are generally < 0.8 mm yr⁻¹. Therefore, conclusions should be drawn with caution, because these differences are not statistically significant and fall within the margin of error obtained by [[20]](https://doi.org/10.1038/s41597-020-00786-7). Consequently, we cannot say with complete certainty which version is better.\n\nThis assessment is mostly based on published material comparing versions vDT2018 and vDT2021, and it does not include the newest version vDT2024."} {"chunk_id": "satellite_satellite-sea-level-global_validation_q01__ace75e365dd7", "report_id": "satellite_satellite-sea-level-global_validation_q01", "dataset_id": "satellite-sea-level-global", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q01", "aspect_base": "validation", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Sea level measurements accuracy assessment from Satellite (observations) in the coastal zone > ℹ️ If you want to know more > Key resources", "title": "Sea level measurements accuracy assessment from Satellite (observations) in the coastal zone", "chunk_index": 11, "token_count": 302, "text_raw": "* CDS entry: [\"Sea level gridded data from satellite observations for the global ocean from 1993 to present\"](https://cds.climate.copernicus.eu/datasets/satellite-sea-level-global?tab=overview)\n* Product User Guide and Specification (PUGS): [vDT2018](https://dast.copernicus-climate.eu/documents/satellite-sea-level/vDT2018/D3.SL.1-v1.2_PUGS_of_v1DT2018_SeaLevel_products_v2.4.pdf) and [vDT2021](https://confluence.ecmwf.int/pages/viewpage.action?pageId=333790908)\n* [Tide gauge dataset](http://www.marineinsitu.eu)\n* [Dynamic Atmospheric Correction dataset (DAC)](https://www.aviso.altimetry.fr/en/data/products/auxiliary-products/dynamic-atmospheric-correction.html)\n* Copernicus climate change indicators: [sea level](https://climate.copernicus.eu/climate-indicators/sea-level)", "text_with_prefix": "EQC Quality Assessment: \"Sea level measurements accuracy assessment from Satellite (observations) in the coastal zone\"\nDataset: satellite-sea-level-global [CDS]\nAspect: validation_q01 | Category: Satellite_ECVs\nSection: Sea level measurements accuracy assessment from Satellite (observations) in the coastal zone > ℹ️ If you want to know more > Key resources\n---\n* CDS entry: [\"Sea level gridded data from satellite observations for the global ocean from 1993 to present\"](https://cds.climate.copernicus.eu/datasets/satellite-sea-level-global?tab=overview)\n* Product User Guide and Specification (PUGS): [vDT2018](https://dast.copernicus-climate.eu/documents/satellite-sea-level/vDT2018/D3.SL.1-v1.2_PUGS_of_v1DT2018_SeaLevel_products_v2.4.pdf) and [vDT2021](https://confluence.ecmwf.int/pages/viewpage.action?pageId=333790908)\n* [Tide gauge dataset](http://www.marineinsitu.eu)\n* [Dynamic Atmospheric Correction dataset (DAC)](https://www.aviso.altimetry.fr/en/data/products/auxiliary-products/dynamic-atmospheric-correction.html)\n* Copernicus climate change indicators: [sea level](https://climate.copernicus.eu/climate-indicators/sea-level)"} {"chunk_id": "satellite_satellite-sea-level-global_validation_q01__c0335431a83f", "report_id": "satellite_satellite-sea-level-global_validation_q01", "dataset_id": "satellite-sea-level-global", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q01", "aspect_base": "validation", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Sea level measurements accuracy assessment from Satellite (observations) in the coastal zone > ℹ️ If you want to know more > References", "title": "Sea level measurements accuracy assessment from Satellite (observations) in the coastal zone", "chunk_index": 12, "token_count": 1077, "text_raw": "[[1]](https://doi.org/10.1007/s10712-019-09569-1) Vignudelli, S., Birol, F., Benveniste, J., Fu, L.-L., Picot, N., Raynal, M., and Roinard, H.: Satellite Altimetry Measurements of Sea Level in the Coastal Zone, Surv. Geophys., 40, 1319–1349, https://doi.org/10.1007/s10712-019-09569-1, 2019\n\n[[2]](https://doi.org/10.3390/rs12233970) Sánchez-Román, A., Pascual, A., Pujol, M.-I., Taburet, G., Marcos, M., and Faugère, Y.: Assessment of DUACS Sentinel-3A Altimetry Data in the Coastal Band of the European Seas: Comparison with Tide Gauge Measurements, Remote Sens., 12, 3970, https://doi.org/10.3390/rs12233970, 2020\n\n[[3]](https://doi.org/10.3390/rs15030793) Pujol, M.-I., Dupuy, S., Vergara, O., Sánchez Román, A., Faugère, Y., Prandi, P., Dabat, M.-L., Dagneaux, Q., Lievin, M., Cadier, E., Dibarboure, G., and Picot, N.: Refining the Resolution of DUACS Along-Track Level-3 Sea Level Altimetry Products, Remote Sens., 15, 1–30, https://doi.org/10.3390/rs15030793, 2023\n\n[[4]](https://doi.org/10.1007/s10712-016-9392-0) Cipollini, P., Calafat, F.-M., Jevrejeva, S., Melet, A., and Prandi, P.: Monitoring sea level in the coastal zone with satellite altimetry and tide gauges, Surv. Geophys., 38, 33–57, https://doi.org/10.1007/s10712-016-9392-0, 2017\n\n[[5]](https://documentation.marine.copernicus.eu/QUID/CMEMS-SL-QUID-008-032-068.pdf) CMEMS-SL-QUID: QUID document for Sea Level TAC DUACS products CMEMS-SL-QUID-008-032-068, https://documentation.marine.copernicus.eu/QUID/CMEMS-SL-QUID-008-032-068.pdf , 2022\n\n[[6]](https://doi.org/10.1002/2015RG000502) Wöppelmann, G. and Marcos, M.: Vertical land motion as a key to understanding sea level change and variability, Rev. Geophys., 54, 64–92, https://doi.org/10.1002/2015RG000502, 2016\n\n[[7]](https://doi.org/10.5194/os-15-1207-2019) Taburet, G., Sanchez-Roman, A., Ballarotta, M., Pujol, M.-I., Legeais, J.-F., Fournier, F., Faugere, Y., and Dibarboure, G.: DUACS DT2018: 25 years of reprocessed sea level altimetry products, Ocean Sci., 15, 1207–1224, https://doi.org/10.5194/os-15-1207-2019, 2019\n\n[[8]](https://doi.org/10.5194/os-19-793-2023) Sánchez-Román, A., Pujol, M. I., Faugère, Y., and Pascual, A.: DUACS DT2021 reprocessed altimetry improves sea level retrieval in the coastal band of the European seas, Ocean Sci., 19, 793–809, https://doi.org/10.5194/os-19-793-2023, 2023\n\n[[9]](https://doi.org/10.1029/2007jc004459) Pascual, A., Marcos, M., and Gomis, D.: Comparing the sea level response to pressure and wind forcing of two barotropic models: validation with tide gauge and altimetry data, J. Geophys. Res., 113, C07011, https://doi.org/10.1029/2007jc004459, 2008", "text_with_prefix": "EQC Quality Assessment: \"Sea level measurements accuracy assessment from Satellite (observations) in the coastal zone\"\nDataset: satellite-sea-level-global [CDS]\nAspect: validation_q01 | Category: Satellite_ECVs\nSection: Sea level measurements accuracy assessment from Satellite (observations) in the coastal zone > ℹ️ If you want to know more > References\n---\n[[1]](https://doi.org/10.1007/s10712-019-09569-1) Vignudelli, S., Birol, F., Benveniste, J., Fu, L.-L., Picot, N., Raynal, M., and Roinard, H.: Satellite Altimetry Measurements of Sea Level in the Coastal Zone, Surv. Geophys., 40, 1319–1349, https://doi.org/10.1007/s10712-019-09569-1, 2019\n\n[[2]](https://doi.org/10.3390/rs12233970) Sánchez-Román, A., Pascual, A., Pujol, M.-I., Taburet, G., Marcos, M., and Faugère, Y.: Assessment of DUACS Sentinel-3A Altimetry Data in the Coastal Band of the European Seas: Comparison with Tide Gauge Measurements, Remote Sens., 12, 3970, https://doi.org/10.3390/rs12233970, 2020\n\n[[3]](https://doi.org/10.3390/rs15030793) Pujol, M.-I., Dupuy, S., Vergara, O., Sánchez Román, A., Faugère, Y., Prandi, P., Dabat, M.-L., Dagneaux, Q., Lievin, M., Cadier, E., Dibarboure, G., and Picot, N.: Refining the Resolution of DUACS Along-Track Level-3 Sea Level Altimetry Products, Remote Sens., 15, 1–30, https://doi.org/10.3390/rs15030793, 2023\n\n[[4]](https://doi.org/10.1007/s10712-016-9392-0) Cipollini, P., Calafat, F.-M., Jevrejeva, S., Melet, A., and Prandi, P.: Monitoring sea level in the coastal zone with satellite altimetry and tide gauges, Surv. Geophys., 38, 33–57, https://doi.org/10.1007/s10712-016-9392-0, 2017\n\n[[5]](https://documentation.marine.copernicus.eu/QUID/CMEMS-SL-QUID-008-032-068.pdf) CMEMS-SL-QUID: QUID document for Sea Level TAC DUACS products CMEMS-SL-QUID-008-032-068, https://documentation.marine.copernicus.eu/QUID/CMEMS-SL-QUID-008-032-068.pdf , 2022\n\n[[6]](https://doi.org/10.1002/2015RG000502) Wöppelmann, G. and Marcos, M.: Vertical land motion as a key to understanding sea level change and variability, Rev. Geophys., 54, 64–92, https://doi.org/10.1002/2015RG000502, 2016\n\n[[7]](https://doi.org/10.5194/os-15-1207-2019) Taburet, G., Sanchez-Roman, A., Ballarotta, M., Pujol, M.-I., Legeais, J.-F., Fournier, F., Faugere, Y., and Dibarboure, G.: DUACS DT2018: 25 years of reprocessed sea level altimetry products, Ocean Sci., 15, 1207–1224, https://doi.org/10.5194/os-15-1207-2019, 2019\n\n[[8]](https://doi.org/10.5194/os-19-793-2023) Sánchez-Román, A., Pujol, M. I., Faugère, Y., and Pascual, A.: DUACS DT2021 reprocessed altimetry improves sea level retrieval in the coastal band of the European seas, Ocean Sci., 19, 793–809, https://doi.org/10.5194/os-19-793-2023, 2023\n\n[[9]](https://doi.org/10.1029/2007jc004459) Pascual, A., Marcos, M., and Gomis, D.: Comparing the sea level response to pressure and wind forcing of two barotropic models: validation with tide gauge and altimetry data, J. Geophys. Res., 113, C07011, https://doi.org/10.1029/2007jc004459, 2008"} {"chunk_id": "satellite_satellite-sea-level-global_validation_q01__e1bb90efea07", "report_id": "satellite_satellite-sea-level-global_validation_q01", "dataset_id": "satellite-sea-level-global", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q01", "aspect_base": "validation", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Sea level measurements accuracy assessment from Satellite (observations) in the coastal zone > ℹ️ If you want to know more > References", "title": "Sea level measurements accuracy assessment from Satellite (observations) in the coastal zone", "chunk_index": 13, "token_count": 1082, "text_raw": "forcing of two barotropic models: validation with tide gauge and altimetry data, J. Geophys. Res., 113, C07011, https://doi.org/10.1029/2007jc004459, 2008\n\n[[10]](https://doi.org/10.1016/j.jmarsys.2016.03.006) Laíz, I., Tejedor, B., Gómez-Enri, J., Aboitiz, A., and Villares, P.: Contributions to the sea level seasonal cycle within the Gulf of Cadiz (Southwestern Iberian Peninsula), J. Mar. Syst., 159, 55– 66, https://doi.org/10.1016/j.jmarsys.2016.03.006, 2016\n\n[[11]](https://doi.org/10.1080/1755876X.2018.1489208) Von Schuckmann, K., Le Traon, P.-Y., Smith, N., et al.: Copernicus Marine Service Ocean State Report, J. Oper. Ocean., 11, S1– S142, https://doi.org/10.1080/1755876X.2018.1489208, 2018\n\n[[12]](https://www.po.gso.uri.edu/codiga/utide/2011Codiga-UTide-Report.pdf) Codiga, D. L.: Unified Tidal Analysis and Prediction Using the UTide Matlab Functions, Technical Report 2011-01, Graduate School of Oceanography, University of Rhode Island, Narragansett, RI, 59 pp., https://www.po.gso.uri.edu/codiga/utide/2011Codiga-UTide-Report.pdf , 2011\n\n[[13]](https://doi.org/10.1175/1520-0426%281999%29016<1279:EOGMAP>2.0.CO;2) Dorandeu, J. and Le Traon, P.-Y.: Effects of global mean atmospheric pressure variations on mean sea level changes from Topex/Poseidon, J. Atmos. Ocean. Technol., 16, 1279–1283, [](https://doi.org/10.1175/1520-0426%281999%29016<1279:EOGMAP>2.0.CO;2), 1999\n\n[[14]](https://doi.org/10.1029/2002GL016473) Carrère, L. and Lyard, F.: Modeling the barotropic response of the global ocean to atmospheric wind and pressure forcing–comparisons with observations, Geophys. Res. Lett., 30, 1275, https://doi.org/10.1029/2002GL016473, 2003\n\n[[15]](https://doi.org/10.1016/j.asr.2015.04.027) Marcos, M., Pascual, A., and Pujol, I.: Improved satellite altimeter mapped sea level anomalies in the Mediterranean Sea: A comparison with tide gauges, Adv. Space Res., 56, 596–604, https://doi.org/10.1016/j.asr.2015.04.027, 2015\n\n[[16]](https://doi.org/10.1029/98RG02638) Peltier, W. R.: Postglacial Variations in the Level of the Sea: Implications for Climate Dynamics and Solid-Earth Geophysics, Rev. Geophys., 36, 603–689, https://doi.org/10.1029/98RG02638, 1998\n\n[[17]](https://doi.org/10.1146/annurev.earth.32.082503.144359) Peltier, W. R.: Global Glacial Isostasy and the Surface of the Ice-Age Earth: The ICE-5G(VM2) model and GRACE, Ann. Rev. Earth. Planet. Sci., 32, 111–149, https://doi.org/10.1146/annurev.earth.32.082503.144359, 2004\n\n[[18]](https://doi.org/10.1175/2008JTECHO556.1) Pascual, A., Boone, C., Larnicol, G., and Le Traon, P. Y.: On the quality of real-time altimeter gridded fields: comparison with in situ data, J. Atmos. Ocean. Technol., 26, 556–569, https://doi.org/10.1175/2008JTECHO556.1, 2009", "text_with_prefix": "EQC Quality Assessment: \"Sea level measurements accuracy assessment from Satellite (observations) in the coastal zone\"\nDataset: satellite-sea-level-global [CDS]\nAspect: validation_q01 | Category: Satellite_ECVs\nSection: Sea level measurements accuracy assessment from Satellite (observations) in the coastal zone > ℹ️ If you want to know more > References\n---\nforcing of two barotropic models: validation with tide gauge and altimetry data, J. Geophys. Res., 113, C07011, https://doi.org/10.1029/2007jc004459, 2008\n\n[[10]](https://doi.org/10.1016/j.jmarsys.2016.03.006) Laíz, I., Tejedor, B., Gómez-Enri, J., Aboitiz, A., and Villares, P.: Contributions to the sea level seasonal cycle within the Gulf of Cadiz (Southwestern Iberian Peninsula), J. Mar. Syst., 159, 55– 66, https://doi.org/10.1016/j.jmarsys.2016.03.006, 2016\n\n[[11]](https://doi.org/10.1080/1755876X.2018.1489208) Von Schuckmann, K., Le Traon, P.-Y., Smith, N., et al.: Copernicus Marine Service Ocean State Report, J. Oper. Ocean., 11, S1– S142, https://doi.org/10.1080/1755876X.2018.1489208, 2018\n\n[[12]](https://www.po.gso.uri.edu/codiga/utide/2011Codiga-UTide-Report.pdf) Codiga, D. L.: Unified Tidal Analysis and Prediction Using the UTide Matlab Functions, Technical Report 2011-01, Graduate School of Oceanography, University of Rhode Island, Narragansett, RI, 59 pp., https://www.po.gso.uri.edu/codiga/utide/2011Codiga-UTide-Report.pdf , 2011\n\n[[13]](https://doi.org/10.1175/1520-0426%281999%29016<1279:EOGMAP>2.0.CO;2) Dorandeu, J. and Le Traon, P.-Y.: Effects of global mean atmospheric pressure variations on mean sea level changes from Topex/Poseidon, J. Atmos. Ocean. Technol., 16, 1279–1283, [](https://doi.org/10.1175/1520-0426%281999%29016<1279:EOGMAP>2.0.CO;2), 1999\n\n[[14]](https://doi.org/10.1029/2002GL016473) Carrère, L. and Lyard, F.: Modeling the barotropic response of the global ocean to atmospheric wind and pressure forcing–comparisons with observations, Geophys. Res. Lett., 30, 1275, https://doi.org/10.1029/2002GL016473, 2003\n\n[[15]](https://doi.org/10.1016/j.asr.2015.04.027) Marcos, M., Pascual, A., and Pujol, I.: Improved satellite altimeter mapped sea level anomalies in the Mediterranean Sea: A comparison with tide gauges, Adv. Space Res., 56, 596–604, https://doi.org/10.1016/j.asr.2015.04.027, 2015\n\n[[16]](https://doi.org/10.1029/98RG02638) Peltier, W. R.: Postglacial Variations in the Level of the Sea: Implications for Climate Dynamics and Solid-Earth Geophysics, Rev. Geophys., 36, 603–689, https://doi.org/10.1029/98RG02638, 1998\n\n[[17]](https://doi.org/10.1146/annurev.earth.32.082503.144359) Peltier, W. R.: Global Glacial Isostasy and the Surface of the Ice-Age Earth: The ICE-5G(VM2) model and GRACE, Ann. Rev. Earth. Planet. Sci., 32, 111–149, https://doi.org/10.1146/annurev.earth.32.082503.144359, 2004\n\n[[18]](https://doi.org/10.1175/2008JTECHO556.1) Pascual, A., Boone, C., Larnicol, G., and Le Traon, P. Y.: On the quality of real-time altimeter gridded fields: comparison with in situ data, J. Atmos. Ocean. Technol., 26, 556–569, https://doi.org/10.1175/2008JTECHO556.1, 2009"} {"chunk_id": "satellite_satellite-sea-level-global_validation_q01__102a56bf8fdb", "report_id": "satellite_satellite-sea-level-global_validation_q01", "dataset_id": "satellite-sea-level-global", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q01", "aspect_base": "validation", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Sea level measurements accuracy assessment from Satellite (observations) in the coastal zone > ℹ️ If you want to know more > References", "title": "Sea level measurements accuracy assessment from Satellite (observations) in the coastal zone", "chunk_index": 14, "token_count": 377, "text_raw": "fields: comparison with in situ data, J. Atmos. Ocean. Technol., 26, 556–569, https://doi.org/10.1175/2008JTECHO556.1, 2009\n\n[[19]](https://doi.org/10.5194/os-15-1091-2019) Ballarotta, M., Ubelmann, C., Pujol, M.-I., Taburet, G., Fournier, F., Legeais, J.-F., Faugère, Y., Delepoulle, A., Chelton, D., Dibarboure, G., and Picot, N.: On the resolutions of ocean altimetry maps, Ocean Sci., 15, 1091–1109, https://doi.org/10.5194/os-15-1091-2019, 2019\n\n[[20]](https://doi.org/10.1038/s41597-020-00786-7) Prandi, P., Meyssignac, B., Ablain, M., Spada, G., Ribes, A., and Benveniste, J.: Local sea level trends, accelerations and uncertainties over 1993–2019. Scientific Data, 8(1), 1, https://doi.org/10.1038/s41597-020-00786-7, 2021", "text_with_prefix": "EQC Quality Assessment: \"Sea level measurements accuracy assessment from Satellite (observations) in the coastal zone\"\nDataset: satellite-sea-level-global [CDS]\nAspect: validation_q01 | Category: Satellite_ECVs\nSection: Sea level measurements accuracy assessment from Satellite (observations) in the coastal zone > ℹ️ If you want to know more > References\n---\nfields: comparison with in situ data, J. Atmos. Ocean. Technol., 26, 556–569, https://doi.org/10.1175/2008JTECHO556.1, 2009\n\n[[19]](https://doi.org/10.5194/os-15-1091-2019) Ballarotta, M., Ubelmann, C., Pujol, M.-I., Taburet, G., Fournier, F., Legeais, J.-F., Faugère, Y., Delepoulle, A., Chelton, D., Dibarboure, G., and Picot, N.: On the resolutions of ocean altimetry maps, Ocean Sci., 15, 1091–1109, https://doi.org/10.5194/os-15-1091-2019, 2019\n\n[[20]](https://doi.org/10.1038/s41597-020-00786-7) Prandi, P., Meyssignac, B., Ablain, M., Spada, G., Ribes, A., and Benveniste, J.: Local sea level trends, accelerations and uncertainties over 1993–2019. Scientific Data, 8(1), 1, https://doi.org/10.1038/s41597-020-00786-7, 2021"} {"chunk_id": "satellite_satellite-sea-surface-temperature_consistency_q01__430562af1f9b", "report_id": "satellite_satellite-sea-surface-temperature_consistency_q01", "dataset_id": "satellite-sea-surface-temperature", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q01", "aspect_base": "consistency", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Consistency Assessment of Satellite Sea Surface Temperature for Climate Monitoring > Quality assessment question(s)", "title": "Consistency Assessment of Satellite Sea Surface Temperature for Climate Monitoring", "chunk_index": 0, "token_count": 226, "text_raw": "* **Do the ESA CCI SST L4 v2.1 and GHRSST Multi-Product Ensemble (GMPE) SST datasets provide consistent representations of SST climatology and its variability?**\n\nWe aim here to evaluate the consistency and representativeness of sea surface temperature (SST) climatology and its variability in satellite-based long-term climate data records (CDRs). \nTo this end, two different SST CDRs have been intercompared on a 30-year long (1982-2011) period. \nThe full period of common coverage (1982-2016) was used when extracting time series, to have more information on each dataset's capability to properly represent large scale climatic oscillations, such as the El-Nino Southern Oscillation (ENSO).", "text_with_prefix": "EQC Quality Assessment: \"Consistency Assessment of Satellite Sea Surface Temperature for Climate Monitoring\"\nDataset: satellite-sea-surface-temperature [CDS]\nAspect: consistency_q01 | Category: Satellite_ECVs\nSection: Consistency Assessment of Satellite Sea Surface Temperature for Climate Monitoring > Quality assessment question(s)\n---\n* **Do the ESA CCI SST L4 v2.1 and GHRSST Multi-Product Ensemble (GMPE) SST datasets provide consistent representations of SST climatology and its variability?**\n\nWe aim here to evaluate the consistency and representativeness of sea surface temperature (SST) climatology and its variability in satellite-based long-term climate data records (CDRs). \nTo this end, two different SST CDRs have been intercompared on a 30-year long (1982-2011) period. \nThe full period of common coverage (1982-2016) was used when extracting time series, to have more information on each dataset's capability to properly represent large scale climatic oscillations, such as the El-Nino Southern Oscillation (ENSO)."} {"chunk_id": "satellite_satellite-sea-surface-temperature_consistency_q01__4eabe983e74a", "report_id": "satellite_satellite-sea-surface-temperature_consistency_q01", "dataset_id": "satellite-sea-surface-temperature", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q01", "aspect_base": "consistency", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Consistency Assessment of Satellite Sea Surface Temperature for Climate Monitoring > Quality assessment statement", "title": "Consistency Assessment of Satellite Sea Surface Temperature for Climate Monitoring", "chunk_index": 1, "token_count": 303, "text_raw": "These are the key outcomes of this assessment\n\n* Both datasets provide consistent results, well reproducing SST mean values and the main patterns of SST variability on a variety of scales, ranging from seasonal to (multi-)annual time scales;\n* Both datasets can be exploited for climate studies. As an example, El-Niño/La-Niña events are well characterized by both GMPE and ESA CCI SST L4 v2.1 products\n;\n* ESA CCI SST L4 v2.1 dataset should be used with caution in the first decade of the CDR. Significant discrepancies with respect to the ensemble mean GMPE dataset were indeed observed in the 1982-1992 period;\n```\n\nattachment:bb2e502f-9d8a-4404-8c19-418886e6e483.png\n---\nheight: 400px\nwidth: 550px\n---\n*Nino 3.4 Time series calculated according to [[6]](https://climatedataguide.ucar.edu/climate-data/nino-sst-indices-nino-12-3-34-4-oni-and-tni)*\n```", "text_with_prefix": "EQC Quality Assessment: \"Consistency Assessment of Satellite Sea Surface Temperature for Climate Monitoring\"\nDataset: satellite-sea-surface-temperature [CDS]\nAspect: consistency_q01 | Category: Satellite_ECVs\nSection: Consistency Assessment of Satellite Sea Surface Temperature for Climate Monitoring > Quality assessment statement\n---\nThese are the key outcomes of this assessment\n\n* Both datasets provide consistent results, well reproducing SST mean values and the main patterns of SST variability on a variety of scales, ranging from seasonal to (multi-)annual time scales;\n* Both datasets can be exploited for climate studies. As an example, El-Niño/La-Niña events are well characterized by both GMPE and ESA CCI SST L4 v2.1 products\n;\n* ESA CCI SST L4 v2.1 dataset should be used with caution in the first decade of the CDR. Significant discrepancies with respect to the ensemble mean GMPE dataset were indeed observed in the 1982-1992 period;\n```\n\nattachment:bb2e502f-9d8a-4404-8c19-418886e6e483.png\n---\nheight: 400px\nwidth: 550px\n---\n*Nino 3.4 Time series calculated according to [[6]](https://climatedataguide.ucar.edu/climate-data/nino-sst-indices-nino-12-3-34-4-oni-and-tni)*\n```"} {"chunk_id": "satellite_satellite-sea-surface-temperature_consistency_q01__de92f7f76379", "report_id": "satellite_satellite-sea-surface-temperature_consistency_q01", "dataset_id": "satellite-sea-surface-temperature", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q01", "aspect_base": "consistency", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Consistency Assessment of Satellite Sea Surface Temperature for Climate Monitoring > Methodology", "title": "Consistency Assessment of Satellite Sea Surface Temperature for Climate Monitoring", "chunk_index": 2, "token_count": 732, "text_raw": "We intercompare the performances of the European Space Agency Climate Change Initiative (ESA CCI) SST dataset v2.1 and the GHRSST multi product ensemble (GMPE) SST.\n\nRespectively, catalogue entries available from the CDS are the following:\n* Sea surface temperature daily gridded data from 1981 to 2016 derived from a multi-product satellite-based ensemble - from the Group for High Resolution Sea Surface Temperature (GHRSST) multi-product ensemble (GMPE) produced by the European Space Agency SST Climate Change Initiative (ESA CCI SST) (GMPE in the following). The GMPE SST is obtained as the ensemble median of 16 global contributing Level-4 (L4) analyses (including the ESA CCI v2.0 and v1.1) [[1]](https://doi.org/10.1016/j.rse.2018.12.015);\n* Sea Surface Temperature daily data from 1981 to present derived from satellite observations, ESA CCI SST L4 dataset v2.1 (ESA CCI SST in the following). This L4 datasets provides daily global SST fields from the merging of space born Infrared SSTs through a variational assimilation algorithm [[2]](https://doi.org/10.1038/s41597-019-0236-x);\n\nThe methodology employed is suitable for investigating the representation of SST variability over climatological time scales.\n\n-The climatology has been defined over a common, standard 30-year long period 1982-2011.\n\n-Annual and seasonal mean, as well as standard deviation, have been computed over the time axis.\n\n-Time series of globally averaged SST anomalies were derived by subtracting the annual cycle from the product’s monthly mean time series, representing the mean of each month over the entire period (1982-2016).\n\nThis processing aims to minimize the impact of the seasonal component, thereby enhancing the contribution of higher and lower frequency components.\n\nThe analysis and results are organised in the following steps, which are detailed in the sections below:'\n\n**[](section-1)**\n\n**[](section-2)**\n\nFunctions for rechunking to optimize caching of intermediate results, get nan values on land and convert to degree celsius in the ocean are defined in this section. Note that chunking is employed to handle more efficiently the large data volume.\nNotice that while in GMPE SST sea-ice covered points have nan values, this is not the case in ESA CCI SST, which has an additional variable called \"mask\". We use this ancillary information to mask out sea-ice covered regions (the variable mask is valued-1 for ocean only points).\n\n**[](section-3)**\n\nTuples and dictionaries containing infomations such as reduction operations (mean and standard deviation) are defined in this section, and the request is set up, transforming the data while downloading them, according to the proposed diagnostics.\n\n**[](section-4)**\n * 4.1 Plot Annual Maps\n * 4.2 Plot Bias maps\n * 4.3 Plot Seasonal Maps\n * 4.4 Plot of global monthly mean time series, Eastern Tropical Pacific SST (El Niño 3), and East Central Tropical Pacific SST (El Niño 3.4)", "text_with_prefix": "EQC Quality Assessment: \"Consistency Assessment of Satellite Sea Surface Temperature for Climate Monitoring\"\nDataset: satellite-sea-surface-temperature [CDS]\nAspect: consistency_q01 | Category: Satellite_ECVs\nSection: Consistency Assessment of Satellite Sea Surface Temperature for Climate Monitoring > Methodology\n---\nWe intercompare the performances of the European Space Agency Climate Change Initiative (ESA CCI) SST dataset v2.1 and the GHRSST multi product ensemble (GMPE) SST.\n\nRespectively, catalogue entries available from the CDS are the following:\n* Sea surface temperature daily gridded data from 1981 to 2016 derived from a multi-product satellite-based ensemble - from the Group for High Resolution Sea Surface Temperature (GHRSST) multi-product ensemble (GMPE) produced by the European Space Agency SST Climate Change Initiative (ESA CCI SST) (GMPE in the following). The GMPE SST is obtained as the ensemble median of 16 global contributing Level-4 (L4) analyses (including the ESA CCI v2.0 and v1.1) [[1]](https://doi.org/10.1016/j.rse.2018.12.015);\n* Sea Surface Temperature daily data from 1981 to present derived from satellite observations, ESA CCI SST L4 dataset v2.1 (ESA CCI SST in the following). This L4 datasets provides daily global SST fields from the merging of space born Infrared SSTs through a variational assimilation algorithm [[2]](https://doi.org/10.1038/s41597-019-0236-x);\n\nThe methodology employed is suitable for investigating the representation of SST variability over climatological time scales.\n\n-The climatology has been defined over a common, standard 30-year long period 1982-2011.\n\n-Annual and seasonal mean, as well as standard deviation, have been computed over the time axis.\n\n-Time series of globally averaged SST anomalies were derived by subtracting the annual cycle from the product’s monthly mean time series, representing the mean of each month over the entire period (1982-2016).\n\nThis processing aims to minimize the impact of the seasonal component, thereby enhancing the contribution of higher and lower frequency components.\n\nThe analysis and results are organised in the following steps, which are detailed in the sections below:'\n\n**[](section-1)**\n\n**[](section-2)**\n\nFunctions for rechunking to optimize caching of intermediate results, get nan values on land and convert to degree celsius in the ocean are defined in this section. Note that chunking is employed to handle more efficiently the large data volume.\nNotice that while in GMPE SST sea-ice covered points have nan values, this is not the case in ESA CCI SST, which has an additional variable called \"mask\". We use this ancillary information to mask out sea-ice covered regions (the variable mask is valued-1 for ocean only points).\n\n**[](section-3)**\n\nTuples and dictionaries containing infomations such as reduction operations (mean and standard deviation) are defined in this section, and the request is set up, transforming the data while downloading them, according to the proposed diagnostics.\n\n**[](section-4)**\n * 4.1 Plot Annual Maps\n * 4.2 Plot Bias maps\n * 4.3 Plot Seasonal Maps\n * 4.4 Plot of global monthly mean time series, Eastern Tropical Pacific SST (El Niño 3), and East Central Tropical Pacific SST (El Niño 3.4)"} {"chunk_id": "satellite_satellite-sea-surface-temperature_consistency_q01__0ca64e6daf19", "report_id": "satellite_satellite-sea-surface-temperature_consistency_q01", "dataset_id": "satellite-sea-surface-temperature", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q01", "aspect_base": "consistency", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Consistency Assessment of Satellite Sea Surface Temperature for Climate Monitoring > Analysis and results > 1. Import packages and define parameters for the requests to the CDS", "title": "Consistency Assessment of Satellite Sea Surface Temperature for Climate Monitoring", "chunk_index": 3, "token_count": 170, "text_raw": "In the following cell the necessary packages are imported, defining the request to the CDS to download ESA CCI and GMPE SST from 1982 to 2011. Regions to calculate nino3 and nino3.4, and global timeseries are defined.\n\nParameters to speed up I/O. For more information, see https://docs.xarray.dev/en/stable/generated/xarray.open_mfdataset.html\n\n(section-2)=", "text_with_prefix": "EQC Quality Assessment: \"Consistency Assessment of Satellite Sea Surface Temperature for Climate Monitoring\"\nDataset: satellite-sea-surface-temperature [CDS]\nAspect: consistency_q01 | Category: Satellite_ECVs\nSection: Consistency Assessment of Satellite Sea Surface Temperature for Climate Monitoring > Analysis and results > 1. Import packages and define parameters for the requests to the CDS\n---\nIn the following cell the necessary packages are imported, defining the request to the CDS to download ESA CCI and GMPE SST from 1982 to 2011. Regions to calculate nino3 and nino3.4, and global timeseries are defined.\n\nParameters to speed up I/O. For more information, see https://docs.xarray.dev/en/stable/generated/xarray.open_mfdataset.html\n\n(section-2)="} {"chunk_id": "satellite_satellite-sea-surface-temperature_consistency_q01__8993fa649f3f", "report_id": "satellite_satellite-sea-surface-temperature_consistency_q01", "dataset_id": "satellite-sea-surface-temperature", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q01", "aspect_base": "consistency", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Consistency Assessment of Satellite Sea Surface Temperature for Climate Monitoring > Analysis and results > 3. Download and transform step", "title": "Consistency Assessment of Satellite Sea Surface Temperature for Climate Monitoring", "chunk_index": 4, "token_count": 378, "text_raw": "Initialize variables\nNote: year from December year-1 to November year\nSeasonal and spatial\n\n```text\nproduct='ESACCI'\n```\n\n```text\nannual: 100%|██████████| 30/30 [00:13<00:00, 2.17it/s]\nspatial reduction: 100%|██████████| 3/3 [00:00<00:00, 5.23it/s]\nseason: 100%|██████████| 4/4 [00:00<00:00, 27.77it/s]\nseason: 100%|██████████| 4/4 [00:00<00:00, 12.79it/s]\n```\n\n```text\nproduct='GMPE'\n```\n\n```text\nannual: 100%|██████████| 30/30 [00:14<00:00, 2.07it/s]\nspatial reduction: 100%|██████████| 3/3 [00:00<00:00, 4.14it/s]\nseason: 100%|██████████| 4/4 [00:00<00:00, 6.46it/s]\nseason: 100%|██████████| 4/4 [00:01<00:00, 3.41it/s]\n```\n\n(section-4)=", "text_with_prefix": "EQC Quality Assessment: \"Consistency Assessment of Satellite Sea Surface Temperature for Climate Monitoring\"\nDataset: satellite-sea-surface-temperature [CDS]\nAspect: consistency_q01 | Category: Satellite_ECVs\nSection: Consistency Assessment of Satellite Sea Surface Temperature for Climate Monitoring > Analysis and results > 3. Download and transform step\n---\nInitialize variables\nNote: year from December year-1 to November year\nSeasonal and spatial\n\n```text\nproduct='ESACCI'\n```\n\n```text\nannual: 100%|██████████| 30/30 [00:13<00:00, 2.17it/s]\nspatial reduction: 100%|██████████| 3/3 [00:00<00:00, 5.23it/s]\nseason: 100%|██████████| 4/4 [00:00<00:00, 27.77it/s]\nseason: 100%|██████████| 4/4 [00:00<00:00, 12.79it/s]\n```\n\n```text\nproduct='GMPE'\n```\n\n```text\nannual: 100%|██████████| 30/30 [00:14<00:00, 2.07it/s]\nspatial reduction: 100%|██████████| 3/3 [00:00<00:00, 4.14it/s]\nseason: 100%|██████████| 4/4 [00:00<00:00, 6.46it/s]\nseason: 100%|██████████| 4/4 [00:01<00:00, 3.41it/s]\n```\n\n(section-4)="} {"chunk_id": "satellite_satellite-sea-surface-temperature_consistency_q01__408627b5b484", "report_id": "satellite_satellite-sea-surface-temperature_consistency_q01", "dataset_id": "satellite-sea-surface-temperature", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q01", "aspect_base": "consistency", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Consistency Assessment of Satellite Sea Surface Temperature for Climate Monitoring > Analysis and results > 4. Plot Annual Maps, Bias, Seasonal Maps and Timeseries, with discussions > 4.1 Plot Annual Maps", "title": "Consistency Assessment of Satellite Sea Surface Temperature for Climate Monitoring", "chunk_index": 5, "token_count": 376, "text_raw": "The global distribution of the mean SST during the complete 30 years from 1982 to 2011 is shown in the top panel. The climatology correctly reveals the dominant latitudinal spatial SST pattern: higher at the tropics, milder at middle latitudes and lower in the polar regions (see, e.g., [[3]](https://doi.org/10.1146/annurev-marine-120408-151453)). The globally averaged SST value is estimated in 20.11 $\\pm$ 0.39 °C in ESA CCI and 20.23 $\\pm$ 0.37 °C in GMPE.\nThe standard deviation map of annual mean SST is shown in the bottom panel. The annual average removes the seasonal variability and better quantifies the magnitude and spatial distribution of the nonseasonal SST variability. The standard deviation map evidences some patterns of the main current systems such as the Gulf Current and the Kuroshio Current, as well as regions such El-Nino in the tropical pacific, where the SST anomaly standard deviation exceeds 1.5°C. The main upwelling systems as, e.g., Perù-Chili, Benguela, NW-African coast and along the southern Saudi Arabia coast are also evident, as well as regions of strong mixing, like the Antarctic Circumpolar Current and the Agulhas Leakage.", "text_with_prefix": "EQC Quality Assessment: \"Consistency Assessment of Satellite Sea Surface Temperature for Climate Monitoring\"\nDataset: satellite-sea-surface-temperature [CDS]\nAspect: consistency_q01 | Category: Satellite_ECVs\nSection: Consistency Assessment of Satellite Sea Surface Temperature for Climate Monitoring > Analysis and results > 4. Plot Annual Maps, Bias, Seasonal Maps and Timeseries, with discussions > 4.1 Plot Annual Maps\n---\nThe global distribution of the mean SST during the complete 30 years from 1982 to 2011 is shown in the top panel. The climatology correctly reveals the dominant latitudinal spatial SST pattern: higher at the tropics, milder at middle latitudes and lower in the polar regions (see, e.g., [[3]](https://doi.org/10.1146/annurev-marine-120408-151453)). The globally averaged SST value is estimated in 20.11 $\\pm$ 0.39 °C in ESA CCI and 20.23 $\\pm$ 0.37 °C in GMPE.\nThe standard deviation map of annual mean SST is shown in the bottom panel. The annual average removes the seasonal variability and better quantifies the magnitude and spatial distribution of the nonseasonal SST variability. The standard deviation map evidences some patterns of the main current systems such as the Gulf Current and the Kuroshio Current, as well as regions such El-Nino in the tropical pacific, where the SST anomaly standard deviation exceeds 1.5°C. The main upwelling systems as, e.g., Perù-Chili, Benguela, NW-African coast and along the southern Saudi Arabia coast are also evident, as well as regions of strong mixing, like the Antarctic Circumpolar Current and the Agulhas Leakage."} {"chunk_id": "satellite_satellite-sea-surface-temperature_consistency_q01__6aa81df561d5", "report_id": "satellite_satellite-sea-surface-temperature_consistency_q01", "dataset_id": "satellite-sea-surface-temperature", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q01", "aspect_base": "consistency", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Consistency Assessment of Satellite Sea Surface Temperature for Climate Monitoring > Analysis and results > 4. Plot Annual Maps, Bias, Seasonal Maps and Timeseries, with discussions > 4.2 Plot Bias Maps", "title": "Consistency Assessment of Satellite Sea Surface Temperature for Climate Monitoring", "chunk_index": 6, "token_count": 490, "text_raw": "Top panel depicts the difference between climatologies in GMPE and ESA CCI SSTs. In general, positive (negative) values stand for GMPE (ESA CCI) being warmer than ESA CCI (GMPE) SSTs. The maximum mean difference is about +1.7°C, especially over the Tropical Eastern Atlantic, Northern Indian Ocean, Western South American and South African coasts. The Gulf Stream presents positive and negative mean SST differences, while the Kuroshio current is mainly characterized by higher GMPE SSTs. On average, GMPE is about 0.02°C warmer than ESA CCI SST. Regarding the standard deviation (bottom panel), SST annual variability is slightly higher in ESA CCI than in GMPE over the tropical Pacific, the North Atlantic, the southern Saudi Arabia coast, and in the Agulhas system. The converse holds true in the Gulf Stream, Antarctic Circumpolar Current, and Indonesian Throughflow, resulting in GMPE having a variance close to 0.2°C greater than the one in ESA CCI. Overall, these results are consistent with the literature (see e.g. [[4]](https://doi.org/10.1175/JCLI-D-20-0793.1), [[5]](https://doi.org/10.1080/1755876X.2016.1273446)).\n\nIt is important to notice that some of the major discrepancies in the mean difference map (e.g. around the African continent and in the Northern Indian Ocean), are likely due to known issues of the ESA CCI dataset, as reported in [[2]](https://doi.org/10.1038/s41597-019-0236-x), being attributed to unscreened dust events in the Advanced Very High Resolution Radiometer (AVHRR) SSTs. This bias is strongly mitigated by GMPE SST, which is the median of an ensemble of satellite SST products.", "text_with_prefix": "EQC Quality Assessment: \"Consistency Assessment of Satellite Sea Surface Temperature for Climate Monitoring\"\nDataset: satellite-sea-surface-temperature [CDS]\nAspect: consistency_q01 | Category: Satellite_ECVs\nSection: Consistency Assessment of Satellite Sea Surface Temperature for Climate Monitoring > Analysis and results > 4. Plot Annual Maps, Bias, Seasonal Maps and Timeseries, with discussions > 4.2 Plot Bias Maps\n---\nTop panel depicts the difference between climatologies in GMPE and ESA CCI SSTs. In general, positive (negative) values stand for GMPE (ESA CCI) being warmer than ESA CCI (GMPE) SSTs. The maximum mean difference is about +1.7°C, especially over the Tropical Eastern Atlantic, Northern Indian Ocean, Western South American and South African coasts. The Gulf Stream presents positive and negative mean SST differences, while the Kuroshio current is mainly characterized by higher GMPE SSTs. On average, GMPE is about 0.02°C warmer than ESA CCI SST. Regarding the standard deviation (bottom panel), SST annual variability is slightly higher in ESA CCI than in GMPE over the tropical Pacific, the North Atlantic, the southern Saudi Arabia coast, and in the Agulhas system. The converse holds true in the Gulf Stream, Antarctic Circumpolar Current, and Indonesian Throughflow, resulting in GMPE having a variance close to 0.2°C greater than the one in ESA CCI. Overall, these results are consistent with the literature (see e.g. [[4]](https://doi.org/10.1175/JCLI-D-20-0793.1), [[5]](https://doi.org/10.1080/1755876X.2016.1273446)).\n\nIt is important to notice that some of the major discrepancies in the mean difference map (e.g. around the African continent and in the Northern Indian Ocean), are likely due to known issues of the ESA CCI dataset, as reported in [[2]](https://doi.org/10.1038/s41597-019-0236-x), being attributed to unscreened dust events in the Advanced Very High Resolution Radiometer (AVHRR) SSTs. This bias is strongly mitigated by GMPE SST, which is the median of an ensemble of satellite SST products."} {"chunk_id": "satellite_satellite-sea-surface-temperature_consistency_q01__c62d89dbe513", "report_id": "satellite_satellite-sea-surface-temperature_consistency_q01", "dataset_id": "satellite-sea-surface-temperature", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q01", "aspect_base": "consistency", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Consistency Assessment of Satellite Sea Surface Temperature for Climate Monitoring > Analysis and results > 4. Plot Annual Maps, Bias, Seasonal Maps and Timeseries, with discussions > 4.3 Plot seasonal maps", "title": "Consistency Assessment of Satellite Sea Surface Temperature for Climate Monitoring", "chunk_index": 7, "token_count": 441, "text_raw": "Seasonal SST climatologies show similar patterns among the two datasets, with a warmer (cooler) Northern Hemisphere during boreal summer (winter), as expected; the converse holds true for the Southern Hemisphere. It can be noticed the strong influence of major western boundary currents, such as the Gulf Stream and the Kuroshio in determining seasonal oscillations. Overall, from this pictures the two dataset are found in good agreement.\n\nRight after, seasonal standard deviation maps for each product are shown. For example, the DJF map is expressed by the following formula\n $\\sqrt{\\frac{1}{N}\\sum_{i=1}^{N}\\left(SST^i_{DJF} - \\langle SST_{DJF}\\rangle\\right)^2},$ where $N$ is the number of observations (i.e. 30 winter seasons in the present case) and $\\langle SST_{DJF} \\rangle$ is the average wintertime map. \nSeasonal standard deviation maps are very similar between the two products, revealing that on average the Northern Hemisphere have a much higher variability than the Southern one, with the Gulf Stream and Kuroshio dominating the patterns year round. Also, signs of high variability are noticeable in both dataset in El-Niño Region and around the Agulhas. While the first region is a reflection of a dominant climate mode, the second one arises as a result of strong mixing and turbulence.\n\nWe also plot difference maps to highlight discrepancies in the two datasets. We can clearly notice that major differences are located within the tropical belt, mainly near Africa and in the Northern Indian Ocean. As stated above, these discrepancies can be partly attributed to unresolved dust events in ESA CCI SST.", "text_with_prefix": "EQC Quality Assessment: \"Consistency Assessment of Satellite Sea Surface Temperature for Climate Monitoring\"\nDataset: satellite-sea-surface-temperature [CDS]\nAspect: consistency_q01 | Category: Satellite_ECVs\nSection: Consistency Assessment of Satellite Sea Surface Temperature for Climate Monitoring > Analysis and results > 4. Plot Annual Maps, Bias, Seasonal Maps and Timeseries, with discussions > 4.3 Plot seasonal maps\n---\nSeasonal SST climatologies show similar patterns among the two datasets, with a warmer (cooler) Northern Hemisphere during boreal summer (winter), as expected; the converse holds true for the Southern Hemisphere. It can be noticed the strong influence of major western boundary currents, such as the Gulf Stream and the Kuroshio in determining seasonal oscillations. Overall, from this pictures the two dataset are found in good agreement.\n\nRight after, seasonal standard deviation maps for each product are shown. For example, the DJF map is expressed by the following formula\n $\\sqrt{\\frac{1}{N}\\sum_{i=1}^{N}\\left(SST^i_{DJF} - \\langle SST_{DJF}\\rangle\\right)^2},$ where $N$ is the number of observations (i.e. 30 winter seasons in the present case) and $\\langle SST_{DJF} \\rangle$ is the average wintertime map. \nSeasonal standard deviation maps are very similar between the two products, revealing that on average the Northern Hemisphere have a much higher variability than the Southern one, with the Gulf Stream and Kuroshio dominating the patterns year round. Also, signs of high variability are noticeable in both dataset in El-Niño Region and around the Agulhas. While the first region is a reflection of a dominant climate mode, the second one arises as a result of strong mixing and turbulence.\n\nWe also plot difference maps to highlight discrepancies in the two datasets. We can clearly notice that major differences are located within the tropical belt, mainly near Africa and in the Northern Indian Ocean. As stated above, these discrepancies can be partly attributed to unresolved dust events in ESA CCI SST."} {"chunk_id": "satellite_satellite-sea-surface-temperature_consistency_q01__d314727688ff", "report_id": "satellite_satellite-sea-surface-temperature_consistency_q01", "dataset_id": "satellite-sea-surface-temperature", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q01", "aspect_base": "consistency", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Consistency Assessment of Satellite Sea Surface Temperature for Climate Monitoring > Analysis and results > 4. Plot Annual Maps, Bias, Seasonal Maps and Timeseries, with discussions > 4.4 Plot of global monthly mean time series, Eastern Tropical Pacific SST (El Niño 3), and East Central Tropical Pacific SST (El Niño 3.4)", "title": "Consistency Assessment of Satellite Sea Surface Temperature for Climate Monitoring", "chunk_index": 8, "token_count": 784, "text_raw": "The time series illustrating the globally averaged monthly mean SST anomalies from 1982 to 2016 is presented. These anomalies were derived by subtracting the annual cycle from the product’s monthly mean time series, representing the mean of each month over the entire period (1982-2016). This processing aims to minimize the impact of the seasonal component, thereby enhancing the contribution of higher and lower frequency components. Notably, variations at an interannual time scale are prominently displayed. These changes are mostly related to strong signatures of El Niño Southern Oscillation (ENSO) variability, with a particular strong and sharp increase (warming) in SST during e.g. the 1982/3, 1987/88, 1991/92, 1994/95, 1997/8 and 2009/10 El Niño events, and negative (cooling) peaks during the 1983/5, 1988/9, 1995/6 and 2006/7 La Niña events. Differences between the two time series which exceed 0.5 degrees characterize the first decade of the period, before 1990. The large and sharp peak in 1982/3 reveals apparently erroneous SSTs, as also evidenced in the SST CCI Climate Assessment Report (CAR, available at https://climate.esa.int/media/documents/SST_CCI_D5.1_CAR_v1.1-signed.pdf). \nThe time series of SST anomalies within the El Niño 3.4 and El Niño 3 regions are also shown (see also [[6]](https://climatedataguide.ucar.edu/climate-data/nino-sst-indices-nino-12-3-34-4-oni-and-tni) for a definition of El Niño Southern Oscillation indices). These time series are used as indexes to monitor the occurrence and variability of El Niño and la Niña events. The El Niño 3.4 index is defined as the average equatorial SST anomalies across the Pacific in the region 5°S-5°N, 170W°-120W°, while El Niño 3 between 150°W-90°W. Peaks corresponding to El-Niño/La-Niña events are well represented by both GMPE and ESA CCI SST products. See [[6]](https://climatedataguide.ucar.edu/climate-data/nino-sst-indices-nino-12-3-34-4-oni-and-tni) for reference.\nFor both the global mean timeseries and El Niño indexes, there are discrepancies in ESA CCI compared to GMPE SST (GMPE is here used as intercomparison benchmark due to its high accuracy, following [[1]](https://doi.org/10.1016/j.rse.2018.12.015), [[7]](https://doi.org/10.1016/j.dsr2.2012.04.013)), in the period 1982 to 1992. Such discrepancies are highly attenuated from 1992 onwards, suggesting that ESA CCI data should be used with caution in the first decade of the CDR.", "text_with_prefix": "EQC Quality Assessment: \"Consistency Assessment of Satellite Sea Surface Temperature for Climate Monitoring\"\nDataset: satellite-sea-surface-temperature [CDS]\nAspect: consistency_q01 | Category: Satellite_ECVs\nSection: Consistency Assessment of Satellite Sea Surface Temperature for Climate Monitoring > Analysis and results > 4. Plot Annual Maps, Bias, Seasonal Maps and Timeseries, with discussions > 4.4 Plot of global monthly mean time series, Eastern Tropical Pacific SST (El Niño 3), and East Central Tropical Pacific SST (El Niño 3.4)\n---\nThe time series illustrating the globally averaged monthly mean SST anomalies from 1982 to 2016 is presented. These anomalies were derived by subtracting the annual cycle from the product’s monthly mean time series, representing the mean of each month over the entire period (1982-2016). This processing aims to minimize the impact of the seasonal component, thereby enhancing the contribution of higher and lower frequency components. Notably, variations at an interannual time scale are prominently displayed. These changes are mostly related to strong signatures of El Niño Southern Oscillation (ENSO) variability, with a particular strong and sharp increase (warming) in SST during e.g. the 1982/3, 1987/88, 1991/92, 1994/95, 1997/8 and 2009/10 El Niño events, and negative (cooling) peaks during the 1983/5, 1988/9, 1995/6 and 2006/7 La Niña events. Differences between the two time series which exceed 0.5 degrees characterize the first decade of the period, before 1990. The large and sharp peak in 1982/3 reveals apparently erroneous SSTs, as also evidenced in the SST CCI Climate Assessment Report (CAR, available at https://climate.esa.int/media/documents/SST_CCI_D5.1_CAR_v1.1-signed.pdf). \nThe time series of SST anomalies within the El Niño 3.4 and El Niño 3 regions are also shown (see also [[6]](https://climatedataguide.ucar.edu/climate-data/nino-sst-indices-nino-12-3-34-4-oni-and-tni) for a definition of El Niño Southern Oscillation indices). These time series are used as indexes to monitor the occurrence and variability of El Niño and la Niña events. The El Niño 3.4 index is defined as the average equatorial SST anomalies across the Pacific in the region 5°S-5°N, 170W°-120W°, while El Niño 3 between 150°W-90°W. Peaks corresponding to El-Niño/La-Niña events are well represented by both GMPE and ESA CCI SST products. See [[6]](https://climatedataguide.ucar.edu/climate-data/nino-sst-indices-nino-12-3-34-4-oni-and-tni) for reference.\nFor both the global mean timeseries and El Niño indexes, there are discrepancies in ESA CCI compared to GMPE SST (GMPE is here used as intercomparison benchmark due to its high accuracy, following [[1]](https://doi.org/10.1016/j.rse.2018.12.015), [[7]](https://doi.org/10.1016/j.dsr2.2012.04.013)), in the period 1982 to 1992. Such discrepancies are highly attenuated from 1992 onwards, suggesting that ESA CCI data should be used with caution in the first decade of the CDR."} {"chunk_id": "satellite_satellite-sea-surface-temperature_consistency_q01__1d0b6efdecac", "report_id": "satellite_satellite-sea-surface-temperature_consistency_q01", "dataset_id": "satellite-sea-surface-temperature", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q01", "aspect_base": "consistency", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Consistency Assessment of Satellite Sea Surface Temperature for Climate Monitoring > ℹ️ If you want to know more > Key resources", "title": "Consistency Assessment of Satellite Sea Surface Temperature for Climate Monitoring", "chunk_index": 9, "token_count": 235, "text_raw": "SST CCI Climate Assessment Report (CAR): https://climate.esa.int/media/documents/SST_CCI_D5.1_CAR_v1.1-signed.pdf\n\nGHRSST Website: https://www.ghrsst.org/\n\nAdditional sources for SST ensemble intercomparison statistics: https://ghrsst-pp.metoffice.gov.uk/ostia-website/gmpe-monitoring.html\n\nCode libraries used:\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)\n* python packages: matplotlib, xarray, cartopy, numpy, tqdm", "text_with_prefix": "EQC Quality Assessment: \"Consistency Assessment of Satellite Sea Surface Temperature for Climate Monitoring\"\nDataset: satellite-sea-surface-temperature [CDS]\nAspect: consistency_q01 | Category: Satellite_ECVs\nSection: Consistency Assessment of Satellite Sea Surface Temperature for Climate Monitoring > ℹ️ If you want to know more > Key resources\n---\nSST CCI Climate Assessment Report (CAR): https://climate.esa.int/media/documents/SST_CCI_D5.1_CAR_v1.1-signed.pdf\n\nGHRSST Website: https://www.ghrsst.org/\n\nAdditional sources for SST ensemble intercomparison statistics: https://ghrsst-pp.metoffice.gov.uk/ostia-website/gmpe-monitoring.html\n\nCode libraries used:\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)\n* python packages: matplotlib, xarray, cartopy, numpy, tqdm"} {"chunk_id": "satellite_satellite-sea-surface-temperature_consistency_q01__dccfee0eac83", "report_id": "satellite_satellite-sea-surface-temperature_consistency_q01", "dataset_id": "satellite-sea-surface-temperature", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q01", "aspect_base": "consistency", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Consistency Assessment of Satellite Sea Surface Temperature for Climate Monitoring > ℹ️ If you want to know more > References", "title": "Consistency Assessment of Satellite Sea Surface Temperature for Climate Monitoring", "chunk_index": 10, "token_count": 831, "text_raw": "[[1]](https://doi.org/10.1016/j.rse.2018.12.015) Fiedler, E. K., McLaren, A., Banzon, V., Brasnett, B., Ishizaki, S., Kennedy, J., ... & Donlon, C. (2019). Intercomparison of long-term sea surface temperature analyses using the GHRSST Multi-Product Ensemble (GMPE) system. Remote sensing of environment, 222, 18-33\n\n[[2]](https://doi.org/10.1038/s41597-019-0236-x) Merchant, C. J., Embury, O., Bulgin, C. E., Block, T., Corlett, G. K., Fiedler, E., ... & Donlon, C. (2019). Satellite-based time-series of sea-surface temperature since 1981 for climate applications. Scientific data, 6(1), 223. https://doi.org/10.1038/s41597-019-0236-x\n\n[[3]](https://doi.org/10.1146/annurev-marine-120408-151453) Deser, C., Alexander, M. A., Xie, S. P., & Phillips, A. S. (2010). Sea surface temperature variability: Patterns and mechanisms. Annual review of marine science, 2, 115-143.\n\n[[4]](https://doi.org/10.1175/JCLI-D-20-0793.1) Yang, C., Leonelli, F. E., Marullo, S., Artale, V., Beggs, H., Nardelli, B. B., ... & Pisano, A. (2021). Sea surface temperature intercomparison in the framework of the Copernicus Climate Change Service (C3S). Journal of Climate, 34(13), 5257-5283\n\n[[5]](https://doi.org/10.1080/1755876X.2016.1273446) Von Schuckmann et al. (2016). The Copernicus Marine Environment Monitoring Service Ocean State Report. Jour. Operational Ocean., vol. 9, 2016, suppl. 2. doi: https://doi.org/10.1080/1755876X.2016.1273446\n\n[[6]](https://climatedataguide.ucar.edu/climate-data/nino-sst-indices-nino-12-3-34-4-oni-and-tni) Trenberth, Kevin & National Center for Atmospheric Research Staff (Eds). Last modified 21 Jan 2020. \"The Climate Data Guide: Nino SST Indices (Nino 1+2, 3, 3.4, 4; ONI and TNI).\" Retrieved from https://climatedataguide.ucar.edu/climate-data/nino-sst-indices-nino-12-3-34-4-oni-and-tni\n\n[[7]](https://doi.org/10.1016/j.dsr2.2012.04.013) Martin, M., Dash, P., Ignatov, A., Banzon, V., Beggs, H., Brasnett, B., ... & Roberts-Jones, J. (2012). Group for High Resolution Sea Surface temperature (GHRSST) analysis fields inter-comparisons. Part 1: A GHRSST multi-product ensemble (GMPE). Deep Sea Research Part II: Topical Studies in Oceanography, 77, 21-30", "text_with_prefix": "EQC Quality Assessment: \"Consistency Assessment of Satellite Sea Surface Temperature for Climate Monitoring\"\nDataset: satellite-sea-surface-temperature [CDS]\nAspect: consistency_q01 | Category: Satellite_ECVs\nSection: Consistency Assessment of Satellite Sea Surface Temperature for Climate Monitoring > ℹ️ If you want to know more > References\n---\n[[1]](https://doi.org/10.1016/j.rse.2018.12.015) Fiedler, E. K., McLaren, A., Banzon, V., Brasnett, B., Ishizaki, S., Kennedy, J., ... & Donlon, C. (2019). Intercomparison of long-term sea surface temperature analyses using the GHRSST Multi-Product Ensemble (GMPE) system. Remote sensing of environment, 222, 18-33\n\n[[2]](https://doi.org/10.1038/s41597-019-0236-x) Merchant, C. J., Embury, O., Bulgin, C. E., Block, T., Corlett, G. K., Fiedler, E., ... & Donlon, C. (2019). Satellite-based time-series of sea-surface temperature since 1981 for climate applications. Scientific data, 6(1), 223. https://doi.org/10.1038/s41597-019-0236-x\n\n[[3]](https://doi.org/10.1146/annurev-marine-120408-151453) Deser, C., Alexander, M. A., Xie, S. P., & Phillips, A. S. (2010). Sea surface temperature variability: Patterns and mechanisms. Annual review of marine science, 2, 115-143.\n\n[[4]](https://doi.org/10.1175/JCLI-D-20-0793.1) Yang, C., Leonelli, F. E., Marullo, S., Artale, V., Beggs, H., Nardelli, B. B., ... & Pisano, A. (2021). Sea surface temperature intercomparison in the framework of the Copernicus Climate Change Service (C3S). Journal of Climate, 34(13), 5257-5283\n\n[[5]](https://doi.org/10.1080/1755876X.2016.1273446) Von Schuckmann et al. (2016). The Copernicus Marine Environment Monitoring Service Ocean State Report. Jour. Operational Ocean., vol. 9, 2016, suppl. 2. doi: https://doi.org/10.1080/1755876X.2016.1273446\n\n[[6]](https://climatedataguide.ucar.edu/climate-data/nino-sst-indices-nino-12-3-34-4-oni-and-tni) Trenberth, Kevin & National Center for Atmospheric Research Staff (Eds). Last modified 21 Jan 2020. \"The Climate Data Guide: Nino SST Indices (Nino 1+2, 3, 3.4, 4; ONI and TNI).\" Retrieved from https://climatedataguide.ucar.edu/climate-data/nino-sst-indices-nino-12-3-34-4-oni-and-tni\n\n[[7]](https://doi.org/10.1016/j.dsr2.2012.04.013) Martin, M., Dash, P., Ignatov, A., Banzon, V., Beggs, H., Brasnett, B., ... & Roberts-Jones, J. (2012). Group for High Resolution Sea Surface temperature (GHRSST) analysis fields inter-comparisons. Part 1: A GHRSST multi-product ensemble (GMPE). Deep Sea Research Part II: Topical Studies in Oceanography, 77, 21-30"} {"chunk_id": "satellite_satellite-sea-surface-temperature_consistency_q02__a05046ac4179", "report_id": "satellite_satellite-sea-surface-temperature_consistency_q02", "dataset_id": "satellite-sea-surface-temperature", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q02", "aspect_base": "consistency", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Sea Surface Temperature Consistency in representing Ocean Warming Trends > Quality assessment question", "title": "Satellite Sea Surface Temperature Consistency in representing Ocean Warming Trends", "chunk_index": 0, "token_count": 169, "text_raw": "* **How well do SST trend estimates obtained from ESA CCI L4 v2.1 and GHRSST Multi-Product Ensemble (GMPE) compare with each other?**\n\nOur aim here is to evaluate the reliability and representativeness of sea surface temperature (SST) linear trends in satellite-based long-term climate data records (CDRs). To this end, two different SST CDRs have been intercompared over a common 30-year long period (1982-2011).", "text_with_prefix": "EQC Quality Assessment: \"Satellite Sea Surface Temperature Consistency in representing Ocean Warming Trends\"\nDataset: satellite-sea-surface-temperature [CDS]\nAspect: consistency_q02 | Category: Satellite_ECVs\nSection: Satellite Sea Surface Temperature Consistency in representing Ocean Warming Trends > Quality assessment question\n---\n* **How well do SST trend estimates obtained from ESA CCI L4 v2.1 and GHRSST Multi-Product Ensemble (GMPE) compare with each other?**\n\nOur aim here is to evaluate the reliability and representativeness of sea surface temperature (SST) linear trends in satellite-based long-term climate data records (CDRs). To this end, two different SST CDRs have been intercompared over a common 30-year long period (1982-2011)."} {"chunk_id": "satellite_satellite-sea-surface-temperature_consistency_q02__f32a994d07be", "report_id": "satellite_satellite-sea-surface-temperature_consistency_q02", "dataset_id": "satellite-sea-surface-temperature", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q02", "aspect_base": "consistency", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Sea Surface Temperature Consistency in representing Ocean Warming Trends > Quality assessment statement", "title": "Satellite Sea Surface Temperature Consistency in representing Ocean Warming Trends", "chunk_index": 1, "token_count": 308, "text_raw": "These are the key outcomes of this assessment\n\n* Both datasets correctly capture the long-term SST trend at both global and regional scale; \t\t\t\t\t\t\t\n* The spatial pattern of the SST trend evidences that the Northern Hemisphere is experiencing a more intense warming compared to the Southern Hemisphere, as reported in literature ([[1]](https://doi.org/10.1007/s00382-014-2147-z), [[2]](https://doi.org/10.1175/JCLI-D-20-0793.1));\n* ESA CCI L4 v2.1 tends to exhibit higher trend estimates compared to GMPE, which serves as a reference dataset due to its nature, particularly in certain ocean basins like the Tropical Atlantic and Northern Indian Ocean;\n* On average, the global mean SST trend estimates provided by the two datasets are consistent within the 95% confidence interval;\n```\n\nattachment:c3c17d69-0c57-4673-a04b-f10787dc8333.png\n---\nheight: 400px\nwidth: 800px\n---\n*Linear trends (1982-2011) of ESA CCI and GMPE SSTs*\n```", "text_with_prefix": "EQC Quality Assessment: \"Satellite Sea Surface Temperature Consistency in representing Ocean Warming Trends\"\nDataset: satellite-sea-surface-temperature [CDS]\nAspect: consistency_q02 | Category: Satellite_ECVs\nSection: Satellite Sea Surface Temperature Consistency in representing Ocean Warming Trends > Quality assessment statement\n---\nThese are the key outcomes of this assessment\n\n* Both datasets correctly capture the long-term SST trend at both global and regional scale; \t\t\t\t\t\t\t\n* The spatial pattern of the SST trend evidences that the Northern Hemisphere is experiencing a more intense warming compared to the Southern Hemisphere, as reported in literature ([[1]](https://doi.org/10.1007/s00382-014-2147-z), [[2]](https://doi.org/10.1175/JCLI-D-20-0793.1));\n* ESA CCI L4 v2.1 tends to exhibit higher trend estimates compared to GMPE, which serves as a reference dataset due to its nature, particularly in certain ocean basins like the Tropical Atlantic and Northern Indian Ocean;\n* On average, the global mean SST trend estimates provided by the two datasets are consistent within the 95% confidence interval;\n```\n\nattachment:c3c17d69-0c57-4673-a04b-f10787dc8333.png\n---\nheight: 400px\nwidth: 800px\n---\n*Linear trends (1982-2011) of ESA CCI and GMPE SSTs*\n```"} {"chunk_id": "satellite_satellite-sea-surface-temperature_consistency_q02__80561358c2e5", "report_id": "satellite_satellite-sea-surface-temperature_consistency_q02", "dataset_id": "satellite-sea-surface-temperature", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q02", "aspect_base": "consistency", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Sea Surface Temperature Consistency in representing Ocean Warming Trends > Methodology", "title": "Satellite Sea Surface Temperature Consistency in representing Ocean Warming Trends", "chunk_index": 2, "token_count": 608, "text_raw": "We evaluate the reliability and representativeness of Sea Surface Temperature (SST) linear trends in satellite-based long-term climate data records (CDRs). To this end, two different SST CDRs have been intercompared over a common 30-year long period (1982-2011). \nRespectively, catalogue entries available from the CDS are the following:\n\n* Sea surface temperature daily gridded data from 1981 to 2016 derived from a multi-product satellite-based ensemble - from the Group for High Resolution Sea Surface Temperature (GHRSST) multi-product ensemble (GMPE) produced by the European Space Agency Climate Change Initiative SST (ESA CCI SST) (GMPE in the following). The GMPE SST is obtained as the ensemble median of 16 global contributing Level-4 (L4) analyses (including the ESA CCI v2.0 and v1.1) [[3]](https://doi.org/10.1016/j.rse.2018.12.015);\n* Sea Surface Temperature daily data from 1981 to present derived from satellite observations, ESA CCI SST L4 dataset v2.1 (ESA CCI SST in the following). This L4 datasets provides daily global SST fields from the merging of space born Infrared SSTs through a variational assimilation algorithm [[4]](https://doi.org/10.1038/s41597-019-0236-x);\n\nThe methodology employed is suitable to monitor surface ocean warming over a climatological standard period.\n\n-Annual and seasonal reductions (linear trend with Mann-Kendall test to estimate its significance at 95% confidence level) have been computed over the time axis, and shown as maps. In these maps, patches denote grid points in which there’s no significant trend;\n\n-To have a closer look at areas of major trend discrepancies between the two datasets, time series of spatially averaged annual mean SST, with their trends, have been computed over Tropical Atlantic and Northern Indian Ocean;\n\n-Linear trend estimates were retrieved with the same method used for producing trend maps, described above;\n\n-Superimposed to the time series, also the absolute differences plot on a twin axis illustrate how discrepancies are not homogenous in time.\n\nThe analysis and results are organised in the following steps, which are detailed in the sections below:\n\n**[](section-1)**\n\n**[](section-2)**\n\n**[](section-3)**\n\n**[](section-4)**\n * 4.1 Plot Annual trend maps\n * 4.2 Seasonal Trends\n * 4.3 SST Time series and trends over Tropical Atlantic and Northern Indian Ocean.", "text_with_prefix": "EQC Quality Assessment: \"Satellite Sea Surface Temperature Consistency in representing Ocean Warming Trends\"\nDataset: satellite-sea-surface-temperature [CDS]\nAspect: consistency_q02 | Category: Satellite_ECVs\nSection: Satellite Sea Surface Temperature Consistency in representing Ocean Warming Trends > Methodology\n---\nWe evaluate the reliability and representativeness of Sea Surface Temperature (SST) linear trends in satellite-based long-term climate data records (CDRs). To this end, two different SST CDRs have been intercompared over a common 30-year long period (1982-2011). \nRespectively, catalogue entries available from the CDS are the following:\n\n* Sea surface temperature daily gridded data from 1981 to 2016 derived from a multi-product satellite-based ensemble - from the Group for High Resolution Sea Surface Temperature (GHRSST) multi-product ensemble (GMPE) produced by the European Space Agency Climate Change Initiative SST (ESA CCI SST) (GMPE in the following). The GMPE SST is obtained as the ensemble median of 16 global contributing Level-4 (L4) analyses (including the ESA CCI v2.0 and v1.1) [[3]](https://doi.org/10.1016/j.rse.2018.12.015);\n* Sea Surface Temperature daily data from 1981 to present derived from satellite observations, ESA CCI SST L4 dataset v2.1 (ESA CCI SST in the following). This L4 datasets provides daily global SST fields from the merging of space born Infrared SSTs through a variational assimilation algorithm [[4]](https://doi.org/10.1038/s41597-019-0236-x);\n\nThe methodology employed is suitable to monitor surface ocean warming over a climatological standard period.\n\n-Annual and seasonal reductions (linear trend with Mann-Kendall test to estimate its significance at 95% confidence level) have been computed over the time axis, and shown as maps. In these maps, patches denote grid points in which there’s no significant trend;\n\n-To have a closer look at areas of major trend discrepancies between the two datasets, time series of spatially averaged annual mean SST, with their trends, have been computed over Tropical Atlantic and Northern Indian Ocean;\n\n-Linear trend estimates were retrieved with the same method used for producing trend maps, described above;\n\n-Superimposed to the time series, also the absolute differences plot on a twin axis illustrate how discrepancies are not homogenous in time.\n\nThe analysis and results are organised in the following steps, which are detailed in the sections below:\n\n**[](section-1)**\n\n**[](section-2)**\n\n**[](section-3)**\n\n**[](section-4)**\n * 4.1 Plot Annual trend maps\n * 4.2 Seasonal Trends\n * 4.3 SST Time series and trends over Tropical Atlantic and Northern Indian Ocean."} {"chunk_id": "satellite_satellite-sea-surface-temperature_consistency_q02__f4917b992c86", "report_id": "satellite_satellite-sea-surface-temperature_consistency_q02", "dataset_id": "satellite-sea-surface-temperature", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q02", "aspect_base": "consistency", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Sea Surface Temperature Consistency in representing Ocean Warming Trends > Analysis and results > 1. Import packages and define parameters for the requests to the CDS", "title": "Satellite Sea Surface Temperature Consistency in representing Ocean Warming Trends", "chunk_index": 3, "token_count": 118, "text_raw": "In the following cell the necessary packages are imported, defining parameters for the request and trend calculation.\n\nChoose timeseries\nParameters to speed up I/O\n\n(section-2)=", "text_with_prefix": "EQC Quality Assessment: \"Satellite Sea Surface Temperature Consistency in representing Ocean Warming Trends\"\nDataset: satellite-sea-surface-temperature [CDS]\nAspect: consistency_q02 | Category: Satellite_ECVs\nSection: Satellite Sea Surface Temperature Consistency in representing Ocean Warming Trends > Analysis and results > 1. Import packages and define parameters for the requests to the CDS\n---\nIn the following cell the necessary packages are imported, defining parameters for the request and trend calculation.\n\nChoose timeseries\nParameters to speed up I/O\n\n(section-2)="} {"chunk_id": "satellite_satellite-sea-surface-temperature_consistency_q02__b42136f49b67", "report_id": "satellite_satellite-sea-surface-temperature_consistency_q02", "dataset_id": "satellite-sea-surface-temperature", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q02", "aspect_base": "consistency", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Sea Surface Temperature Consistency in representing Ocean Warming Trends > Analysis and results > 2. Define functions to cache", "title": "Satellite Sea Surface Temperature Consistency in representing Ocean Warming Trends", "chunk_index": 4, "token_count": 113, "text_raw": "Functions for rechunking to optimize caching of intermediate results, get nan values on land, convert to degree Celsius in the ocean and calculate the trend are defined below.\n\n(section-3)=", "text_with_prefix": "EQC Quality Assessment: \"Satellite Sea Surface Temperature Consistency in representing Ocean Warming Trends\"\nDataset: satellite-sea-surface-temperature [CDS]\nAspect: consistency_q02 | Category: Satellite_ECVs\nSection: Satellite Sea Surface Temperature Consistency in representing Ocean Warming Trends > Analysis and results > 2. Define functions to cache\n---\nFunctions for rechunking to optimize caching of intermediate results, get nan values on land, convert to degree Celsius in the ocean and calculate the trend are defined below.\n\n(section-3)="} {"chunk_id": "satellite_satellite-sea-surface-temperature_consistency_q02__e441c12c5bb3", "report_id": "satellite_satellite-sea-surface-temperature_consistency_q02", "dataset_id": "satellite-sea-surface-temperature", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q02", "aspect_base": "consistency", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Sea Surface Temperature Consistency in representing Ocean Warming Trends > Analysis and results > 3. Download and transform", "title": "Satellite Sea Surface Temperature Consistency in representing Ocean Warming Trends", "chunk_index": 5, "token_count": 105, "text_raw": "In this step the data are requested, retrieved and processed in one step.\n\nprint(f\"{product=}\")\nMann Kendall settings\n\n(section-4)=", "text_with_prefix": "EQC Quality Assessment: \"Satellite Sea Surface Temperature Consistency in representing Ocean Warming Trends\"\nDataset: satellite-sea-surface-temperature [CDS]\nAspect: consistency_q02 | Category: Satellite_ECVs\nSection: Satellite Sea Surface Temperature Consistency in representing Ocean Warming Trends > Analysis and results > 3. Download and transform\n---\nIn this step the data are requested, retrieved and processed in one step.\n\nprint(f\"{product=}\")\nMann Kendall settings\n\n(section-4)="} {"chunk_id": "satellite_satellite-sea-surface-temperature_consistency_q02__991f86aea007", "report_id": "satellite_satellite-sea-surface-temperature_consistency_q02", "dataset_id": "satellite-sea-surface-temperature", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q02", "aspect_base": "consistency", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Sea Surface Temperature Consistency in representing Ocean Warming Trends > Analysis and results > 4. Plot Annual and Seasonal trends, SST Time series and trends over Tropical Atlantic and Northern Indian Ocean, with discussions > 4.1 Plot Annual trend maps", "title": "Satellite Sea Surface Temperature Consistency in representing Ocean Warming Trends", "chunk_index": 6, "token_count": 729, "text_raw": "In this section we plot the annual trend maps for each product, discussing differences.\n\n```text\nESACCI in Mediterranean Sea: 0.042 ± 0.014 °C/year\nESACCI in Baltic Sea: 0.044 ± 0.016 °C/year\nESACCI in Black Sea: 0.058 ± 0.018 °C/year\nESACCI in Global: 0.022 ± 0.013 °C/year\nESACCI in Tropical Atlantic: 0.033 ± 0.01 °C/year\nESACCI in Northern Indian Ocean: 0.03 ± 0.008 °C/year\nGMPE in Mediterranean Sea: 0.037 ± 0.012 °C/year\nGMPE in Baltic Sea: 0.053 ± 0.015 °C/year\nGMPE in Black Sea: 0.056 ± 0.015 °C/year\nGMPE in Global: 0.018 ± 0.01 °C/year\nGMPE in Tropical Atlantic: 0.025 ± 0.0095 °C/year\nGMPE in Northern Indian Ocean: 0.019 ± 0.0083 °C/year\n```\n\nThe global SST trend map has been computed over 30 years from 1982 to 2011 at the 95% confidence level (grey dots mark non-significant trend areas). SST exhibits an overall positive (warming) trend over the Northern Hemisphere, where it can reach $\\simeq$0.05 °C/year, while the Southern Hemisphere is characterized by less significant and negative values. A large area covering the Eastern Pacific Ocean is characterized by no significant trends (i.e., p $\\geq$ 0.05) with few sparse significant values in both ESA CCI and GMPE. The North Atlantic ocean and the European seas stand out showing the most intense SST trend values, as e.g. the Mediterranean Sea (GMPE = 0.04 $\\pm$ 0.01 °C/year, ESA CCI = 0.04 $\\pm$ 0.01 °C/year), Baltic Sea (GMPE = 0.05 $\\pm$ 0.02 °C/year, ESA CCI = 0.04 $\\pm$ 0.02 °C/year) and Black Sea (GMPE = 0.06 $\\pm$ 0.01 °C/year, ESA CCI = 0.06 $\\pm$ 0.02 °C/year). The Mediterranean SST trend is in good agreement with the one reported in [[5]](https://doi.org/10.3390/rs12010132), while Baltic Sea and Black Sea trends are similar to those found in [[6]](https://doi.org/10.1080/1755876X.2018.1489208), but over a shorter period. In the Black and Mediterranean Sea, trend estimates between ESA CCI and GMPE are more in agreement than in the Baltic Sea.", "text_with_prefix": "EQC Quality Assessment: \"Satellite Sea Surface Temperature Consistency in representing Ocean Warming Trends\"\nDataset: satellite-sea-surface-temperature [CDS]\nAspect: consistency_q02 | Category: Satellite_ECVs\nSection: Satellite Sea Surface Temperature Consistency in representing Ocean Warming Trends > Analysis and results > 4. Plot Annual and Seasonal trends, SST Time series and trends over Tropical Atlantic and Northern Indian Ocean, with discussions > 4.1 Plot Annual trend maps\n---\nIn this section we plot the annual trend maps for each product, discussing differences.\n\n```text\nESACCI in Mediterranean Sea: 0.042 ± 0.014 °C/year\nESACCI in Baltic Sea: 0.044 ± 0.016 °C/year\nESACCI in Black Sea: 0.058 ± 0.018 °C/year\nESACCI in Global: 0.022 ± 0.013 °C/year\nESACCI in Tropical Atlantic: 0.033 ± 0.01 °C/year\nESACCI in Northern Indian Ocean: 0.03 ± 0.008 °C/year\nGMPE in Mediterranean Sea: 0.037 ± 0.012 °C/year\nGMPE in Baltic Sea: 0.053 ± 0.015 °C/year\nGMPE in Black Sea: 0.056 ± 0.015 °C/year\nGMPE in Global: 0.018 ± 0.01 °C/year\nGMPE in Tropical Atlantic: 0.025 ± 0.0095 °C/year\nGMPE in Northern Indian Ocean: 0.019 ± 0.0083 °C/year\n```\n\nThe global SST trend map has been computed over 30 years from 1982 to 2011 at the 95% confidence level (grey dots mark non-significant trend areas). SST exhibits an overall positive (warming) trend over the Northern Hemisphere, where it can reach $\\simeq$0.05 °C/year, while the Southern Hemisphere is characterized by less significant and negative values. A large area covering the Eastern Pacific Ocean is characterized by no significant trends (i.e., p $\\geq$ 0.05) with few sparse significant values in both ESA CCI and GMPE. The North Atlantic ocean and the European seas stand out showing the most intense SST trend values, as e.g. the Mediterranean Sea (GMPE = 0.04 $\\pm$ 0.01 °C/year, ESA CCI = 0.04 $\\pm$ 0.01 °C/year), Baltic Sea (GMPE = 0.05 $\\pm$ 0.02 °C/year, ESA CCI = 0.04 $\\pm$ 0.02 °C/year) and Black Sea (GMPE = 0.06 $\\pm$ 0.01 °C/year, ESA CCI = 0.06 $\\pm$ 0.02 °C/year). The Mediterranean SST trend is in good agreement with the one reported in [[5]](https://doi.org/10.3390/rs12010132), while Baltic Sea and Black Sea trends are similar to those found in [[6]](https://doi.org/10.1080/1755876X.2018.1489208), but over a shorter period. In the Black and Mediterranean Sea, trend estimates between ESA CCI and GMPE are more in agreement than in the Baltic Sea."} {"chunk_id": "satellite_satellite-sea-surface-temperature_consistency_q02__60d7497a2e28", "report_id": "satellite_satellite-sea-surface-temperature_consistency_q02", "dataset_id": "satellite-sea-surface-temperature", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q02", "aspect_base": "consistency", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Sea Surface Temperature Consistency in representing Ocean Warming Trends > Analysis and results > 4. Plot Annual and Seasonal trends, SST Time series and trends over Tropical Atlantic and Northern Indian Ocean, with discussions > 4.2 Seasonal Trends", "title": "Satellite Sea Surface Temperature Consistency in representing Ocean Warming Trends", "chunk_index": 7, "token_count": 506, "text_raw": "In this section we plot seasonal trends for each product, discussing the differences.\n\nChoose timeseries\nprint(f\"{product=}\")\nMann Kendall settings\nPlot seasonal trends\n\nSeasonal maps show an intensification of the trend from Spring to Autumn and an attenuation during winter. Overall the agreement between GMPE and ESA CCI SST is quite good though the former dataset provides less intense trend estimates. Both datasets’ estimates yield a Northern Hemisphere which has warmed up at a higher rate than the Southern Hemisphere, especially in Boreal Summer and Fall.\n\nThe globally averaged mean trend is estimated in 0.018 $\\pm$ 0.01 °C/year for GMPE, while for ESA CCI we have 0.022 $\\pm$ 0.013 °C/year, which corresponds to an average total increase of about 0.54 °C in GMPE and 0.66 °C in ESA CCI over this 30-year period (1982-2011). Within the 95% confidence interval, both datasets provide consistent estimates of global warming trends. These values are also reasonably consistent (within respective errors) with the mean global trend estimate of 0.011 °C/year from 1980 to 2005 as reported in one of the last IPCC reports [[7]](http://hdl.handle.net/10013/epic.45156.d001), given the different time period on which the trend is calculated. Regionally, estimates of trend slopes yield values for the Northern Indian Ocean (GMPE=0.019 $\\pm$ 0.008 °C/year, ESA CCI=0.030 $\\pm$ 0.008 °C/year) and Tropical Atlantic (GMPE= 0.025 $\\pm$ 0.009 °C/year, ESA CCI=0.003 $\\pm$ 0.01 °C/year), which are consistent with each others in the limits of their respective uncertainties, though are areas of main disagreement for trend estimates.", "text_with_prefix": "EQC Quality Assessment: \"Satellite Sea Surface Temperature Consistency in representing Ocean Warming Trends\"\nDataset: satellite-sea-surface-temperature [CDS]\nAspect: consistency_q02 | Category: Satellite_ECVs\nSection: Satellite Sea Surface Temperature Consistency in representing Ocean Warming Trends > Analysis and results > 4. Plot Annual and Seasonal trends, SST Time series and trends over Tropical Atlantic and Northern Indian Ocean, with discussions > 4.2 Seasonal Trends\n---\nIn this section we plot seasonal trends for each product, discussing the differences.\n\nChoose timeseries\nprint(f\"{product=}\")\nMann Kendall settings\nPlot seasonal trends\n\nSeasonal maps show an intensification of the trend from Spring to Autumn and an attenuation during winter. Overall the agreement between GMPE and ESA CCI SST is quite good though the former dataset provides less intense trend estimates. Both datasets’ estimates yield a Northern Hemisphere which has warmed up at a higher rate than the Southern Hemisphere, especially in Boreal Summer and Fall.\n\nThe globally averaged mean trend is estimated in 0.018 $\\pm$ 0.01 °C/year for GMPE, while for ESA CCI we have 0.022 $\\pm$ 0.013 °C/year, which corresponds to an average total increase of about 0.54 °C in GMPE and 0.66 °C in ESA CCI over this 30-year period (1982-2011). Within the 95% confidence interval, both datasets provide consistent estimates of global warming trends. These values are also reasonably consistent (within respective errors) with the mean global trend estimate of 0.011 °C/year from 1980 to 2005 as reported in one of the last IPCC reports [[7]](http://hdl.handle.net/10013/epic.45156.d001), given the different time period on which the trend is calculated. Regionally, estimates of trend slopes yield values for the Northern Indian Ocean (GMPE=0.019 $\\pm$ 0.008 °C/year, ESA CCI=0.030 $\\pm$ 0.008 °C/year) and Tropical Atlantic (GMPE= 0.025 $\\pm$ 0.009 °C/year, ESA CCI=0.003 $\\pm$ 0.01 °C/year), which are consistent with each others in the limits of their respective uncertainties, though are areas of main disagreement for trend estimates."} {"chunk_id": "satellite_satellite-sea-surface-temperature_consistency_q02__551d2229e860", "report_id": "satellite_satellite-sea-surface-temperature_consistency_q02", "dataset_id": "satellite-sea-surface-temperature", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q02", "aspect_base": "consistency", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Sea Surface Temperature Consistency in representing Ocean Warming Trends > Analysis and results > 4. Plot Annual and Seasonal trends, SST Time series and trends over Tropical Atlantic and Northern Indian Ocean, with discussions > 4.3 SST Time series and trends over Tropical Atlantic and Northern Indian Ocean", "title": "Satellite Sea Surface Temperature Consistency in representing Ocean Warming Trends", "chunk_index": 8, "token_count": 1127, "text_raw": "Trend analyses shown in the previous section evidenced the existence of disagreement areas between ESA CCI and GMPE SST datasets. In particular, the Northern Indian Ocean and the Tropical Atlantic exhibit a larger SST trend for the ESA CCI dataset, when compared to GMPE. The ESA CCI and GMPE SST monthly time series were then intercompared in those areas, in order to have further insights on these discrepancies.\n\n```text\nproduct='ESACCI' region='global'\n```\n\n```text\nyear: 100%|██████████| 35/35 [00:09<00:00, 3.72it/s]\n```\n\n```text\nproduct='ESACCI' region='med sea'\n```\n\n```text\nyear: 100%|██████████| 35/35 [00:09<00:00, 3.58it/s]\n```\n\n```text\nproduct='ESACCI' region='baltic sea'\n```\n\n```text\nyear: 100%|██████████| 35/35 [00:05<00:00, 6.27it/s]\n```\n\n```text\nproduct='ESACCI' region='black sea'\n```\n\n```text\nyear: 100%|██████████| 35/35 [00:05<00:00, 6.41it/s]\n```\n\n```text\nproduct='ESACCI' region='Tropical Atlantic'\n```\n\n```text\nyear: 100%|██████████| 35/35 [00:03<00:00, 9.27it/s]\n```\n\n```text\nproduct='ESACCI' region='Northern Indian Ocean'\n```\n\n```text\nyear: 100%|██████████| 35/35 [00:03<00:00, 8.81it/s]\n```\n\n```text\nproduct='GMPE' region='global'\n```\n\n```text\nyear: 100%|██████████| 35/35 [00:06<00:00, 5.48it/s]\n```\n\n```text\nproduct='GMPE' region='med sea'\n```\n\n```text\nyear: 100%|██████████| 35/35 [00:04<00:00, 7.60it/s]\n```\n\n```text\nproduct='GMPE' region='baltic sea'\n```\n\n```text\nyear: 100%|██████████| 35/35 [00:04<00:00, 7.88it/s]\n```\n\n```text\nproduct='GMPE' region='black sea'\n```\n\n```text\nyear: 100%|██████████| 35/35 [00:05<00:00, 6.97it/s]\n```\n\n```text\nproduct='GMPE' region='Tropical Atlantic'\n```\n\n```text\nyear: 100%|██████████| 35/35 [00:07<00:00, 4.54it/s]\n```\n\n```text\nproduct='GMPE' region='Northern Indian Ocean'\n```\n\n```text\nyear: 100%|██████████| 35/35 [00:54<00:00, 1.56s/it]\n```\n\n```text\nESACCI in Northern Indian Ocean: 0.0018 ± 0.0004 °C/year\nGMPE in Northern Indian Ocean: 0.0012 ± 0.0005 °C/year\n```\n\n```text\nESACCI in Tropical Atlantic: 0.0024 ± 0.0004 °C/year\nGMPE in Tropical Atlantic: 0.002 ± 0.0004 °C/year\n```\n\nIn the 1982-1991 decade, the ESA CCI SSTs are generally colder than the GMPE ones, as sketched by the monthly time series presented in this section. This is also highlighted by the ∆SST time series (given in yellow), whose values can reach 1°C/0.6°C in the Indian Ocean and Tropical Atlantic, respectively, in the aforementioned temporal window. Such discrepancies are significantly reduced afterwards, thus resulting in a larger SST slope from the ESA CCI CDR, when compared with GMPE. This is a known issue of the ESA CCI dataset, as reported in [[4]](https://doi.org/10.1038/s41597-019-0236-x): unscreened dust events caused negative biases in the Advanced Very High Resolution Radiometer (AVHRR) SSTs in the north East Tropical Atlantic and in the Arabian/Red Sea. This issue covers the first decade of the CDR, resulting in anomalously high positive SST trends in these regions.", "text_with_prefix": "EQC Quality Assessment: \"Satellite Sea Surface Temperature Consistency in representing Ocean Warming Trends\"\nDataset: satellite-sea-surface-temperature [CDS]\nAspect: consistency_q02 | Category: Satellite_ECVs\nSection: Satellite Sea Surface Temperature Consistency in representing Ocean Warming Trends > Analysis and results > 4. Plot Annual and Seasonal trends, SST Time series and trends over Tropical Atlantic and Northern Indian Ocean, with discussions > 4.3 SST Time series and trends over Tropical Atlantic and Northern Indian Ocean\n---\nTrend analyses shown in the previous section evidenced the existence of disagreement areas between ESA CCI and GMPE SST datasets. In particular, the Northern Indian Ocean and the Tropical Atlantic exhibit a larger SST trend for the ESA CCI dataset, when compared to GMPE. The ESA CCI and GMPE SST monthly time series were then intercompared in those areas, in order to have further insights on these discrepancies.\n\n```text\nproduct='ESACCI' region='global'\n```\n\n```text\nyear: 100%|██████████| 35/35 [00:09<00:00, 3.72it/s]\n```\n\n```text\nproduct='ESACCI' region='med sea'\n```\n\n```text\nyear: 100%|██████████| 35/35 [00:09<00:00, 3.58it/s]\n```\n\n```text\nproduct='ESACCI' region='baltic sea'\n```\n\n```text\nyear: 100%|██████████| 35/35 [00:05<00:00, 6.27it/s]\n```\n\n```text\nproduct='ESACCI' region='black sea'\n```\n\n```text\nyear: 100%|██████████| 35/35 [00:05<00:00, 6.41it/s]\n```\n\n```text\nproduct='ESACCI' region='Tropical Atlantic'\n```\n\n```text\nyear: 100%|██████████| 35/35 [00:03<00:00, 9.27it/s]\n```\n\n```text\nproduct='ESACCI' region='Northern Indian Ocean'\n```\n\n```text\nyear: 100%|██████████| 35/35 [00:03<00:00, 8.81it/s]\n```\n\n```text\nproduct='GMPE' region='global'\n```\n\n```text\nyear: 100%|██████████| 35/35 [00:06<00:00, 5.48it/s]\n```\n\n```text\nproduct='GMPE' region='med sea'\n```\n\n```text\nyear: 100%|██████████| 35/35 [00:04<00:00, 7.60it/s]\n```\n\n```text\nproduct='GMPE' region='baltic sea'\n```\n\n```text\nyear: 100%|██████████| 35/35 [00:04<00:00, 7.88it/s]\n```\n\n```text\nproduct='GMPE' region='black sea'\n```\n\n```text\nyear: 100%|██████████| 35/35 [00:05<00:00, 6.97it/s]\n```\n\n```text\nproduct='GMPE' region='Tropical Atlantic'\n```\n\n```text\nyear: 100%|██████████| 35/35 [00:07<00:00, 4.54it/s]\n```\n\n```text\nproduct='GMPE' region='Northern Indian Ocean'\n```\n\n```text\nyear: 100%|██████████| 35/35 [00:54<00:00, 1.56s/it]\n```\n\n```text\nESACCI in Northern Indian Ocean: 0.0018 ± 0.0004 °C/year\nGMPE in Northern Indian Ocean: 0.0012 ± 0.0005 °C/year\n```\n\n```text\nESACCI in Tropical Atlantic: 0.0024 ± 0.0004 °C/year\nGMPE in Tropical Atlantic: 0.002 ± 0.0004 °C/year\n```\n\nIn the 1982-1991 decade, the ESA CCI SSTs are generally colder than the GMPE ones, as sketched by the monthly time series presented in this section. This is also highlighted by the ∆SST time series (given in yellow), whose values can reach 1°C/0.6°C in the Indian Ocean and Tropical Atlantic, respectively, in the aforementioned temporal window. Such discrepancies are significantly reduced afterwards, thus resulting in a larger SST slope from the ESA CCI CDR, when compared with GMPE. This is a known issue of the ESA CCI dataset, as reported in [[4]](https://doi.org/10.1038/s41597-019-0236-x): unscreened dust events caused negative biases in the Advanced Very High Resolution Radiometer (AVHRR) SSTs in the north East Tropical Atlantic and in the Arabian/Red Sea. This issue covers the first decade of the CDR, resulting in anomalously high positive SST trends in these regions."} {"chunk_id": "satellite_satellite-sea-surface-temperature_consistency_q02__1d0b6efdecac", "report_id": "satellite_satellite-sea-surface-temperature_consistency_q02", "dataset_id": "satellite-sea-surface-temperature", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q02", "aspect_base": "consistency", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Sea Surface Temperature Consistency in representing Ocean Warming Trends > ℹ️ If you want to know more > Key resources", "title": "Satellite Sea Surface Temperature Consistency in representing Ocean Warming Trends", "chunk_index": 9, "token_count": 238, "text_raw": "SST CCI Climate Assessment Report (CAR): https://climate.esa.int/media/documents/SST_CCI_D5.1_CAR_v1.1-signed.pdf\n\nGHRSST Website: https://www.ghrsst.org/\n\nAdditional sources for SST ensemble intercomparison statistics: https://ghrsst-pp.metoffice.gov.uk/ostia-website/gmpe-monitoring.html\n\nCode libraries used:\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)\n* python packages: matplotlib, xarray, cartopy, numpy, tqdm", "text_with_prefix": "EQC Quality Assessment: \"Satellite Sea Surface Temperature Consistency in representing Ocean Warming Trends\"\nDataset: satellite-sea-surface-temperature [CDS]\nAspect: consistency_q02 | Category: Satellite_ECVs\nSection: Satellite Sea Surface Temperature Consistency in representing Ocean Warming Trends > ℹ️ If you want to know more > Key resources\n---\nSST CCI Climate Assessment Report (CAR): https://climate.esa.int/media/documents/SST_CCI_D5.1_CAR_v1.1-signed.pdf\n\nGHRSST Website: https://www.ghrsst.org/\n\nAdditional sources for SST ensemble intercomparison statistics: https://ghrsst-pp.metoffice.gov.uk/ostia-website/gmpe-monitoring.html\n\nCode libraries used:\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)\n* python packages: matplotlib, xarray, cartopy, numpy, tqdm"} {"chunk_id": "satellite_satellite-sea-surface-temperature_consistency_q02__8be6f4c3bbea", "report_id": "satellite_satellite-sea-surface-temperature_consistency_q02", "dataset_id": "satellite-sea-surface-temperature", "store": "CDS", "doc_type": "EQC_QA", "aspect": "consistency_q02", "aspect_base": "consistency", "category": "Satellite_ECVs", "match_confidence": "exact", "section": "Satellite Sea Surface Temperature Consistency in representing Ocean Warming Trends > ℹ️ If you want to know more > References", "title": "Satellite Sea Surface Temperature Consistency in representing Ocean Warming Trends", "chunk_index": 10, "token_count": 999, "text_raw": "[[1]](https://doi.org/10.1007/s00382-014-2147-z) Kang, S.M., Seager, R., Frierson, D.M.W. et al. Croll revisited: Why is the northern hemisphere warmer than the southern hemisphere?. Clim Dyn 44, 1457–1472 (2015). https://doi.org/10.1007/s00382-014-2147-z\n\n[[2]](https://doi.org/10.1175/JCLI-D-20-0793.1) Yang, C., Leonelli, F. E., Marullo, S., Artale, V., Beggs, H., Nardelli, B. B., ... & Pisano, A. (2021). Sea surface temperature intercomparison in the framework of the Copernicus Climate Change Service (C3S). Journal of Climate, 34(13), 5257-5283\n\n[[3]](https://doi.org/10.1016/j.rse.2018.12.015) Fiedler, E. K., McLaren, A., Banzon, V., Brasnett, B., Ishizaki, S., Kennedy, J., ... & Donlon, C. (2019). Intercomparison of long-term sea surface temperature analyses using the GHRSST Multi-Product Ensemble (GMPE) system. Remote sensing of environment, 222, 18-33\n\n[[4]](https://doi.org/10.1038/s41597-019-0236-x) Merchant, C. J., Embury, O., Bulgin, C. E., Block, T., Corlett, G. K., Fiedler, E., ... & Donlon, C. (2019). Satellite-based time-series of sea-surface temperature since 1981 for climate applications. Scientific data, 6(1), 223. https://doi.org/10.1038/s41597-019-0236-x\n\n[[5]](https://doi.org/10.3390/rs12010132) Pisano, A., Marullo, S., Artale, V., Falcini, F., Yang, C., Leonelli, F. E., Santoleri, R. and Buongiorno Nardelli, B.: New Evidence of Mediterranean Climate Change and Variability from Sea Surface Temperature Observations, Remote Sens., 12(1), 132, doi:10.3390/rs12010132, 2020\n\n[[6]](https://doi.org/10.1080/1755876X.2018.1489208) Mulet, S., Buongiorno Nardelli, B., Good, S., Pisano, A., Greiner, E., Monier, M., Autret, E., Axell, L., Boberg, F., Ciliberti, S., Drévillon, M., Droghei, R., Embury, O., Gourrion, J., Høyer, J., Juza, M., Kennedy, J., Lemieux-Dudon, B., Peneva, E., Reid, R., Simoncelli, S., Storto, A., Tinker, J., Von Schuckmann, K., Wakelin, S. L., 2018. Ocean temperature and salinity. In: Copernicus Marine Service Ocean State Report, Issue 2, Journal of Operational Oceanography, 11:sup1, s5–s13, DOI: 10.1080/1755876X.2018.1489208\n\n[[7]](http://hdl.handle.net/10013/epic.45156.d001) Pachauri, R. K., Allen, M. R., Barros, V. R., Broome, J., Cramer, W., Christ, R., Church, J. A., Clarke, L., Dahe, Q., Dasgupta, P., Dubash, N. K., et al.: Climate Change 2014: Synthesis Report. Contribution of Working Groups I, II and III to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change, edited by: Pachauri, R. and Meyer, L., Geneva, Switzerland, IPCC, ISBN 978-92-9169-143-2, 2014", "text_with_prefix": "EQC Quality Assessment: \"Satellite Sea Surface Temperature Consistency in representing Ocean Warming Trends\"\nDataset: satellite-sea-surface-temperature [CDS]\nAspect: consistency_q02 | Category: Satellite_ECVs\nSection: Satellite Sea Surface Temperature Consistency in representing Ocean Warming Trends > ℹ️ If you want to know more > References\n---\n[[1]](https://doi.org/10.1007/s00382-014-2147-z) Kang, S.M., Seager, R., Frierson, D.M.W. et al. Croll revisited: Why is the northern hemisphere warmer than the southern hemisphere?. Clim Dyn 44, 1457–1472 (2015). https://doi.org/10.1007/s00382-014-2147-z\n\n[[2]](https://doi.org/10.1175/JCLI-D-20-0793.1) Yang, C., Leonelli, F. E., Marullo, S., Artale, V., Beggs, H., Nardelli, B. B., ... & Pisano, A. (2021). Sea surface temperature intercomparison in the framework of the Copernicus Climate Change Service (C3S). Journal of Climate, 34(13), 5257-5283\n\n[[3]](https://doi.org/10.1016/j.rse.2018.12.015) Fiedler, E. K., McLaren, A., Banzon, V., Brasnett, B., Ishizaki, S., Kennedy, J., ... & Donlon, C. (2019). Intercomparison of long-term sea surface temperature analyses using the GHRSST Multi-Product Ensemble (GMPE) system. Remote sensing of environment, 222, 18-33\n\n[[4]](https://doi.org/10.1038/s41597-019-0236-x) Merchant, C. J., Embury, O., Bulgin, C. E., Block, T., Corlett, G. K., Fiedler, E., ... & Donlon, C. (2019). Satellite-based time-series of sea-surface temperature since 1981 for climate applications. Scientific data, 6(1), 223. https://doi.org/10.1038/s41597-019-0236-x\n\n[[5]](https://doi.org/10.3390/rs12010132) Pisano, A., Marullo, S., Artale, V., Falcini, F., Yang, C., Leonelli, F. E., Santoleri, R. and Buongiorno Nardelli, B.: New Evidence of Mediterranean Climate Change and Variability from Sea Surface Temperature Observations, Remote Sens., 12(1), 132, doi:10.3390/rs12010132, 2020\n\n[[6]](https://doi.org/10.1080/1755876X.2018.1489208) Mulet, S., Buongiorno Nardelli, B., Good, S., Pisano, A., Greiner, E., Monier, M., Autret, E., Axell, L., Boberg, F., Ciliberti, S., Drévillon, M., Droghei, R., Embury, O., Gourrion, J., Høyer, J., Juza, M., Kennedy, J., Lemieux-Dudon, B., Peneva, E., Reid, R., Simoncelli, S., Storto, A., Tinker, J., Von Schuckmann, K., Wakelin, S. L., 2018. Ocean temperature and salinity. In: Copernicus Marine Service Ocean State Report, Issue 2, Journal of Operational Oceanography, 11:sup1, s5–s13, DOI: 10.1080/1755876X.2018.1489208\n\n[[7]](http://hdl.handle.net/10013/epic.45156.d001) Pachauri, R. K., Allen, M. R., Barros, V. R., Broome, J., Cramer, W., Christ, R., Church, J. A., Clarke, L., Dahe, Q., Dasgupta, P., Dubash, N. K., et al.: Climate Change 2014: Synthesis Report. Contribution of Working Groups I, II and III to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change, edited by: Pachauri, R. and Meyer, L., Geneva, Switzerland, IPCC, ISBN 978-92-9169-143-2, 2014"} {"chunk_id": "seasonal_seasonal-monthly-ocean_extremes-detection_q07__5d82eab0c5e8", "report_id": "seasonal_seasonal-monthly-ocean_extremes-detection_q07", "dataset_id": "seasonal-monthly-ocean", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extremes-detection_q07", "aspect_base": "extremes-detection", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Seasonal Forecast Prediction of Upper Ocean Heat Content Extremes > Quality assessment question", "title": "Seasonal Forecast Prediction of Upper Ocean Heat Content Extremes", "chunk_index": 0, "token_count": 474, "text_raw": "* **What is the forecast skill of the seasonal prediction system for detecting extreme upper ocean heat content events in the NDJ season, from October initializations?**\n\nThis notebook evaluates the skill of the Météo-France System-9 seasonal forecasting system in predicting upper ocean heat content extremes. We chose this system as an example since it had high data coverage of 300-m integrated potential temperature. The analysis uses:\n- **Seasonal Forecast Data**: Monthly depth-averaged potential temperature for the upper 300 m, from Météo-France System-9 forecasts. [[doi:10.24381/cds.2f9be611]](https://doi.org/10.24381/cds.2f9be611)\n- **Reanalysis Data**: Ocean heat content from the ORAS5 global ocean and sea-ice reanalysis dataset. [[doi:10.24381/cds.67e8eeb7]](https://doi.org/10.24381/cds.67e8eeb7)\n\nWe consider the skill in predicting upper ocean heat content extremes in November, December, and January between 1993 and 2023 for (re-)forecasts with a nominal start date October 1st, using metrics such as the Symmetric Extremal Dependence Index (SEDI) and the Brier Skill Score (BSS). The metrics considered follow Jacox et. al. 2022, [[1]](https://doi.org/10.1038/s41586-022-04573-9), however in this work we look at upper ocean heat content and not sea surface temperature. Since ocean heat content (OHC) anomalies may persist for several months, this variable is an important component of seasonal predictability, containing relevant information even at a monthly time-scale (McAdam et. al. 2022, [[2]](https://doi.org/10.1007/S00382-021-06101-3)).", "text_with_prefix": "EQC Quality Assessment: \"Seasonal Forecast Prediction of Upper Ocean Heat Content Extremes\"\nDataset: seasonal-monthly-ocean [CDS]\nAspect: extremes-detection_q07 | Category: Seasonal_Forecasts\nSection: Seasonal Forecast Prediction of Upper Ocean Heat Content Extremes > Quality assessment question\n---\n* **What is the forecast skill of the seasonal prediction system for detecting extreme upper ocean heat content events in the NDJ season, from October initializations?**\n\nThis notebook evaluates the skill of the Météo-France System-9 seasonal forecasting system in predicting upper ocean heat content extremes. We chose this system as an example since it had high data coverage of 300-m integrated potential temperature. The analysis uses:\n- **Seasonal Forecast Data**: Monthly depth-averaged potential temperature for the upper 300 m, from Météo-France System-9 forecasts. [[doi:10.24381/cds.2f9be611]](https://doi.org/10.24381/cds.2f9be611)\n- **Reanalysis Data**: Ocean heat content from the ORAS5 global ocean and sea-ice reanalysis dataset. [[doi:10.24381/cds.67e8eeb7]](https://doi.org/10.24381/cds.67e8eeb7)\n\nWe consider the skill in predicting upper ocean heat content extremes in November, December, and January between 1993 and 2023 for (re-)forecasts with a nominal start date October 1st, using metrics such as the Symmetric Extremal Dependence Index (SEDI) and the Brier Skill Score (BSS). The metrics considered follow Jacox et. al. 2022, [[1]](https://doi.org/10.1038/s41586-022-04573-9), however in this work we look at upper ocean heat content and not sea surface temperature. Since ocean heat content (OHC) anomalies may persist for several months, this variable is an important component of seasonal predictability, containing relevant information even at a monthly time-scale (McAdam et. al. 2022, [[2]](https://doi.org/10.1007/S00382-021-06101-3))."} {"chunk_id": "seasonal_seasonal-monthly-ocean_extremes-detection_q07__d35191827340", "report_id": "seasonal_seasonal-monthly-ocean_extremes-detection_q07", "dataset_id": "seasonal-monthly-ocean", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extremes-detection_q07", "aspect_base": "extremes-detection", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Seasonal Forecast Prediction of Upper Ocean Heat Content Extremes > Quality assessment statement", "title": "Seasonal Forecast Prediction of Upper Ocean Heat Content Extremes", "chunk_index": 1, "token_count": 1048, "text_raw": "These are the key outcomes of this assessment\n\n- The Météo-France System-9 forecast model demonstrates skill in detecting the risk of NDJ (November-January) upper ocean heatwaves. The model successfully captures 33% of observed heatwaves in November, with hit rates remaining consistent through December (30%) and January (27%).\n- The forecast is useful for identifying broad, large-scale patterns of risk, particularly during major El Niño events. However, the Brier Skill Score (BSS) highlights that its probabilistic reliability varies by region. Using a multi-model forecast ensemble would likely increase the spread of the forecasts and provide more reliable results outside the ENSO region.\n- The system has potential as a tool for anticipating and building resilience to marine heatwaves, especially in ENSO-sensitive regions. This analysis could be extended to other C3S seasonal systems for which the required data are available.\n\nattachment:0fa94a15-3175-4725-863a-fc42e5bcb6c5.png\n---\nheight: 600px\n---\nFigure 1: Spatial Forecast Skill for November-December-January (NDJ) Upper Ocean Heatwaves (1993–2023).\nSkill of Météo-France System-9 forecasts (with nominal start date October 1st) verified against ORAS5 reanalysis. Heatwaves are defined as upper 300m ocean heat content (OHC) anomalies exceeding the 90th percentile within the 1993-2023 period. Rows show: (A) Symmetric Extremal Dependence Index (SEDI), (B) Brier Skill Score (BSS); and (C) Accuracy. In all maps, red areas indicate positive skill (forecast is better than the reference), while blue areas indicate negative skill. Masked areas are shown in grey. For the Accuracy maps (C) the areas with an accuracy lower than what would be expected by random chance (82%) are also shown with a grey color.\n\n```\n\n## 📋 Methodology\n\nThe following steps were applied to evaluate the skill of the seasonal forecasting system in predicting heatwave extremes:\n\n[](seasonal_ohc:section-1)\n\nThis section includes:\n- imports, settings, parameter definitions, and reusable functions required to initialize the analysis. Here the extent of the Nino3.4 area is given (5°N–5°S, 170°–120°W, also delineated in most map plots, e.g. [](fig-corr-map)). This area is selected to represent the tropical Pacific and ENSO, providing a key benchmark for seasonal predictability. The data is also loaded in this section.\n\n- **Météo-France System-9**: The seasonal forecast data includes monthly depth-averaged potential temperature for the upper 300m. We use the hindcasts with a nominal start date of October 1st, between 1993 and 2023. For consistency only the hindcast period is used, since the model system has a different number of ensemble members in the forecast period compared to the hindcast period. *depth_average_potential_temperature_of_upper_300m* ($\\bar{\\Theta}$) was converted to OHC using the similar constants as used by ORAS5 (pers. comm.):\n$\\text{OHC} =\\rho \\, c_p \\, \\bar{ \\Theta} \\, dz $, where $ \\rho = 1026 \\, \\text{kg/m}^3$, $ c_p = 3991 \\, \\text{J/kg/K} $, and $ dz = 300 \\, \\text{m} $. \n- **ORAS5**: Reanalysis OHC for the upper 300m is used as the observational reference. Data is regridded to a common 1° spatial resolution using conservative normalization in xESMF.\n\n- The seasonal forecast and reanalysis data were processed for the period from **1993 to 2023**; and we consider forecasts **with a nominal start date October 1st, valid for November, December, and January**. For instance, the last season considers forecasts with a nominal start date October 1st 2023 for November and December 2023, and January 2024. By choosing an October initialization, we avoid compounding the task of extreme event detection with the lower predictability associated with other initialization windows, such as the boreal spring period in the Eastern Tropical Pacific (Jacox et. al. 2022, [[1]](https://doi.org/10.1038/s41586-022-04573-9)).\n\n- **Anomalies**: Monthly anomalies are calculated by subtracting the calendar month climatology.", "text_with_prefix": "EQC Quality Assessment: \"Seasonal Forecast Prediction of Upper Ocean Heat Content Extremes\"\nDataset: seasonal-monthly-ocean [CDS]\nAspect: extremes-detection_q07 | Category: Seasonal_Forecasts\nSection: Seasonal Forecast Prediction of Upper Ocean Heat Content Extremes > Quality assessment statement\n---\nThese are the key outcomes of this assessment\n\n- The Météo-France System-9 forecast model demonstrates skill in detecting the risk of NDJ (November-January) upper ocean heatwaves. The model successfully captures 33% of observed heatwaves in November, with hit rates remaining consistent through December (30%) and January (27%).\n- The forecast is useful for identifying broad, large-scale patterns of risk, particularly during major El Niño events. However, the Brier Skill Score (BSS) highlights that its probabilistic reliability varies by region. Using a multi-model forecast ensemble would likely increase the spread of the forecasts and provide more reliable results outside the ENSO region.\n- The system has potential as a tool for anticipating and building resilience to marine heatwaves, especially in ENSO-sensitive regions. This analysis could be extended to other C3S seasonal systems for which the required data are available.\n\nattachment:0fa94a15-3175-4725-863a-fc42e5bcb6c5.png\n---\nheight: 600px\n---\nFigure 1: Spatial Forecast Skill for November-December-January (NDJ) Upper Ocean Heatwaves (1993–2023).\nSkill of Météo-France System-9 forecasts (with nominal start date October 1st) verified against ORAS5 reanalysis. Heatwaves are defined as upper 300m ocean heat content (OHC) anomalies exceeding the 90th percentile within the 1993-2023 period. Rows show: (A) Symmetric Extremal Dependence Index (SEDI), (B) Brier Skill Score (BSS); and (C) Accuracy. In all maps, red areas indicate positive skill (forecast is better than the reference), while blue areas indicate negative skill. Masked areas are shown in grey. For the Accuracy maps (C) the areas with an accuracy lower than what would be expected by random chance (82%) are also shown with a grey color.\n\n```\n\n## 📋 Methodology\n\nThe following steps were applied to evaluate the skill of the seasonal forecasting system in predicting heatwave extremes:\n\n[](seasonal_ohc:section-1)\n\nThis section includes:\n- imports, settings, parameter definitions, and reusable functions required to initialize the analysis. Here the extent of the Nino3.4 area is given (5°N–5°S, 170°–120°W, also delineated in most map plots, e.g. [](fig-corr-map)). This area is selected to represent the tropical Pacific and ENSO, providing a key benchmark for seasonal predictability. The data is also loaded in this section.\n\n- **Météo-France System-9**: The seasonal forecast data includes monthly depth-averaged potential temperature for the upper 300m. We use the hindcasts with a nominal start date of October 1st, between 1993 and 2023. For consistency only the hindcast period is used, since the model system has a different number of ensemble members in the forecast period compared to the hindcast period. *depth_average_potential_temperature_of_upper_300m* ($\\bar{\\Theta}$) was converted to OHC using the similar constants as used by ORAS5 (pers. comm.):\n$\\text{OHC} =\\rho \\, c_p \\, \\bar{ \\Theta} \\, dz $, where $ \\rho = 1026 \\, \\text{kg/m}^3$, $ c_p = 3991 \\, \\text{J/kg/K} $, and $ dz = 300 \\, \\text{m} $. \n- **ORAS5**: Reanalysis OHC for the upper 300m is used as the observational reference. Data is regridded to a common 1° spatial resolution using conservative normalization in xESMF.\n\n- The seasonal forecast and reanalysis data were processed for the period from **1993 to 2023**; and we consider forecasts **with a nominal start date October 1st, valid for November, December, and January**. For instance, the last season considers forecasts with a nominal start date October 1st 2023 for November and December 2023, and January 2024. By choosing an October initialization, we avoid compounding the task of extreme event detection with the lower predictability associated with other initialization windows, such as the boreal spring period in the Eastern Tropical Pacific (Jacox et. al. 2022, [[1]](https://doi.org/10.1038/s41586-022-04573-9)).\n\n- **Anomalies**: Monthly anomalies are calculated by subtracting the calendar month climatology."} {"chunk_id": "seasonal_seasonal-monthly-ocean_extremes-detection_q07__c6cf9b67bf94", "report_id": "seasonal_seasonal-monthly-ocean_extremes-detection_q07", "dataset_id": "seasonal-monthly-ocean", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extremes-detection_q07", "aspect_base": "extremes-detection", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Seasonal Forecast Prediction of Upper Ocean Heat Content Extremes > Quality assessment statement", "title": "Seasonal Forecast Prediction of Upper Ocean Heat Content Extremes", "chunk_index": 2, "token_count": 1080, "text_raw": ". 2022, [[1]](https://doi.org/10.1038/s41586-022-04573-9)).\n\n- **Anomalies**: Monthly anomalies are calculated by subtracting the calendar month climatology.\n\n- We also apply a common land mask to the interpolated products, also omitting areas where at any time the seasonal forecasts has temperatures below 0°C (we could have used a lower threshold than 0°C, but use this threshold as it at least will mask out all ice covered areas).\n\n- The seasonal forecast system (System 9) is initialized by nudging the ocean and sea-ice models towards the Mercator Ocean International GLORYS12V1 reanalysis, except for wihtin 225 km of coastlines or in the presence of sea ice (Specq et. al. 2024, [[3]](https://doi.org/10.5281/zenodo.15518106)). In contrast, ORAS5 is used as the reference reanalysis in this study. This makes the reference data somewhat independent, however, the data assimilated in ORAS5 and GLORYS12 likely overlap to a great extent. Differences in means will not affect the conclusions as anomalies are considered. Further, the limits for a heatwave is calculated separately for the seasonal forecast data and the reference data, allowing them to have different distributions.\n\n[](seasonal_ohc:section-2)\n\n- We compare the data in map plots and in area aggregated time-series plots. Initial inspection showed strong temporal trends in OHC during the time period. The linear trends were removed from the raw data, so that the trends would not inflate the skill scores. Further, the strong temporal trends would skew the diagnosis of heatwaves to primarily occur in the more recent years.\n\n[](seasonal_ohc:section-3)\n\nThe OHC anomalies based on detrended data were used to assess heatwaves. We show that the reference data and seasonal forecasting data have different data distributions. To align our diagnosed heatwaves, the defined percentiles of a heatwave are calculated separately depending on both the data source and the calendar month. To avoid unrealistic jumps, and to align our analysis with Jacox et. al. 2022, [[1]](https://doi.org/10.1038/s41586-022-04573-9) we use a three-month centered rolling window to set the monthly heatwave thresholds. For instance, January thresholds are derived from pooled December, January, and February anomalies. Heatwaves are defined as anomalies exceeding the 90th percentile.\n\nIn the subsequent sections we evaluate the forecast over different dimensions:\n\n### Pooling and grouping strategies\n\n| Analysis Goal\t| Grouping Strategy\t| Question It Answers |\n|---------------|-------------------|---------------------| \n | Skill Maps | Pool time and ensemble for each grid point.\t| \"What is the forecast skill at this specific location over the entire 30-year period?\"|\n| Annual Time Series |\tPooled horizontally and over the ensemble for each year. |\t\"What was the overall global forecast accuracy in the year 1997?\"|\n| Summary Tables |\tPool time, space, and ensemble for each lead-time (month). |\t\"What was the overall global forecast skill for all Novembers combined in the record?\" |\n\n[](seasonal_ohc:section-4)\n\nWe evaluate the forecasting system using the metrics outlined below.\n\n### SEDI - the Symmetric Extremal Dependence Index\n\nSEDI quantifies the forecast skill for rare binary events (e.g., heatwaves). It is the Symmetric Extremal Dependence Index (Ferro & Stephenson 2011 [[4]](https://doi.org/10.1175/WAF-D-10-05030.1)). SEDI has a maximum value of one, and a minimum value of -1. Scores above zero indicate a forecast better than random chance.\n\nSEDI is built on the output of a contingency table. A contingency table provides: sample size, n; base rate, p; hit rate, H; and false-alarm rate, F. The ensemble members are pooled and each adds to the counts in the contingency table before the SEDI score is calculated.\n\n| | Event observed | Non-event observed | |\n | --- | --- | --- | --- | \n |Event forecasted | a = Hpn | b = F(1-p)n | a + b | \n |Non-event forecasted | c = (1-H)pn | d = (1-F)(1-p)n | c + d |\n |Sum | a + c = pn | b + d = (1-p)n | n |\n\nHit rate (H) = $ \\frac{a}{a+c} $, False alarm rate (F) = $ \\frac{b}{b+d} $", "text_with_prefix": "EQC Quality Assessment: \"Seasonal Forecast Prediction of Upper Ocean Heat Content Extremes\"\nDataset: seasonal-monthly-ocean [CDS]\nAspect: extremes-detection_q07 | Category: Seasonal_Forecasts\nSection: Seasonal Forecast Prediction of Upper Ocean Heat Content Extremes > Quality assessment statement\n---\n. 2022, [[1]](https://doi.org/10.1038/s41586-022-04573-9)).\n\n- **Anomalies**: Monthly anomalies are calculated by subtracting the calendar month climatology.\n\n- We also apply a common land mask to the interpolated products, also omitting areas where at any time the seasonal forecasts has temperatures below 0°C (we could have used a lower threshold than 0°C, but use this threshold as it at least will mask out all ice covered areas).\n\n- The seasonal forecast system (System 9) is initialized by nudging the ocean and sea-ice models towards the Mercator Ocean International GLORYS12V1 reanalysis, except for wihtin 225 km of coastlines or in the presence of sea ice (Specq et. al. 2024, [[3]](https://doi.org/10.5281/zenodo.15518106)). In contrast, ORAS5 is used as the reference reanalysis in this study. This makes the reference data somewhat independent, however, the data assimilated in ORAS5 and GLORYS12 likely overlap to a great extent. Differences in means will not affect the conclusions as anomalies are considered. Further, the limits for a heatwave is calculated separately for the seasonal forecast data and the reference data, allowing them to have different distributions.\n\n[](seasonal_ohc:section-2)\n\n- We compare the data in map plots and in area aggregated time-series plots. Initial inspection showed strong temporal trends in OHC during the time period. The linear trends were removed from the raw data, so that the trends would not inflate the skill scores. Further, the strong temporal trends would skew the diagnosis of heatwaves to primarily occur in the more recent years.\n\n[](seasonal_ohc:section-3)\n\nThe OHC anomalies based on detrended data were used to assess heatwaves. We show that the reference data and seasonal forecasting data have different data distributions. To align our diagnosed heatwaves, the defined percentiles of a heatwave are calculated separately depending on both the data source and the calendar month. To avoid unrealistic jumps, and to align our analysis with Jacox et. al. 2022, [[1]](https://doi.org/10.1038/s41586-022-04573-9) we use a three-month centered rolling window to set the monthly heatwave thresholds. For instance, January thresholds are derived from pooled December, January, and February anomalies. Heatwaves are defined as anomalies exceeding the 90th percentile.\n\nIn the subsequent sections we evaluate the forecast over different dimensions:\n\n### Pooling and grouping strategies\n\n| Analysis Goal\t| Grouping Strategy\t| Question It Answers |\n|---------------|-------------------|---------------------| \n | Skill Maps | Pool time and ensemble for each grid point.\t| \"What is the forecast skill at this specific location over the entire 30-year period?\"|\n| Annual Time Series |\tPooled horizontally and over the ensemble for each year. |\t\"What was the overall global forecast accuracy in the year 1997?\"|\n| Summary Tables |\tPool time, space, and ensemble for each lead-time (month). |\t\"What was the overall global forecast skill for all Novembers combined in the record?\" |\n\n[](seasonal_ohc:section-4)\n\nWe evaluate the forecasting system using the metrics outlined below.\n\n### SEDI - the Symmetric Extremal Dependence Index\n\nSEDI quantifies the forecast skill for rare binary events (e.g., heatwaves). It is the Symmetric Extremal Dependence Index (Ferro & Stephenson 2011 [[4]](https://doi.org/10.1175/WAF-D-10-05030.1)). SEDI has a maximum value of one, and a minimum value of -1. Scores above zero indicate a forecast better than random chance.\n\nSEDI is built on the output of a contingency table. A contingency table provides: sample size, n; base rate, p; hit rate, H; and false-alarm rate, F. The ensemble members are pooled and each adds to the counts in the contingency table before the SEDI score is calculated.\n\n| | Event observed | Non-event observed | |\n | --- | --- | --- | --- | \n |Event forecasted | a = Hpn | b = F(1-p)n | a + b | \n |Non-event forecasted | c = (1-H)pn | d = (1-F)(1-p)n | c + d |\n |Sum | a + c = pn | b + d = (1-p)n | n |\n\nHit rate (H) = $ \\frac{a}{a+c} $, False alarm rate (F) = $ \\frac{b}{b+d} $"} {"chunk_id": "seasonal_seasonal-monthly-ocean_extremes-detection_q07__99097840e41b", "report_id": "seasonal_seasonal-monthly-ocean_extremes-detection_q07", "dataset_id": "seasonal-monthly-ocean", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extremes-detection_q07", "aspect_base": "extremes-detection", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Seasonal Forecast Prediction of Upper Ocean Heat Content Extremes > Quality assessment statement", "title": "Seasonal Forecast Prediction of Upper Ocean Heat Content Extremes", "chunk_index": 3, "token_count": 883, "text_raw": "+ d = (1-p)n | n |\n\nHit rate (H) = $ \\frac{a}{a+c} $, False alarm rate (F) = $ \\frac{b}{b+d} $\n\nSEDI = $ \\frac{logF - logH -log(1-F) + log(1-H)}{logF + logH +log(1-F) + log(1-H)} $\n\n### The Brier Skill Score (BSS)\n\nThe Brier Skill Score (BSS) compares the forecast skill against a reference forecast (e.g., climatology), which for heatwaves defined as values above 90% of those observed, would be that there always was about a 10% chance of a heatwave.\n\nIt is calculated as:\n $\n \\text{BSS} = 1 - \\frac{\\text{Brier Score (BrS)}}{\\text{BrS}_{\\text{ref}}}\n $\n\nIt is based on the Brier Score, which measures the mean squared error between forecast probabilities and observed binary events.\n\nThe BSS's upper limit is 1, meanwhile it has no lower limit. Scores above zero indicate a forecast better than the climatology, which is around a 10% chance.\n\n### Accuracy\n\nAccuracy measures the proportion of correct predictions (both true positives and true negatives) out of the total number of forecasts made. The formula is:\n\n$ \\text{Accuracy} = \\frac{\\text{True Positives (TP)} + \\text{True Negatives (TN)}}{\\text{Total Predictions (N)}} $\n\nFor rare events like heatwaves (which occur about 10 % of the time in this study), a simple accuracy score can be misleading. A forecast that always predicts \"no heatwave\" would be 90 % accurate but have zero skill. Therefore, we compare the forecast's accuracy to a random chance benchmark. For an event with a 10% probability (p=0.1), the accuracy of a random forecast is calculated as:\n\n$ \\text{Random Accuracy} = p^2+(1−p)^2 = 0.1^2+(0.9)^2 = 0.82 $\n\nThis means the forecasting system must achieve an accuracy greater than 0.82 (or 82%) to be considered more skillful than random chance.\n\n[](seasonal_ohc:section-5)\nThe contingency tables for all November, December, and January's are presented and discussed.\n\n[](seasonal_ohc:section-6)\nFinally, we display the annual metrics as time-series, also showing the components of the contingency table, allowing a discussion of the components of the metrics.\n\n[](seasonal_ohc:section-7)\n\n## 📈 Analysis and results\n(seasonal_ohc:section-1)= \n### 1. Data Preparation\nThis section includes imports, settings, parameter definitions, and reusable functions required to initialize the analysis.\n\n#### Imports and Settings\n\n### Imports and Settings\n\nSelect realizations for ensemble\nWhether to detrend anomalies or not\nAnalysis time period\nUse all realizations for ensemble\nVarious parameters\nPhysical constants for ocean heat content calculations as used in ORAS5\nConversion factor for heat content (J/m^2)\ndelta_level * cp * rho0 * (Seasonal_data - Tf)\n\n#### Definition of Parameters\n\nWe have chosen Meteo-France System-9, and focus on the hindcast period, which are available from 1993 until 2024, to have a consistent ensemble throughout the analysis.\n\nWe focus on the forecasts initiated October 1st every year, and consider one, two, and three months leadtimes - i.e. the forecasted values for November, December, and January. In the initial processing also October and February values are used, as the heatwave thresholds are centered a three-month month moving window.\n\n#### Helper Functions for Data Processing and Plotting", "text_with_prefix": "EQC Quality Assessment: \"Seasonal Forecast Prediction of Upper Ocean Heat Content Extremes\"\nDataset: seasonal-monthly-ocean [CDS]\nAspect: extremes-detection_q07 | Category: Seasonal_Forecasts\nSection: Seasonal Forecast Prediction of Upper Ocean Heat Content Extremes > Quality assessment statement\n---\n+ d = (1-p)n | n |\n\nHit rate (H) = $ \\frac{a}{a+c} $, False alarm rate (F) = $ \\frac{b}{b+d} $\n\nSEDI = $ \\frac{logF - logH -log(1-F) + log(1-H)}{logF + logH +log(1-F) + log(1-H)} $\n\n### The Brier Skill Score (BSS)\n\nThe Brier Skill Score (BSS) compares the forecast skill against a reference forecast (e.g., climatology), which for heatwaves defined as values above 90% of those observed, would be that there always was about a 10% chance of a heatwave.\n\nIt is calculated as:\n $\n \\text{BSS} = 1 - \\frac{\\text{Brier Score (BrS)}}{\\text{BrS}_{\\text{ref}}}\n $\n\nIt is based on the Brier Score, which measures the mean squared error between forecast probabilities and observed binary events.\n\nThe BSS's upper limit is 1, meanwhile it has no lower limit. Scores above zero indicate a forecast better than the climatology, which is around a 10% chance.\n\n### Accuracy\n\nAccuracy measures the proportion of correct predictions (both true positives and true negatives) out of the total number of forecasts made. The formula is:\n\n$ \\text{Accuracy} = \\frac{\\text{True Positives (TP)} + \\text{True Negatives (TN)}}{\\text{Total Predictions (N)}} $\n\nFor rare events like heatwaves (which occur about 10 % of the time in this study), a simple accuracy score can be misleading. A forecast that always predicts \"no heatwave\" would be 90 % accurate but have zero skill. Therefore, we compare the forecast's accuracy to a random chance benchmark. For an event with a 10% probability (p=0.1), the accuracy of a random forecast is calculated as:\n\n$ \\text{Random Accuracy} = p^2+(1−p)^2 = 0.1^2+(0.9)^2 = 0.82 $\n\nThis means the forecasting system must achieve an accuracy greater than 0.82 (or 82%) to be considered more skillful than random chance.\n\n[](seasonal_ohc:section-5)\nThe contingency tables for all November, December, and January's are presented and discussed.\n\n[](seasonal_ohc:section-6)\nFinally, we display the annual metrics as time-series, also showing the components of the contingency table, allowing a discussion of the components of the metrics.\n\n[](seasonal_ohc:section-7)\n\n## 📈 Analysis and results\n(seasonal_ohc:section-1)= \n### 1. Data Preparation\nThis section includes imports, settings, parameter definitions, and reusable functions required to initialize the analysis.\n\n#### Imports and Settings\n\n### Imports and Settings\n\nSelect realizations for ensemble\nWhether to detrend anomalies or not\nAnalysis time period\nUse all realizations for ensemble\nVarious parameters\nPhysical constants for ocean heat content calculations as used in ORAS5\nConversion factor for heat content (J/m^2)\ndelta_level * cp * rho0 * (Seasonal_data - Tf)\n\n#### Definition of Parameters\n\nWe have chosen Meteo-France System-9, and focus on the hindcast period, which are available from 1993 until 2024, to have a consistent ensemble throughout the analysis.\n\nWe focus on the forecasts initiated October 1st every year, and consider one, two, and three months leadtimes - i.e. the forecasted values for November, December, and January. In the initial processing also October and February values are used, as the heatwave thresholds are centered a three-month month moving window.\n\n#### Helper Functions for Data Processing and Plotting"} {"chunk_id": "seasonal_seasonal-monthly-ocean_extremes-detection_q07__d2bb2e69a205", "report_id": "seasonal_seasonal-monthly-ocean_extremes-detection_q07", "dataset_id": "seasonal-monthly-ocean", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extremes-detection_q07", "aspect_base": "extremes-detection", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Seasonal Forecast Prediction of Upper Ocean Heat Content Extremes > Quality assessment statement", "title": "Seasonal Forecast Prediction of Upper Ocean Heat Content Extremes", "chunk_index": 4, "token_count": 1060, "text_raw": "and January. In the initial processing also October and February values are used, as the heatwave thresholds are centered a three-month month moving window.\n\n#### Helper Functions for Data Processing and Plotting\n\nmap plot helper\nFrom https://github.com/COSIMA/ocean-regrid/blob/master/nemo_grid.py\nThese are the top righ-hand corner of t cells.\nExtend south so that Southern most cells can have bottom corners.\nRepeat first longitude so that Western most cells have left corners.\nCorners of t points. Index 0 is bottom left and then\nanti-clockwise.\nCheck if 'leadtime' is a dim, 'time' is a coord,\nand 'time' is indexed by 'leadtime'.\nSet 'bnds' variables as coordinates\nClean up 'realization' coordinate and ensure it's a dimension\n--- Standard Monthly Climatology ---\nCalculate mean for each month (Jan, Feb, etc.)\nCalculate Anomalies\n--- Month-by-Month Detrending ---\n1. Group by month and fit trend using .map()\nExtract the coefficients DataArray.\nReconstruct the full trend line (mx + c)\nIsolate the slope component.\nSubtract the slope\nprint(f\"Number of 'True' points in sesonal static_sea_ice_proxy_mask: {static_not_sea_ice_proxy_mask_s.sum().compute().item()}\")\nstatic_not_sea_ice_proxy_mask_s.plot()\nStep 1: Compute anomalies\napply the mask here\n--------------------------\nStep 2: Reindex to (year, month), assing year to fc_start\nStep 1: Compute anomalies\n--------------------------\nCreate the 'season_year'\nthe month coord will be [1, 2, 3, 4, 10, 11, 12]\n\n#### Data Loading\nWe download the reanalysis and seasonal forecast datasets, regridding the reanalysis data to align with the seasonal forecast grid, and computing the detrended anomalies. \nThese transformations are performed using the *c3s_eqc_automatic_quality_control* *download.download_and_transform* function and custom preprocessing steps. xesmf conservative regridding was used to transform the ORAS5 ORCA025 grid to a global 1x1 latlon grid. Sensitivity tests were conducted to ensure that the regridding process (comparing area-conservative interpolation used for ORAS5 vs. a version of the two step distance-weighted interpolation that System 9 underwent) produced similar changes to the mean and variance of the original data (not shown). \n - The data is combined and reindexed so that year and calendar month are the dataset coordinates.\n - The long-term climatology and possibly the monthly trend is subtracted.\n - We apply a common land mask and also mask out areas where seasonal forecast data have values corresponding to a monthly mean potential temperature below 0°C.\n - The seasonal forecast vertically integrated potential temperature is converted to OHC, following the convention used in ORAS5. \n - The variable names are changed to reflect their current state.\n\nrequests_reanalysis = {leadtime: [] for leadtime in leadtimes}\npost processing: conversion and applying the common mask\n(the seas fc has an additional mask now, removing cold areas,\ncomputed before converted to anomalies )\nconvert the seasonal values to potential temperature expressed as OHC\nreaname the variables\nadd inn the common mask again, also using an ice-proxy from the seasonal fc data\n\n```text\nBuilding a flat list of unique reanalysis requests...\nFound 155 unique months of reanalysis to download.\n```\n\n(seasonal_ohc:section-2)= \n### 2. OHC Time Series and Trend Analysis\n\nCompares the model's OHC and heatwave representation against the reanalysis. The time series of globally averaged OHC anomalies clearly shows a strong positive trend over the 1993-2023 period ([](ffig-anom-ts)). This warming trend is present in both the ORAS5 reanalysis and the Météo-France System 9 seasonal forecasts, though it is slightly stornger in Météo-France System 9. Because this predictable, long-term signal could artificially inflate skill scores, the linear trend is removed from both datasets for all subsequent heatwave and skill analyses. The detrended time series shows the interannual variability, such as the strong El Niño events, more clearly.\n\nTime series plotting\nSort the data\nPerform the mean operations (using ensemble mean for forecast)\n--- Compute Monthly Correlations ---\nWe take the ensemble mean of the forecast first, then correlate with reanalysis per month\nThis produces a correlation value for each month [11, 12, 1]\nFormat the correlation strings for the title\n--- Plotting ---\nConvert to DataFrames for seaborn\nPanel 0: Global\nPanel 1: Nino 3.4\n--- Execute plotting ---", "text_with_prefix": "EQC Quality Assessment: \"Seasonal Forecast Prediction of Upper Ocean Heat Content Extremes\"\nDataset: seasonal-monthly-ocean [CDS]\nAspect: extremes-detection_q07 | Category: Seasonal_Forecasts\nSection: Seasonal Forecast Prediction of Upper Ocean Heat Content Extremes > Quality assessment statement\n---\nand January. In the initial processing also October and February values are used, as the heatwave thresholds are centered a three-month month moving window.\n\n#### Helper Functions for Data Processing and Plotting\n\nmap plot helper\nFrom https://github.com/COSIMA/ocean-regrid/blob/master/nemo_grid.py\nThese are the top righ-hand corner of t cells.\nExtend south so that Southern most cells can have bottom corners.\nRepeat first longitude so that Western most cells have left corners.\nCorners of t points. Index 0 is bottom left and then\nanti-clockwise.\nCheck if 'leadtime' is a dim, 'time' is a coord,\nand 'time' is indexed by 'leadtime'.\nSet 'bnds' variables as coordinates\nClean up 'realization' coordinate and ensure it's a dimension\n--- Standard Monthly Climatology ---\nCalculate mean for each month (Jan, Feb, etc.)\nCalculate Anomalies\n--- Month-by-Month Detrending ---\n1. Group by month and fit trend using .map()\nExtract the coefficients DataArray.\nReconstruct the full trend line (mx + c)\nIsolate the slope component.\nSubtract the slope\nprint(f\"Number of 'True' points in sesonal static_sea_ice_proxy_mask: {static_not_sea_ice_proxy_mask_s.sum().compute().item()}\")\nstatic_not_sea_ice_proxy_mask_s.plot()\nStep 1: Compute anomalies\napply the mask here\n--------------------------\nStep 2: Reindex to (year, month), assing year to fc_start\nStep 1: Compute anomalies\n--------------------------\nCreate the 'season_year'\nthe month coord will be [1, 2, 3, 4, 10, 11, 12]\n\n#### Data Loading\nWe download the reanalysis and seasonal forecast datasets, regridding the reanalysis data to align with the seasonal forecast grid, and computing the detrended anomalies. \nThese transformations are performed using the *c3s_eqc_automatic_quality_control* *download.download_and_transform* function and custom preprocessing steps. xesmf conservative regridding was used to transform the ORAS5 ORCA025 grid to a global 1x1 latlon grid. Sensitivity tests were conducted to ensure that the regridding process (comparing area-conservative interpolation used for ORAS5 vs. a version of the two step distance-weighted interpolation that System 9 underwent) produced similar changes to the mean and variance of the original data (not shown). \n - The data is combined and reindexed so that year and calendar month are the dataset coordinates.\n - The long-term climatology and possibly the monthly trend is subtracted.\n - We apply a common land mask and also mask out areas where seasonal forecast data have values corresponding to a monthly mean potential temperature below 0°C.\n - The seasonal forecast vertically integrated potential temperature is converted to OHC, following the convention used in ORAS5. \n - The variable names are changed to reflect their current state.\n\nrequests_reanalysis = {leadtime: [] for leadtime in leadtimes}\npost processing: conversion and applying the common mask\n(the seas fc has an additional mask now, removing cold areas,\ncomputed before converted to anomalies )\nconvert the seasonal values to potential temperature expressed as OHC\nreaname the variables\nadd inn the common mask again, also using an ice-proxy from the seasonal fc data\n\n```text\nBuilding a flat list of unique reanalysis requests...\nFound 155 unique months of reanalysis to download.\n```\n\n(seasonal_ohc:section-2)= \n### 2. OHC Time Series and Trend Analysis\n\nCompares the model's OHC and heatwave representation against the reanalysis. The time series of globally averaged OHC anomalies clearly shows a strong positive trend over the 1993-2023 period ([](ffig-anom-ts)). This warming trend is present in both the ORAS5 reanalysis and the Météo-France System 9 seasonal forecasts, though it is slightly stornger in Météo-France System 9. Because this predictable, long-term signal could artificially inflate skill scores, the linear trend is removed from both datasets for all subsequent heatwave and skill analyses. The detrended time series shows the interannual variability, such as the strong El Niño events, more clearly.\n\nTime series plotting\nSort the data\nPerform the mean operations (using ensemble mean for forecast)\n--- Compute Monthly Correlations ---\nWe take the ensemble mean of the forecast first, then correlate with reanalysis per month\nThis produces a correlation value for each month [11, 12, 1]\nFormat the correlation strings for the title\n--- Plotting ---\nConvert to DataFrames for seaborn\nPanel 0: Global\nPanel 1: Nino 3.4\n--- Execute plotting ---"} {"chunk_id": "seasonal_seasonal-monthly-ocean_extremes-detection_q07__cfef2ef4667c", "report_id": "seasonal_seasonal-monthly-ocean_extremes-detection_q07", "dataset_id": "seasonal-monthly-ocean", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extremes-detection_q07", "aspect_base": "extremes-detection", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Seasonal Forecast Prediction of Upper Ocean Heat Content Extremes > Quality assessment statement", "title": "Seasonal Forecast Prediction of Upper Ocean Heat Content Extremes", "chunk_index": 5, "token_count": 1046, "text_raw": ", 12, 1]\nFormat the correlation strings for the title\n--- Plotting ---\nConvert to DataFrames for seaborn\nPanel 0: Global\nPanel 1: Nino 3.4\n--- Execute plotting ---\n\n(ffig-anom-ts)=\n#### Figure 2 \nAnnual time series of reanalysis and forecast anomalies according to calendar month. The forecast fields have a nominal start date October 1st. The single dots with a stronger color shows the reanalysis data, while the more lightly colored plot clouds show the seasonal forecasts (one dot per realization). The upper two rows show the anomalies in the original time-series for the global and Niño3.4 area, while the lower panels show the detrended data. The subplot-titles provide the Pearson correlation coefficients of the area averaged anomalies. Note that the y-axis range varies.\n\n---\n\n(seasonal_ohc:section-3)= \n### 3. Model Performance vs. Reanalysis\n\nThis section directly compares the model's OHC and heatwave representation against the reanalysis, after having detrended both.\n\n#### Correlation of OHC Anomalies\n\n--- 1. Compute monthly correlations ---\nCompute Pearson correlation between reanalysis and forecast ensemble mean per calendar month\n--- 2. Visualization (3-panel plot) ---\nExtract correlation map for the specific lead month\nPlot correlation\nMap configuration\nAdd shared horizontal colorbar at the bottom\nClean up memory\n\n```text\n/data/common/miniforge3/envs/wp3/lib/python3.12/site-packages/dask/array/numpy_compat.py:58: RuntimeWarning: invalid value encountered in divide\n x = np.divide(x1, x2, out)\n```\n\n(fig-corr-map)=\n##### Figure 3\nMap plot showing the temporal anomaly correlation coefficients (Pearson's) between the forecast ensemble mean and the reanalysis 1993-2023 NDJ OHC every valid grid box.\n\n---\n\n#### Comparison of distributions \nWe inspect the OHC anomalies in the two data sources to see if they share a similar distribution. We look at November, December, and January simultaneously and pool the seasonal forecast ensemble members. Identical distributions and tails would allow for the use of the same physical magnitudes for a heatwave threshold.\n\nThe differences in the probability density functions, the longer tails in the reanalysis versus the more blunt tails in the forecasts justify the use of relative heatwave thresholds depending on the specific data source.\n\n1. Data Selection (NDJ) and Masking\nSelecting late autumn/winter months and applying the common ocean mask\n2. Spatial Subsampling\nCompute early to bring data into memory for weighted statistics and quantiles\n3. Area Weights Calculation\nUsing cosine of latitude to account for grid cell area convergence at poles\nBroadcast weights to match the 2D/3D shape of the data for flattening\n4. Compute Weighted Statistics\nCalculate weighted quantiles for the Q-Q plot (focusing on the extreme upper tail)\n5. Prepare Data for PDF/Kurtosis (flattening and dropping NaNs)\nCalculate Fisher Kurtosis (Normal distribution = 0)\nPositive = Leptokurtic (fat tails), Negative = Platykurtic (thin tails)\nKolmogorov-Smirnov test (unweighted) to check if distributions are identical\nCalculate the number of ensemble members\nCalculate points per realization for the forecast\nPrinting Results Table\n7. Visualization\nLeft: Area-Weighted Probability Density Function (KDE)\nRight: Q-Q Plot for the Upper Tail Comparison\nExecution call\n\n```text\nStarting statistical analysis (subsampling step: 5)...\nComputing weighted statistics...\n\n--- Statistics Summary ---\n Metric Forecast (S9) Reanalysis (ORAS5)\n StdDev (J/m²) 6.40e+08 6.71e+08\n Kurtosis 4.18 6.50\n Total Points 3687357 118947\nPoints per Member 118947 118947\nK-S Test p-value: 1.73e-47\n```\n\n(fig-pdfs)=\n##### Figure 4\nDistribution analysis of pooled OHC anomalies for the NDJ season. The left plot shows the area-weighted probability density functions (PDFs) for the forecast (blue) and reanalysis (red), where the broader \"shoulders\" of the blue curve indicate a blunter distribution. The right plot shows a quantile-quantile comparison of the upper tail ($P_{85}$ to $P_{99}$) of the two data sources. The black line falling below the diagonal indicate higher intensity values (longer tails) in the reanalysis data compared to the forecast.\n\n---", "text_with_prefix": "EQC Quality Assessment: \"Seasonal Forecast Prediction of Upper Ocean Heat Content Extremes\"\nDataset: seasonal-monthly-ocean [CDS]\nAspect: extremes-detection_q07 | Category: Seasonal_Forecasts\nSection: Seasonal Forecast Prediction of Upper Ocean Heat Content Extremes > Quality assessment statement\n---\n, 12, 1]\nFormat the correlation strings for the title\n--- Plotting ---\nConvert to DataFrames for seaborn\nPanel 0: Global\nPanel 1: Nino 3.4\n--- Execute plotting ---\n\n(ffig-anom-ts)=\n#### Figure 2 \nAnnual time series of reanalysis and forecast anomalies according to calendar month. The forecast fields have a nominal start date October 1st. The single dots with a stronger color shows the reanalysis data, while the more lightly colored plot clouds show the seasonal forecasts (one dot per realization). The upper two rows show the anomalies in the original time-series for the global and Niño3.4 area, while the lower panels show the detrended data. The subplot-titles provide the Pearson correlation coefficients of the area averaged anomalies. Note that the y-axis range varies.\n\n---\n\n(seasonal_ohc:section-3)= \n### 3. Model Performance vs. Reanalysis\n\nThis section directly compares the model's OHC and heatwave representation against the reanalysis, after having detrended both.\n\n#### Correlation of OHC Anomalies\n\n--- 1. Compute monthly correlations ---\nCompute Pearson correlation between reanalysis and forecast ensemble mean per calendar month\n--- 2. Visualization (3-panel plot) ---\nExtract correlation map for the specific lead month\nPlot correlation\nMap configuration\nAdd shared horizontal colorbar at the bottom\nClean up memory\n\n```text\n/data/common/miniforge3/envs/wp3/lib/python3.12/site-packages/dask/array/numpy_compat.py:58: RuntimeWarning: invalid value encountered in divide\n x = np.divide(x1, x2, out)\n```\n\n(fig-corr-map)=\n##### Figure 3\nMap plot showing the temporal anomaly correlation coefficients (Pearson's) between the forecast ensemble mean and the reanalysis 1993-2023 NDJ OHC every valid grid box.\n\n---\n\n#### Comparison of distributions \nWe inspect the OHC anomalies in the two data sources to see if they share a similar distribution. We look at November, December, and January simultaneously and pool the seasonal forecast ensemble members. Identical distributions and tails would allow for the use of the same physical magnitudes for a heatwave threshold.\n\nThe differences in the probability density functions, the longer tails in the reanalysis versus the more blunt tails in the forecasts justify the use of relative heatwave thresholds depending on the specific data source.\n\n1. Data Selection (NDJ) and Masking\nSelecting late autumn/winter months and applying the common ocean mask\n2. Spatial Subsampling\nCompute early to bring data into memory for weighted statistics and quantiles\n3. Area Weights Calculation\nUsing cosine of latitude to account for grid cell area convergence at poles\nBroadcast weights to match the 2D/3D shape of the data for flattening\n4. Compute Weighted Statistics\nCalculate weighted quantiles for the Q-Q plot (focusing on the extreme upper tail)\n5. Prepare Data for PDF/Kurtosis (flattening and dropping NaNs)\nCalculate Fisher Kurtosis (Normal distribution = 0)\nPositive = Leptokurtic (fat tails), Negative = Platykurtic (thin tails)\nKolmogorov-Smirnov test (unweighted) to check if distributions are identical\nCalculate the number of ensemble members\nCalculate points per realization for the forecast\nPrinting Results Table\n7. Visualization\nLeft: Area-Weighted Probability Density Function (KDE)\nRight: Q-Q Plot for the Upper Tail Comparison\nExecution call\n\n```text\nStarting statistical analysis (subsampling step: 5)...\nComputing weighted statistics...\n\n--- Statistics Summary ---\n Metric Forecast (S9) Reanalysis (ORAS5)\n StdDev (J/m²) 6.40e+08 6.71e+08\n Kurtosis 4.18 6.50\n Total Points 3687357 118947\nPoints per Member 118947 118947\nK-S Test p-value: 1.73e-47\n```\n\n(fig-pdfs)=\n##### Figure 4\nDistribution analysis of pooled OHC anomalies for the NDJ season. The left plot shows the area-weighted probability density functions (PDFs) for the forecast (blue) and reanalysis (red), where the broader \"shoulders\" of the blue curve indicate a blunter distribution. The right plot shows a quantile-quantile comparison of the upper tail ($P_{85}$ to $P_{99}$) of the two data sources. The black line falling below the diagonal indicate higher intensity values (longer tails) in the reanalysis data compared to the forecast.\n\n---"} {"chunk_id": "seasonal_seasonal-monthly-ocean_extremes-detection_q07__53a30a9b03a4", "report_id": "seasonal_seasonal-monthly-ocean_extremes-detection_q07", "dataset_id": "seasonal-monthly-ocean", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extremes-detection_q07", "aspect_base": "extremes-detection", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Seasonal Forecast Prediction of Upper Ocean Heat Content Extremes > Quality assessment statement", "title": "Seasonal Forecast Prediction of Upper Ocean Heat Content Extremes", "chunk_index": 6, "token_count": 1050, "text_raw": "tail ($P_{85}$ to $P_{99}$) of the two data sources. The black line falling below the diagonal indicate higher intensity values (longer tails) in the reanalysis data compared to the forecast.\n\n---\n\n#### Comparison of Heatwave Representation\n**Heatwaves** are defined as monthly anomalies exceeding a 90th percentile threshold. The thresholds are calculated from a three-month centered moving window, and separately for each data source. For reanalysis data, the threshold is calculated across all years at each grid point. For the seasonal forecast, the threshold is calculated across all years and all ensemble members at each grid point.\n\n--- 1. Data Preparation ---\n--- 2. Threshold Calculation (Centered Windows) ---\nReanalysis thresholds (ORAS5 reference)\nSeasonal Forecast thresholds (System 9)\nCompute p90 values\n--- 3. Unified Scaling and Plotting (2x3 Grid) ---\nUnified color limits for all 6 panels to ensure fair comparison\nRow 0: Reanalysis Reference (ORAS5)\nRow 1: Forecast Thresholds (System 9)\nShared colorbar for all subplots\n--- 4. Mask Generation ---\n--- Call the function ---\n\n```text\nComputing p90 thresholds...\n```\n\n```text\n/data/common/miniforge3/envs/wp3/lib/python3.12/site-packages/numpy/lib/_nanfunctions_impl.py:1617: RuntimeWarning: All-NaN slice encountered\n return fnb._ureduce(a,\n```\n\n```text\nThresholds computed.\nGenerating heatwave masks...\n```\n\n##### Figure 5\nMap plot showing the definition of heatwave (or the grid cell threshold to be exceeded) in the OHC anomalies for the seasonal forecasts (left) and the reanalysis (right).\n\n---\n\n#### Diagnostic Plots.\n\nWe choose 1997 as our year to inspect the spatial patterns of the anomalies.\n\n--- 1. Selection & Computation ---\nBring small spatial slices into memory\n--- 2. Figure Setup ---\nConsistent scale for rows 1 & 2\nSeparate scale for the difference row\n--- 3. Plotting Grid ---\nRow 1: Forecast (Météo-France System 9)\nRow 2: Reanalysis (ORAS5)\nRow 3: Difference (Bias)\nStyling and horizontal colorbars\nPositioning colorbars at the bottom\nRun the plot\n\n##### Figure 6 \nMap plots showing ensemble mean OHC anomalies from the seasonal forecast with a nomianal start date 1 October 1997 (upper row), the corresponding reanalysis fields (centre row), and their difference (forecast minus reanalysis, bottom row).\n\n---\n\n(seasonal_ohc:section-4)=\n### 4. Metrics: SEDI, BSS, Accuracy\n\nThis section presents the primary skill scores (SEDI, BSS, Accuracy).\n\n#### Symmetric Extremal Dependence Index (SEDI)\n\nSEDI (Ferro & Stephenson 2011, [[4]](https://doi.org/10.1175/WAF-D-10-05030.1)) is calculated using the hit rate (H) and false alarm rate (F), as can be found in a contingency table:\n\n$\n \\text{SEDI} = \\frac{\\log(F) - \\log(H) - \\log(1-F) + \\log(1-H)}{\\log(F) + \\log(H) + \\log(1-F) + \\log(1-H)}\n $\n\nTo avoid numerical issues, logarithms are computed using a lower bound of $10^{-6}$.\n\n#### The Brier Skill Score (BSS)\n\nThe Brier Skill Score (BSS) here compares the forecast's Brier score against the Brier score of using the climatology of the reference data set as a forecast. The climatologically based heatwave forecast is near a 10% chance, but is here calculated explicitly per grid point. The BSS’s upper limit is 1, meanwhile it has no lower limit. Scores above zero indicate a forecast better than climatology.\n\n#### Accuracy Assessment\nProportion of correct predictions relative to total events. Forecast accuracy was calculated as:\n $\n \\text{Accuracy} = \\frac{\\text{True Positives} + \\text{True Negatives}}{N}\n $\n where $N$ is the total number of events. Accuracy above a random threshold (e.g., 0.82 for a base rate of 0.1) indicates skill better than chance.\n\n#### Pooling Strategy for Skill Metrics in Map Form\n\nTo calculate robust skill scores, data is pooled in different ways depending on the metric. The BSS pools the ensemble members at once, while this is done in a later step for SEDI and Accuracy:", "text_with_prefix": "EQC Quality Assessment: \"Seasonal Forecast Prediction of Upper Ocean Heat Content Extremes\"\nDataset: seasonal-monthly-ocean [CDS]\nAspect: extremes-detection_q07 | Category: Seasonal_Forecasts\nSection: Seasonal Forecast Prediction of Upper Ocean Heat Content Extremes > Quality assessment statement\n---\ntail ($P_{85}$ to $P_{99}$) of the two data sources. The black line falling below the diagonal indicate higher intensity values (longer tails) in the reanalysis data compared to the forecast.\n\n---\n\n#### Comparison of Heatwave Representation\n**Heatwaves** are defined as monthly anomalies exceeding a 90th percentile threshold. The thresholds are calculated from a three-month centered moving window, and separately for each data source. For reanalysis data, the threshold is calculated across all years at each grid point. For the seasonal forecast, the threshold is calculated across all years and all ensemble members at each grid point.\n\n--- 1. Data Preparation ---\n--- 2. Threshold Calculation (Centered Windows) ---\nReanalysis thresholds (ORAS5 reference)\nSeasonal Forecast thresholds (System 9)\nCompute p90 values\n--- 3. Unified Scaling and Plotting (2x3 Grid) ---\nUnified color limits for all 6 panels to ensure fair comparison\nRow 0: Reanalysis Reference (ORAS5)\nRow 1: Forecast Thresholds (System 9)\nShared colorbar for all subplots\n--- 4. Mask Generation ---\n--- Call the function ---\n\n```text\nComputing p90 thresholds...\n```\n\n```text\n/data/common/miniforge3/envs/wp3/lib/python3.12/site-packages/numpy/lib/_nanfunctions_impl.py:1617: RuntimeWarning: All-NaN slice encountered\n return fnb._ureduce(a,\n```\n\n```text\nThresholds computed.\nGenerating heatwave masks...\n```\n\n##### Figure 5\nMap plot showing the definition of heatwave (or the grid cell threshold to be exceeded) in the OHC anomalies for the seasonal forecasts (left) and the reanalysis (right).\n\n---\n\n#### Diagnostic Plots.\n\nWe choose 1997 as our year to inspect the spatial patterns of the anomalies.\n\n--- 1. Selection & Computation ---\nBring small spatial slices into memory\n--- 2. Figure Setup ---\nConsistent scale for rows 1 & 2\nSeparate scale for the difference row\n--- 3. Plotting Grid ---\nRow 1: Forecast (Météo-France System 9)\nRow 2: Reanalysis (ORAS5)\nRow 3: Difference (Bias)\nStyling and horizontal colorbars\nPositioning colorbars at the bottom\nRun the plot\n\n##### Figure 6 \nMap plots showing ensemble mean OHC anomalies from the seasonal forecast with a nomianal start date 1 October 1997 (upper row), the corresponding reanalysis fields (centre row), and their difference (forecast minus reanalysis, bottom row).\n\n---\n\n(seasonal_ohc:section-4)=\n### 4. Metrics: SEDI, BSS, Accuracy\n\nThis section presents the primary skill scores (SEDI, BSS, Accuracy).\n\n#### Symmetric Extremal Dependence Index (SEDI)\n\nSEDI (Ferro & Stephenson 2011, [[4]](https://doi.org/10.1175/WAF-D-10-05030.1)) is calculated using the hit rate (H) and false alarm rate (F), as can be found in a contingency table:\n\n$\n \\text{SEDI} = \\frac{\\log(F) - \\log(H) - \\log(1-F) + \\log(1-H)}{\\log(F) + \\log(H) + \\log(1-F) + \\log(1-H)}\n $\n\nTo avoid numerical issues, logarithms are computed using a lower bound of $10^{-6}$.\n\n#### The Brier Skill Score (BSS)\n\nThe Brier Skill Score (BSS) here compares the forecast's Brier score against the Brier score of using the climatology of the reference data set as a forecast. The climatologically based heatwave forecast is near a 10% chance, but is here calculated explicitly per grid point. The BSS’s upper limit is 1, meanwhile it has no lower limit. Scores above zero indicate a forecast better than climatology.\n\n#### Accuracy Assessment\nProportion of correct predictions relative to total events. Forecast accuracy was calculated as:\n $\n \\text{Accuracy} = \\frac{\\text{True Positives} + \\text{True Negatives}}{N}\n $\n where $N$ is the total number of events. Accuracy above a random threshold (e.g., 0.82 for a base rate of 0.1) indicates skill better than chance.\n\n#### Pooling Strategy for Skill Metrics in Map Form\n\nTo calculate robust skill scores, data is pooled in different ways depending on the metric. The BSS pools the ensemble members at once, while this is done in a later step for SEDI and Accuracy:"} {"chunk_id": "seasonal_seasonal-monthly-ocean_extremes-detection_q07__ae8dcdcf75b4", "report_id": "seasonal_seasonal-monthly-ocean_extremes-detection_q07", "dataset_id": "seasonal-monthly-ocean", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extremes-detection_q07", "aspect_base": "extremes-detection", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Seasonal Forecast Prediction of Upper Ocean Heat Content Extremes > Quality assessment statement", "title": "Seasonal Forecast Prediction of Upper Ocean Heat Content Extremes", "chunk_index": 7, "token_count": 1093, "text_raw": "Metrics in Map Form\n\nTo calculate robust skill scores, data is pooled in different ways depending on the metric. The BSS pools the ensemble members at once, while this is done in a later step for SEDI and Accuracy:\n\n* **For SEDI and Accuracy**: These metrics are calculated from a contingency table (i.e., hits, misses, false alarms, correct rejections). For these every **ensemble member** in every **year** is treated as an independent event. This means we are **pooling across the time dimension (1993-2023) and the ensemble dimension**. This results in a single, robust skill score map that represents the overall performance for that calendar month across the entire period.\n\n* **For the Brier Skill Score (BSS)**: This metric evaluates the performance of a probabilistic forecast compared to a reference forecast (here climatology) for each calendar month and grid cell. The calculation is done in steps:\n 1. First, for each year, the ensemble members are pooled to create a single **forecast probability** for that year (e.g., if 5 of 25 members predict a heatwave, the probability is 0.2).\n 2. The Brier Score is then calculated for each year by comparing that probability to the single observed outcome (`1` or `0`).\n 3. Finally, these annual Brier Scores are averaged across all years to get the final score. The BSS then compares this to the reference score from climatology.\n\n#### Findings\nA key finding of this analysis is a divergence between the metrics in several regions ([](fig-global-maps)). The SEDI and Accuracy scores ( > 0.82) indicate that the forecast is better than a reference climatology, while BSS is weakly negative in many regions. A negative BSS indicates that the forecast is less skillful than the long-term climatological average, i.e. assuming a ~10% chance of a heatwave everywhere, every year.\n\nThis suggests that while the model has the ability to correctly identify many heatwave events (demonstrated in the positive SEDI), it likely forecasts with too much confidence or has a high \"false alarm rate,\" where it predicts heatwaves that don't occur. The BSS penalizes this lack of reliability, while SEDI, which focuses only on the co-occurrence of extreme events, is less sensitive to it.\n\nThe highest skill (SEDI > 0.5) is consistently found in the tropical Pacific Ocean. This region's variability is dominated by the El Niño-Southern Oscillation (ENSO), the primary source of seasonal predictability in the climate system. The model successfully captures the OHC anomalies associated with strong ENSO events.\nThe skill is lower in the mid-to-high latitudes. These regions are governed by more chaotic and faster-moving atmospheric and oceanic processes, making them more difficult to predict months in advance.\n\nFunction to compute skill score maps\n--- Step 1: Brier Skill Score ---\nfloat32 conversion\nCalculate probabilities\nCalculate Brier Scores\nCalculate final BSS\n--- Execute the BSS calculation ---\n--- Free memory ---\n--- Step 2: SEDI and Accuracy ---\nlazy and uses int8\nSum over 'year' and 'realization' to create the maps\n--- Execute the SEDI/Accuracy calculation ---\n--- Calculate final metrics---\n--- Combine ---\n--- Execute the skill calculation ---\n\nThe main plotting cell with mean values in titles\ndictionaries\nDefine the labels you want for each row\n-------------------------\n\n(fig-global-maps)=\n##### Figure 7 \nSkill of Météo-France System-9 forecasts with a nominal start date October 1st for November, December, and January, using the ORAS5 reanalysis as the reference dataset. Heatwaves are defined as upper 300m ocean heat content (OHC) anomalies exceeding the 90th percentile within the 1993-2023 period. Rows show: (A) Symmetric Extremal Dependence Index (SEDI), (B) Brier Skill Score (BSS); and (C) Accuracy. In all maps, red areas indicate positive skill (forecast is better than the reference), while blue areas indicate negative skill. Masked areas are shown in grey. For the Accuracy maps (C) the areas with an accuracy lower than what would be expected by random chance (82%) are also shown with a grey color.\n\n---\n\n(seasonal_ohc:section-5)= \n### 5. Contingency Table Analysis\nWe examine the spatially-pooled contingency tables for each month to better understand the forecast behavior.\n\nCreate boolean fields.\nDefine dimensions to sum over\nExecute all calculations in a single Dask operation\nprint(\"Computing contingency table\")\nFormat to 2 decimal places with a '%' sign\n\n```text\n--- Contingency Tables in Percent (%) ---\n\nNovember:\n```\n\n```text\nDecember:\n```\n\n```text\nJanuary:\n```", "text_with_prefix": "EQC Quality Assessment: \"Seasonal Forecast Prediction of Upper Ocean Heat Content Extremes\"\nDataset: seasonal-monthly-ocean [CDS]\nAspect: extremes-detection_q07 | Category: Seasonal_Forecasts\nSection: Seasonal Forecast Prediction of Upper Ocean Heat Content Extremes > Quality assessment statement\n---\nMetrics in Map Form\n\nTo calculate robust skill scores, data is pooled in different ways depending on the metric. The BSS pools the ensemble members at once, while this is done in a later step for SEDI and Accuracy:\n\n* **For SEDI and Accuracy**: These metrics are calculated from a contingency table (i.e., hits, misses, false alarms, correct rejections). For these every **ensemble member** in every **year** is treated as an independent event. This means we are **pooling across the time dimension (1993-2023) and the ensemble dimension**. This results in a single, robust skill score map that represents the overall performance for that calendar month across the entire period.\n\n* **For the Brier Skill Score (BSS)**: This metric evaluates the performance of a probabilistic forecast compared to a reference forecast (here climatology) for each calendar month and grid cell. The calculation is done in steps:\n 1. First, for each year, the ensemble members are pooled to create a single **forecast probability** for that year (e.g., if 5 of 25 members predict a heatwave, the probability is 0.2).\n 2. The Brier Score is then calculated for each year by comparing that probability to the single observed outcome (`1` or `0`).\n 3. Finally, these annual Brier Scores are averaged across all years to get the final score. The BSS then compares this to the reference score from climatology.\n\n#### Findings\nA key finding of this analysis is a divergence between the metrics in several regions ([](fig-global-maps)). The SEDI and Accuracy scores ( > 0.82) indicate that the forecast is better than a reference climatology, while BSS is weakly negative in many regions. A negative BSS indicates that the forecast is less skillful than the long-term climatological average, i.e. assuming a ~10% chance of a heatwave everywhere, every year.\n\nThis suggests that while the model has the ability to correctly identify many heatwave events (demonstrated in the positive SEDI), it likely forecasts with too much confidence or has a high \"false alarm rate,\" where it predicts heatwaves that don't occur. The BSS penalizes this lack of reliability, while SEDI, which focuses only on the co-occurrence of extreme events, is less sensitive to it.\n\nThe highest skill (SEDI > 0.5) is consistently found in the tropical Pacific Ocean. This region's variability is dominated by the El Niño-Southern Oscillation (ENSO), the primary source of seasonal predictability in the climate system. The model successfully captures the OHC anomalies associated with strong ENSO events.\nThe skill is lower in the mid-to-high latitudes. These regions are governed by more chaotic and faster-moving atmospheric and oceanic processes, making them more difficult to predict months in advance.\n\nFunction to compute skill score maps\n--- Step 1: Brier Skill Score ---\nfloat32 conversion\nCalculate probabilities\nCalculate Brier Scores\nCalculate final BSS\n--- Execute the BSS calculation ---\n--- Free memory ---\n--- Step 2: SEDI and Accuracy ---\nlazy and uses int8\nSum over 'year' and 'realization' to create the maps\n--- Execute the SEDI/Accuracy calculation ---\n--- Calculate final metrics---\n--- Combine ---\n--- Execute the skill calculation ---\n\nThe main plotting cell with mean values in titles\ndictionaries\nDefine the labels you want for each row\n-------------------------\n\n(fig-global-maps)=\n##### Figure 7 \nSkill of Météo-France System-9 forecasts with a nominal start date October 1st for November, December, and January, using the ORAS5 reanalysis as the reference dataset. Heatwaves are defined as upper 300m ocean heat content (OHC) anomalies exceeding the 90th percentile within the 1993-2023 period. Rows show: (A) Symmetric Extremal Dependence Index (SEDI), (B) Brier Skill Score (BSS); and (C) Accuracy. In all maps, red areas indicate positive skill (forecast is better than the reference), while blue areas indicate negative skill. Masked areas are shown in grey. For the Accuracy maps (C) the areas with an accuracy lower than what would be expected by random chance (82%) are also shown with a grey color.\n\n---\n\n(seasonal_ohc:section-5)= \n### 5. Contingency Table Analysis\nWe examine the spatially-pooled contingency tables for each month to better understand the forecast behavior.\n\nCreate boolean fields.\nDefine dimensions to sum over\nExecute all calculations in a single Dask operation\nprint(\"Computing contingency table\")\nFormat to 2 decimal places with a '%' sign\n\n```text\n--- Contingency Tables in Percent (%) ---\n\nNovember:\n```\n\n```text\nDecember:\n```\n\n```text\nJanuary:\n```"} {"chunk_id": "seasonal_seasonal-monthly-ocean_extremes-detection_q07__12827b674bcd", "report_id": "seasonal_seasonal-monthly-ocean_extremes-detection_q07", "dataset_id": "seasonal-monthly-ocean", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extremes-detection_q07", "aspect_base": "extremes-detection", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Seasonal Forecast Prediction of Upper Ocean Heat Content Extremes > Quality assessment statement", "title": "Seasonal Forecast Prediction of Upper Ocean Heat Content Extremes", "chunk_index": 8, "token_count": 1054, "text_raw": "print(\"Computing contingency table\")\nFormat to 2 decimal places with a '%' sign\n\n```text\n--- Contingency Tables in Percent (%) ---\n\nNovember:\n```\n\n```text\nDecember:\n```\n\n```text\nJanuary:\n```\n\nFor all Novembers (one month leadtime, or 1-2 months, to be exact), when the model predicts a marine heatwave, that prediction is correct 32% of the time (Precision). Conversely, the False Discovery Rate is 68%, meaning roughly two-thirds of forecasted heatwaves did not materialize in observations. The predictive skill degrades slightly with lead time; by January (three month leadtime), the probability of a forecasted heatwave actually occurring drops to 26%.\n\nRegarding the capture of observed events (Hit Rate), the model successfully anticipates 33% of the heatwaves that occurred in November. This capability decreases marginally to 30% in December and 27% in January. Overall, the system demonstrates an ability to capture approximately one-third of heatwave events at a one-month lead time, though users must be mindful of the high false alarm ratio.\n\n(seasonal_ohc:section-6)= \n### 6. Annual Variability and ENSO Influence\nWe now investigate the year-to-year skill for each month and its potential connection to ENSO.\n\n---BSS in float ---\n--- free memory ---\n-------------------\nSEDI & Rates from the int8 arrays\nThe int-based computations\n--- free mem ---\n-------------------\n--- STEP 3: Combine results -------\n--- combine and format ---\n\n#### Annual metrics - global\n\n--- Plotting function ---\n--- Panel A: RONI NDJ index (narrow bar chart) ---\n--- Panel B: MHW Occurrence ---\n--- Panel C: Skill Scores ---\n--- Panel D: Contingency Rates ---\nShared x-axis limits\n2. Run global calculation and plot\n\n(fig-global-ts-in)=\n##### Figure 8 \nAnnual time series of globally-averaged forecast performance for the NDJ season. The upper panel depticts occurrence of El Niño according to the NCEP NOAA Relative Oceanic Niño Index (RONI), defines as the three month running average of the relative Niño 3.4 index. The second row depicts global marine heatwave (MHW) occurrence each year, i.e. the percentage of gridcells in heatwave conditions. The third plot shows the SEDI and the BSS in NDJ averaged over the valid ocean region. The contingency table components, the False or True Positives and Negatives each year for a selected month, November, are shown in the lower plot. The figure is similar to Extended Data Fig. 5 in Jacox et. al. 2022, [[1]](https://doi.org/10.1038/s41586-022-04573-9).\n\n---\n\nAn analysis of global year-to-year variability in SEDI and BSS ([](fig-global-ts-in)) reveals a divergence in skill metrics. The SEDI scores (orange lines) remain positive and notably peak during major El Niño events (e.g., 1997-98, 2015-16). This indicates that the forecast system is highly effective at signal detection; it successfully identifies high-risk periods during these major climate drivers. Conversely, the BSS (purple lines), which measures probabilistic skill, often deteriorates during these same events. This might indicate that the single-model system is overconfident. The distribution analysis ([](fig-pdfs)) shows that the reference reanalysis exhibits a sharper tail than the forecast ensemble. While the model often captures the correct warming signal, this lower extreme-tail variability may be linked to the observed overconfidence in its probabilistic assignments.\n\nWhile the model correctly predicts the risk of a heatwave (high SEDI), the probability it assigns is less accurate. This is a common limitation of single-model systems, where the ensemble spread is insufficient to represent true uncertainty. Transitioning to a multi-model ensemble would likely improve reliability for decision-making by increasing that spread.\n\nWe focus on a key area: the Niño 3.4 region (5°N-5°S, 170-120°W). The Niño 3.4 region was chosen as it represents the tropical pacific and the ENSO.\n\n#### Annual metrics - Niño 3.4 region\n\n1. Select the data for the Niño 3.4 region\n4. Clean up large intermediate masks\ndel nino34_mask_forecast, nino34_mask_reanalysis\ngc.collect();\n\n```text\n--- Running Case Study: Niño 3.4 Region ---\n```\n\n(fig-nino-ts)=\n##### Figure 9 \nAs in [](fig-global-ts-in), but for the Niño 3.4 region.\n\n---", "text_with_prefix": "EQC Quality Assessment: \"Seasonal Forecast Prediction of Upper Ocean Heat Content Extremes\"\nDataset: seasonal-monthly-ocean [CDS]\nAspect: extremes-detection_q07 | Category: Seasonal_Forecasts\nSection: Seasonal Forecast Prediction of Upper Ocean Heat Content Extremes > Quality assessment statement\n---\nprint(\"Computing contingency table\")\nFormat to 2 decimal places with a '%' sign\n\n```text\n--- Contingency Tables in Percent (%) ---\n\nNovember:\n```\n\n```text\nDecember:\n```\n\n```text\nJanuary:\n```\n\nFor all Novembers (one month leadtime, or 1-2 months, to be exact), when the model predicts a marine heatwave, that prediction is correct 32% of the time (Precision). Conversely, the False Discovery Rate is 68%, meaning roughly two-thirds of forecasted heatwaves did not materialize in observations. The predictive skill degrades slightly with lead time; by January (three month leadtime), the probability of a forecasted heatwave actually occurring drops to 26%.\n\nRegarding the capture of observed events (Hit Rate), the model successfully anticipates 33% of the heatwaves that occurred in November. This capability decreases marginally to 30% in December and 27% in January. Overall, the system demonstrates an ability to capture approximately one-third of heatwave events at a one-month lead time, though users must be mindful of the high false alarm ratio.\n\n(seasonal_ohc:section-6)= \n### 6. Annual Variability and ENSO Influence\nWe now investigate the year-to-year skill for each month and its potential connection to ENSO.\n\n---BSS in float ---\n--- free memory ---\n-------------------\nSEDI & Rates from the int8 arrays\nThe int-based computations\n--- free mem ---\n-------------------\n--- STEP 3: Combine results -------\n--- combine and format ---\n\n#### Annual metrics - global\n\n--- Plotting function ---\n--- Panel A: RONI NDJ index (narrow bar chart) ---\n--- Panel B: MHW Occurrence ---\n--- Panel C: Skill Scores ---\n--- Panel D: Contingency Rates ---\nShared x-axis limits\n2. Run global calculation and plot\n\n(fig-global-ts-in)=\n##### Figure 8 \nAnnual time series of globally-averaged forecast performance for the NDJ season. The upper panel depticts occurrence of El Niño according to the NCEP NOAA Relative Oceanic Niño Index (RONI), defines as the three month running average of the relative Niño 3.4 index. The second row depicts global marine heatwave (MHW) occurrence each year, i.e. the percentage of gridcells in heatwave conditions. The third plot shows the SEDI and the BSS in NDJ averaged over the valid ocean region. The contingency table components, the False or True Positives and Negatives each year for a selected month, November, are shown in the lower plot. The figure is similar to Extended Data Fig. 5 in Jacox et. al. 2022, [[1]](https://doi.org/10.1038/s41586-022-04573-9).\n\n---\n\nAn analysis of global year-to-year variability in SEDI and BSS ([](fig-global-ts-in)) reveals a divergence in skill metrics. The SEDI scores (orange lines) remain positive and notably peak during major El Niño events (e.g., 1997-98, 2015-16). This indicates that the forecast system is highly effective at signal detection; it successfully identifies high-risk periods during these major climate drivers. Conversely, the BSS (purple lines), which measures probabilistic skill, often deteriorates during these same events. This might indicate that the single-model system is overconfident. The distribution analysis ([](fig-pdfs)) shows that the reference reanalysis exhibits a sharper tail than the forecast ensemble. While the model often captures the correct warming signal, this lower extreme-tail variability may be linked to the observed overconfidence in its probabilistic assignments.\n\nWhile the model correctly predicts the risk of a heatwave (high SEDI), the probability it assigns is less accurate. This is a common limitation of single-model systems, where the ensemble spread is insufficient to represent true uncertainty. Transitioning to a multi-model ensemble would likely improve reliability for decision-making by increasing that spread.\n\nWe focus on a key area: the Niño 3.4 region (5°N-5°S, 170-120°W). The Niño 3.4 region was chosen as it represents the tropical pacific and the ENSO.\n\n#### Annual metrics - Niño 3.4 region\n\n1. Select the data for the Niño 3.4 region\n4. Clean up large intermediate masks\ndel nino34_mask_forecast, nino34_mask_reanalysis\ngc.collect();\n\n```text\n--- Running Case Study: Niño 3.4 Region ---\n```\n\n(fig-nino-ts)=\n##### Figure 9 \nAs in [](fig-global-ts-in), but for the Niño 3.4 region.\n\n---"} {"chunk_id": "seasonal_seasonal-monthly-ocean_extremes-detection_q07__625f82364b4f", "report_id": "seasonal_seasonal-monthly-ocean_extremes-detection_q07", "dataset_id": "seasonal-monthly-ocean", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extremes-detection_q07", "aspect_base": "extremes-detection", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Seasonal Forecast Prediction of Upper Ocean Heat Content Extremes > Quality assessment statement", "title": "Seasonal Forecast Prediction of Upper Ocean Heat Content Extremes", "chunk_index": 9, "token_count": 1087, "text_raw": "--- Running Case Study: Niño 3.4 Region ---\n```\n\n(fig-nino-ts)=\n##### Figure 9 \nAs in [](fig-global-ts-in), but for the Niño 3.4 region.\n\n---\n\nIn contrast, the skill within the Niño 3.4 region ([](fig-nino-ts)) is much higher and more reliable. Although the time-series exhibits higher variability, which is expected a dynamic and smaller region, the BSS is consistently positive, approaching 1.0 during the major 1997-98 and 2015-16 events. The high SEDI scores further confirm the model's ability to capture these extremes. This confirms that the probabilistic forecasts are both skillful and reliable regarding marine heatwaves during ENSO events, which are the primary driver of seasonal predictability.\n\n(seasonal_ohc:section-7)= \n### Summary of Results\n\nThis study compares the Météo-France System-9 seasonal forecasts against the ORAS5 ocean reanalysis for forecasts initialized around October 1st for mean monthly values in November, December, and January. The analysis focused on the forecast skill for extreme upper OHC events (anomalies >90th percentile) after removing long-term trends from both datasets. By looking at anomalies, detrending the data, and applying lead-time (or calendar month) dependent percentiles we do our best to match the seasonal data with the reference data. We assess skill 1–3 months after the forecasts nominal start date using metrics such as the Symmetric Extremal Dependence Index (SEDI) and the Brier Skill Score (BSS). The metrics considered follow Jacox et al. (2022) [[1]](https://doi.org/10.1038/s41586-022-04573-9), however, in this work we look at OHC and not SST. It should be noted that this evaluation and use-case considers a single ocean reanalysis as a reference dataset. Using an ensemble of observational datasets could provide a more robust assessment.\n\nThe forecast system demonstrates clear, positive skill in detecting events. The Symmetric Extremal Dependence Index (SEDI) is positive globally and spikes during major El Niño events (e.g., 1997-98, 2015-16). This is corroborated by the contingency rate analysis ([](fig-global-ts-in) lower panel), which shows the True Positive (Hit) Rate increasing dramatically in these years, confirming the model successfully captures the El Niño signal, and the corresponding effect on the MHW index derived from the ocean heat content.\n\nThe system provides a valuable tool for anticipation of NDJ upper ocean heatwaves, with its utility and reliability being highest in ENSO-dominated regions. However, outside this region the Brier Skill Score (BSS) is often negative, indicating that the forecast probabilities are less reliable than the static climatological reference forecast (always a 10% chance). This might indicate that the model has too little ensemble spread in these regions. This makes the forecast a good \"heads-up\" warning for high-risk periods, but for higher reliability outside the main ENSO regions a multi-model system might be useful.\n\n## ℹ️ If you want to know more\n\n### Key resources\n\n* Seasonal forecast monthly averages of ocean variables DOI: [10.24381/cds.2f9be611](https://doi.org/10.24381/cds.2f9be611)\n- ORAS5 global ocean reanalysis monthly data from 1958 to present DOI: [10.24381/cds.67e8eeb7](https://doi.org/10.24381/cds.67e8eeb7)\n\nCode libraries used:\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared and advised on by [B-Open](https://www.bopen.eu/)\n\n* xesmf\n\n* [ocean-regrid](https://github.com/COSIMA/ocean-regrid/tree/master) `make_corners` to provide cell edge coordinates for the ORCA025 tripolar grid\n\n* xskillscore\n\n* dask\n\n* xarray\n\n* pandas\n\n* matplotlib\n\n* seaborn\n\n* scipy\n\n* numpy\n\n* cartopy\n\n### References\n\n[[1]](https://doi.org/10.1038/s41586-022-04573-9) Jacox, M. G., Alexander, M. A., Amaya, D., Becker, E., Bograd, S. J., Brodie, S., Hazen, E. L., Pozo Buil, M., & Tommasi, D. (2022). Global seasonal forecasts of marine heatwaves. Nature, 604(7906), 486–490.", "text_with_prefix": "EQC Quality Assessment: \"Seasonal Forecast Prediction of Upper Ocean Heat Content Extremes\"\nDataset: seasonal-monthly-ocean [CDS]\nAspect: extremes-detection_q07 | Category: Seasonal_Forecasts\nSection: Seasonal Forecast Prediction of Upper Ocean Heat Content Extremes > Quality assessment statement\n---\n--- Running Case Study: Niño 3.4 Region ---\n```\n\n(fig-nino-ts)=\n##### Figure 9 \nAs in [](fig-global-ts-in), but for the Niño 3.4 region.\n\n---\n\nIn contrast, the skill within the Niño 3.4 region ([](fig-nino-ts)) is much higher and more reliable. Although the time-series exhibits higher variability, which is expected a dynamic and smaller region, the BSS is consistently positive, approaching 1.0 during the major 1997-98 and 2015-16 events. The high SEDI scores further confirm the model's ability to capture these extremes. This confirms that the probabilistic forecasts are both skillful and reliable regarding marine heatwaves during ENSO events, which are the primary driver of seasonal predictability.\n\n(seasonal_ohc:section-7)= \n### Summary of Results\n\nThis study compares the Météo-France System-9 seasonal forecasts against the ORAS5 ocean reanalysis for forecasts initialized around October 1st for mean monthly values in November, December, and January. The analysis focused on the forecast skill for extreme upper OHC events (anomalies >90th percentile) after removing long-term trends from both datasets. By looking at anomalies, detrending the data, and applying lead-time (or calendar month) dependent percentiles we do our best to match the seasonal data with the reference data. We assess skill 1–3 months after the forecasts nominal start date using metrics such as the Symmetric Extremal Dependence Index (SEDI) and the Brier Skill Score (BSS). The metrics considered follow Jacox et al. (2022) [[1]](https://doi.org/10.1038/s41586-022-04573-9), however, in this work we look at OHC and not SST. It should be noted that this evaluation and use-case considers a single ocean reanalysis as a reference dataset. Using an ensemble of observational datasets could provide a more robust assessment.\n\nThe forecast system demonstrates clear, positive skill in detecting events. The Symmetric Extremal Dependence Index (SEDI) is positive globally and spikes during major El Niño events (e.g., 1997-98, 2015-16). This is corroborated by the contingency rate analysis ([](fig-global-ts-in) lower panel), which shows the True Positive (Hit) Rate increasing dramatically in these years, confirming the model successfully captures the El Niño signal, and the corresponding effect on the MHW index derived from the ocean heat content.\n\nThe system provides a valuable tool for anticipation of NDJ upper ocean heatwaves, with its utility and reliability being highest in ENSO-dominated regions. However, outside this region the Brier Skill Score (BSS) is often negative, indicating that the forecast probabilities are less reliable than the static climatological reference forecast (always a 10% chance). This might indicate that the model has too little ensemble spread in these regions. This makes the forecast a good \"heads-up\" warning for high-risk periods, but for higher reliability outside the main ENSO regions a multi-model system might be useful.\n\n## ℹ️ If you want to know more\n\n### Key resources\n\n* Seasonal forecast monthly averages of ocean variables DOI: [10.24381/cds.2f9be611](https://doi.org/10.24381/cds.2f9be611)\n- ORAS5 global ocean reanalysis monthly data from 1958 to present DOI: [10.24381/cds.67e8eeb7](https://doi.org/10.24381/cds.67e8eeb7)\n\nCode libraries used:\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared and advised on by [B-Open](https://www.bopen.eu/)\n\n* xesmf\n\n* [ocean-regrid](https://github.com/COSIMA/ocean-regrid/tree/master) `make_corners` to provide cell edge coordinates for the ORCA025 tripolar grid\n\n* xskillscore\n\n* dask\n\n* xarray\n\n* pandas\n\n* matplotlib\n\n* seaborn\n\n* scipy\n\n* numpy\n\n* cartopy\n\n### References\n\n[[1]](https://doi.org/10.1038/s41586-022-04573-9) Jacox, M. G., Alexander, M. A., Amaya, D., Becker, E., Bograd, S. J., Brodie, S., Hazen, E. L., Pozo Buil, M., & Tommasi, D. (2022). Global seasonal forecasts of marine heatwaves. Nature, 604(7906), 486–490."} {"chunk_id": "seasonal_seasonal-monthly-ocean_extremes-detection_q07__0109a0494170", "report_id": "seasonal_seasonal-monthly-ocean_extremes-detection_q07", "dataset_id": "seasonal-monthly-ocean", "store": "CDS", "doc_type": "EQC_QA", "aspect": "extremes-detection_q07", "aspect_base": "extremes-detection", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Seasonal Forecast Prediction of Upper Ocean Heat Content Extremes > Quality assessment statement", "title": "Seasonal Forecast Prediction of Upper Ocean Heat Content Extremes", "chunk_index": 10, "token_count": 362, "text_raw": "S., Hazen, E. L., Pozo Buil, M., & Tommasi, D. (2022). Global seasonal forecasts of marine heatwaves. Nature, 604(7906), 486–490.\n\n[[2]](https://doi.org/10.1007/S00382-021-06101-3) McAdam, R., Masina, S., Balmaseda, M., Gualdi, S., Senan, R., & Mayer, M. (2022). Seasonal forecast skill of upper-ocean heat content in coupled high-resolution systems. Climate Dynamics, 58(11–12), 3335–3350.\n\n[[3]](https://doi.org/10.5281/zenodo.15518106) Specq, D., Dorel, L., Beuvier, J., Ardilouze, C., & Batté, L. (2024). Documentation of the Météo-France seasonal forecasting system 9. Zenodo.\n\n[[4]](https://doi.org/10.1175/WAF-D-10-05030.1) Ferro, C. A. T., & Stephenson, D. B. (2011). Extremal Dependence Indices: Improved Verification Measures for Deterministic Forecasts of Rare Binary Events. Weather and Forecasting, 26(5), 699–713.", "text_with_prefix": "EQC Quality Assessment: \"Seasonal Forecast Prediction of Upper Ocean Heat Content Extremes\"\nDataset: seasonal-monthly-ocean [CDS]\nAspect: extremes-detection_q07 | Category: Seasonal_Forecasts\nSection: Seasonal Forecast Prediction of Upper Ocean Heat Content Extremes > Quality assessment statement\n---\nS., Hazen, E. L., Pozo Buil, M., & Tommasi, D. (2022). Global seasonal forecasts of marine heatwaves. Nature, 604(7906), 486–490.\n\n[[2]](https://doi.org/10.1007/S00382-021-06101-3) McAdam, R., Masina, S., Balmaseda, M., Gualdi, S., Senan, R., & Mayer, M. (2022). Seasonal forecast skill of upper-ocean heat content in coupled high-resolution systems. Climate Dynamics, 58(11–12), 3335–3350.\n\n[[3]](https://doi.org/10.5281/zenodo.15518106) Specq, D., Dorel, L., Beuvier, J., Ardilouze, C., & Batté, L. (2024). Documentation of the Météo-France seasonal forecasting system 9. Zenodo.\n\n[[4]](https://doi.org/10.1175/WAF-D-10-05030.1) Ferro, C. A. T., & Stephenson, D. B. (2011). Extremal Dependence Indices: Improved Verification Measures for Deterministic Forecasts of Rare Binary Events. Weather and Forecasting, 26(5), 699–713."} {"chunk_id": "seasonal_seasonal-monthly-single-levels_forecast-skill_q02__8dc6a77b84b9", "report_id": "seasonal_seasonal-monthly-single-levels_forecast-skill_q02", "dataset_id": "seasonal-monthly-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "forecast-skill_q02", "aspect_base": "forecast-skill", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Seasonal forecasts bias assessment for impact models > Quality assessment questions", "title": "Seasonal forecasts bias assessment for impact models", "chunk_index": 0, "token_count": 376, "text_raw": "* **Do I have to correct systematic errors in temperature and precipitation before using seasonal forecasts as an input for my crop model?**\n* **Can I assume that climate models always produce the same systematic error for seasonal forecasts over a certain area?**\n\nSeasonal forecasts can be used as an input to crop models for agricultural risk assessment [[1]](https://doi.org/10.1038/s41598-018-20628-2) [[2]](https://doi.org/10.1016/j.agrformet.2018.11.029) [[3]](https://doi.org/10.1016/j.cliser.2022.100324). However, the presence of systematic errors can undermine the performance of crop models in a number of ways, particularly when critical thresholds in the plant growth process are considered. Similar problems can arise in other sectoral applications, such as the management of renewable energy facilities [[4]](https://doi.org/10.1016/j.renene.2019.04.135), or the alerting of national health services for the occurrence of summer heat waves [[5]](https://doi.org/10.1007/s00382-021-05828-3). In addition to temperature and precipitation, the most commonly used climate variables for risk assessment in all sectoral applications, the notebook also considers wind speed and dew point temperature. While these variables are commonly used in agricultural applications to calculate potential evapotranspiration, the results of the assessment are also valuable for several other sectors.", "text_with_prefix": "EQC Quality Assessment: \"Seasonal forecasts bias assessment for impact models\"\nDataset: seasonal-monthly-single-levels [CDS]\nAspect: forecast-skill_q02 | Category: Seasonal_Forecasts\nSection: Seasonal forecasts bias assessment for impact models > Quality assessment questions\n---\n* **Do I have to correct systematic errors in temperature and precipitation before using seasonal forecasts as an input for my crop model?**\n* **Can I assume that climate models always produce the same systematic error for seasonal forecasts over a certain area?**\n\nSeasonal forecasts can be used as an input to crop models for agricultural risk assessment [[1]](https://doi.org/10.1038/s41598-018-20628-2) [[2]](https://doi.org/10.1016/j.agrformet.2018.11.029) [[3]](https://doi.org/10.1016/j.cliser.2022.100324). However, the presence of systematic errors can undermine the performance of crop models in a number of ways, particularly when critical thresholds in the plant growth process are considered. Similar problems can arise in other sectoral applications, such as the management of renewable energy facilities [[4]](https://doi.org/10.1016/j.renene.2019.04.135), or the alerting of national health services for the occurrence of summer heat waves [[5]](https://doi.org/10.1007/s00382-021-05828-3). In addition to temperature and precipitation, the most commonly used climate variables for risk assessment in all sectoral applications, the notebook also considers wind speed and dew point temperature. While these variables are commonly used in agricultural applications to calculate potential evapotranspiration, the results of the assessment are also valuable for several other sectors."} {"chunk_id": "seasonal_seasonal-monthly-single-levels_forecast-skill_q02__f716a3dc5f92", "report_id": "seasonal_seasonal-monthly-single-levels_forecast-skill_q02", "dataset_id": "seasonal-monthly-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "forecast-skill_q02", "aspect_base": "forecast-skill", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Seasonal forecasts bias assessment for impact models > Quality assessment statement", "title": "Seasonal forecasts bias assessment for impact models", "chunk_index": 1, "token_count": 454, "text_raw": "These are the key outcomes of this assessment\n\n* Seasonal forecasts should be bias-corrected before being used as input to climate impact models. Depending on the impact model, it might be advisable to bias-correct the resulting impact indicator [[6]](https://doi.org/10.1175/JHM-D-19-0042.1)\n* Model drifts and bias tend to be a complex function of both the start time of the forecast and of its valid time, with a slightly stronger dependence on the valid time for the case of temperature and precipitation. Therefore, the most appropriate approach to produce unbiased forecasts for use in impact models must be selected on a case-by-case basis.[[7]](https://doi.org/10.1007/s00382-019-04640-4) [[8]](https://doi.org/10.1007/s00382-017-3962-9)\n* For the implementation of bias correction procedures, it is recommended that reference climatologies are computed for time scales no longer than one month, since the typical model bias changes on a monthly basis. Instead, the correction of model biases on a seasonal (i.e. three months) timescale may lead to a loss of quality.\n* Similarly, the calculation of multi-model statistics requires a recalibration of the model output in order to align all models around the same climatological distribution. [[9]](https://doi.org/10.1007/s00382-020-05314-2)\n```\n\nattachment:8568344f-a005-4668-b14a-05d8ff7f3b96.png\n---\nheight: 500px\n---\nThe figure shows a sample of the ECMWF monthly temperature bias for Western North America as a function of the starting month and of the valid month of the forecast. More details on how to interpret the results are available below in the notebook.\n```", "text_with_prefix": "EQC Quality Assessment: \"Seasonal forecasts bias assessment for impact models\"\nDataset: seasonal-monthly-single-levels [CDS]\nAspect: forecast-skill_q02 | Category: Seasonal_Forecasts\nSection: Seasonal forecasts bias assessment for impact models > Quality assessment statement\n---\nThese are the key outcomes of this assessment\n\n* Seasonal forecasts should be bias-corrected before being used as input to climate impact models. Depending on the impact model, it might be advisable to bias-correct the resulting impact indicator [[6]](https://doi.org/10.1175/JHM-D-19-0042.1)\n* Model drifts and bias tend to be a complex function of both the start time of the forecast and of its valid time, with a slightly stronger dependence on the valid time for the case of temperature and precipitation. Therefore, the most appropriate approach to produce unbiased forecasts for use in impact models must be selected on a case-by-case basis.[[7]](https://doi.org/10.1007/s00382-019-04640-4) [[8]](https://doi.org/10.1007/s00382-017-3962-9)\n* For the implementation of bias correction procedures, it is recommended that reference climatologies are computed for time scales no longer than one month, since the typical model bias changes on a monthly basis. Instead, the correction of model biases on a seasonal (i.e. three months) timescale may lead to a loss of quality.\n* Similarly, the calculation of multi-model statistics requires a recalibration of the model output in order to align all models around the same climatological distribution. [[9]](https://doi.org/10.1007/s00382-020-05314-2)\n```\n\nattachment:8568344f-a005-4668-b14a-05d8ff7f3b96.png\n---\nheight: 500px\n---\nThe figure shows a sample of the ECMWF monthly temperature bias for Western North America as a function of the starting month and of the valid month of the forecast. More details on how to interpret the results are available below in the notebook.\n```"} {"chunk_id": "seasonal_seasonal-monthly-single-levels_forecast-skill_q02__ee7d0a32180c", "report_id": "seasonal_seasonal-monthly-single-levels_forecast-skill_q02", "dataset_id": "seasonal-monthly-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "forecast-skill_q02", "aspect_base": "forecast-skill", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Seasonal forecasts bias assessment for impact models > Methodology", "title": "Seasonal forecasts bias assessment for impact models", "chunk_index": 2, "token_count": 479, "text_raw": "This notebook provides a comprehensive assessment of the systematic errors of the forecast systems by comparing the model predictions with the ERA5 reanalysis. The analysis is carried out for eight regions selected from those used in the IPCC-AR6 (see figure below), which capture the spatial scale of the main systematic errors already analysed in previous studies [[10]](https://doi.org/10.1007/s00382-019-04640-4)\n\nThe analysis and results are organised in the following steps, which are detailed in the sections below:\n\n**[](section-1):**\n * Import required packages\n * Define parameters describing regions, variables, and data\n * Plot the chosen regions\n * Define required functions\n * Show a sample global map of temperature biases\n\n**[](section-2):**\n * Retrieve monthly ERA5 data and compute the regional mean climatology, i.e. 12 values for each area\n * Compute area averages\n\n**[](section-3):**\n * Retrieve monthly seasonal forecast data from all available centres and for all starting dates, compute the monthly mean climatology (i.e. 6 values, since each forecast is valid for 6 months) and subtract the corresponding ERA5 climatology.\n * Compute area averages\n * Plot a sample map\n\n**[](section-4)**\n\n**[](section-5):** \n * Plots are shown in a comprehensive set of panels that provide an overview of the multimodel performance \n * The results are summarised and discussed\n\nNote that crop models and other sectoral applications tend to use daily values rather than the monthly statistics considered in this notebook, and they focus on smaller areas compared to the regions considered below. However, the analysis of monthly data over the IPCC-AR6 regions provides a useful overview of the overall characteristics of the dataset.\n\nNote that for ECCC, only one of the two systems has been analysed (system 3). This is different from the approach taken in the C3S charts, and CanSIPSv2.1, where both systems are combined (system 2 and system 3).", "text_with_prefix": "EQC Quality Assessment: \"Seasonal forecasts bias assessment for impact models\"\nDataset: seasonal-monthly-single-levels [CDS]\nAspect: forecast-skill_q02 | Category: Seasonal_Forecasts\nSection: Seasonal forecasts bias assessment for impact models > Methodology\n---\nThis notebook provides a comprehensive assessment of the systematic errors of the forecast systems by comparing the model predictions with the ERA5 reanalysis. The analysis is carried out for eight regions selected from those used in the IPCC-AR6 (see figure below), which capture the spatial scale of the main systematic errors already analysed in previous studies [[10]](https://doi.org/10.1007/s00382-019-04640-4)\n\nThe analysis and results are organised in the following steps, which are detailed in the sections below:\n\n**[](section-1):**\n * Import required packages\n * Define parameters describing regions, variables, and data\n * Plot the chosen regions\n * Define required functions\n * Show a sample global map of temperature biases\n\n**[](section-2):**\n * Retrieve monthly ERA5 data and compute the regional mean climatology, i.e. 12 values for each area\n * Compute area averages\n\n**[](section-3):**\n * Retrieve monthly seasonal forecast data from all available centres and for all starting dates, compute the monthly mean climatology (i.e. 6 values, since each forecast is valid for 6 months) and subtract the corresponding ERA5 climatology.\n * Compute area averages\n * Plot a sample map\n\n**[](section-4)**\n\n**[](section-5):** \n * Plots are shown in a comprehensive set of panels that provide an overview of the multimodel performance \n * The results are summarised and discussed\n\nNote that crop models and other sectoral applications tend to use daily values rather than the monthly statistics considered in this notebook, and they focus on smaller areas compared to the regions considered below. However, the analysis of monthly data over the IPCC-AR6 regions provides a useful overview of the overall characteristics of the dataset.\n\nNote that for ECCC, only one of the two systems has been analysed (system 3). This is different from the approach taken in the C3S charts, and CanSIPSv2.1, where both systems are combined (system 2 and system 3)."} {"chunk_id": "seasonal_seasonal-monthly-single-levels_forecast-skill_q02__904d9b2b3edb", "report_id": "seasonal_seasonal-monthly-single-levels_forecast-skill_q02", "dataset_id": "seasonal-monthly-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "forecast-skill_q02", "aspect_base": "forecast-skill", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Seasonal forecasts bias assessment for impact models > Analysis and results > 1. Choose the data to use and setup the code > Import required packages", "title": "Seasonal forecasts bias assessment for impact models", "chunk_index": 3, "token_count": 155, "text_raw": "We use the `regionmask package` to extract climatologies for a few selected regiones that have already e been adopted in several climate studies, including the analysis presented in the latest IPCC report.\n\nA description of the regions adopted for climate assesements in the scientific literature is available [here](https://regionmask.readthedocs.io/en/stable/defined_scientific.html).", "text_with_prefix": "EQC Quality Assessment: \"Seasonal forecasts bias assessment for impact models\"\nDataset: seasonal-monthly-single-levels [CDS]\nAspect: forecast-skill_q02 | Category: Seasonal_Forecasts\nSection: Seasonal forecasts bias assessment for impact models > Analysis and results > 1. Choose the data to use and setup the code > Import required packages\n---\nWe use the `regionmask package` to extract climatologies for a few selected regiones that have already e been adopted in several climate studies, including the analysis presented in the latest IPCC report.\n\nA description of the regions adopted for climate assesements in the scientific literature is available [here](https://regionmask.readthedocs.io/en/stable/defined_scientific.html)."} {"chunk_id": "seasonal_seasonal-monthly-single-levels_forecast-skill_q02__eaa60e91949b", "report_id": "seasonal_seasonal-monthly-single-levels_forecast-skill_q02", "dataset_id": "seasonal-monthly-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "forecast-skill_q02", "aspect_base": "forecast-skill", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Seasonal forecasts bias assessment for impact models > Analysis and results > 1. Choose the data to use and setup the code > Define parameters describing regions, variables, and data", "title": "Seasonal forecasts bias assessment for impact models", "chunk_index": 4, "token_count": 257, "text_raw": "In this notebook the analysis of the model bias is done for 5 variables from all 8 originating centres. The SREX regions selected for the analysis are:\n\n- NEAF: Northern-East Africa\n- ENA: Eastern North America\n- MED: Southern Europe/Mediterranean\n- NEB: North-eastern Brazil\n- SAS: Southern Asia\n- SEA: South-Eastern Asia\n- WNA: Western North America\n- NWS: Western Coast of South America\n\nA `common_request` is defined in order to cover the entire global domain and the entire hindcast period from 1993 to 2016. Specific data requests are then defined for the reanalysis and for the seasonal forecasts, so that all available months and lead times are retrieved for the analysis.\n\nDefine centres with missing variables\nPlotting settings\nDefine data requests", "text_with_prefix": "EQC Quality Assessment: \"Seasonal forecasts bias assessment for impact models\"\nDataset: seasonal-monthly-single-levels [CDS]\nAspect: forecast-skill_q02 | Category: Seasonal_Forecasts\nSection: Seasonal forecasts bias assessment for impact models > Analysis and results > 1. Choose the data to use and setup the code > Define parameters describing regions, variables, and data\n---\nIn this notebook the analysis of the model bias is done for 5 variables from all 8 originating centres. The SREX regions selected for the analysis are:\n\n- NEAF: Northern-East Africa\n- ENA: Eastern North America\n- MED: Southern Europe/Mediterranean\n- NEB: North-eastern Brazil\n- SAS: Southern Asia\n- SEA: South-Eastern Asia\n- WNA: Western North America\n- NWS: Western Coast of South America\n\nA `common_request` is defined in order to cover the entire global domain and the entire hindcast period from 1993 to 2016. Specific data requests are then defined for the reanalysis and for the seasonal forecasts, so that all available months and lead times are retrieved for the analysis.\n\nDefine centres with missing variables\nPlotting settings\nDefine data requests"} {"chunk_id": "seasonal_seasonal-monthly-single-levels_forecast-skill_q02__01d0300a8e0b", "report_id": "seasonal_seasonal-monthly-single-levels_forecast-skill_q02", "dataset_id": "seasonal-monthly-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "forecast-skill_q02", "aspect_base": "forecast-skill", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Seasonal forecasts bias assessment for impact models > Analysis and results > 1. Choose the data to use and setup the code > Plot the chosen regions", "title": "Seasonal forecasts bias assessment for impact models", "chunk_index": 5, "token_count": 141, "text_raw": "For the bias analysis we use 8 out of the 45 regions defined in the `regionmask` package in order to assess the systematic errors of the modelling systems under different climatic regimes (tropical, extratropical, continental, maritime) with a number of cases maneageable within the notebook.", "text_with_prefix": "EQC Quality Assessment: \"Seasonal forecasts bias assessment for impact models\"\nDataset: seasonal-monthly-single-levels [CDS]\nAspect: forecast-skill_q02 | Category: Seasonal_Forecasts\nSection: Seasonal forecasts bias assessment for impact models > Analysis and results > 1. Choose the data to use and setup the code > Plot the chosen regions\n---\nFor the bias analysis we use 8 out of the 45 regions defined in the `regionmask` package in order to assess the systematic errors of the modelling systems under different climatic regimes (tropical, extratropical, continental, maritime) with a number of cases maneageable within the notebook."} {"chunk_id": "seasonal_seasonal-monthly-single-levels_forecast-skill_q02__23169a335aec", "report_id": "seasonal_seasonal-monthly-single-levels_forecast-skill_q02", "dataset_id": "seasonal-monthly-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "forecast-skill_q02", "aspect_base": "forecast-skill", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Seasonal forecasts bias assessment for impact models > Analysis and results > 1. Choose the data to use and setup the code > Define the required functions", "title": "Seasonal forecasts bias assessment for impact models", "chunk_index": 6, "token_count": 367, "text_raw": "- The `regionalised_spatial_weighted_mean` function extracts the regional means over the selected domains. It uses spatial weighting to account for the latitudinal dependence of the grid size in the lon-lat grids used for the reanalysis and for the forecast models. The bias is then calculated later in the notebook by simply subtracting the regional means.\n\n- The `postprocess_dataarray` function performs two important postprocessings operations required to align the arrays containing the reanalsysis and the forecast data.\n\n- The time dimensions are renamed to allow the broadcasting of the dimension `valid_time`, which is used to compute the bias, independently of the corresponding `lead_times` in the forecast.\n\n- Furthermore, the cumulative variables have to be rescaled in order to make ERA5 data and seasonal forecast data comparable on a monthly timescale. A [conversion table](https://confluence.ecmwf.int/pages/viewpage.action?pageId=197702790) for accumulated variables is available in the [Copernicus Knowledge Base](https://confluence.ecmwf.int/pages/viewpage.action?pageId=55116796) repository.\n\n- The function `get_seasonal_maps` is used to derive a sample bias map to illustrate how the analysis is performed.\n\n- The `monthly_mean` function will be used to derive the monthly values at the appropriate `monthly_mean` before calculating the bias.\n\nConvert units\n\n(section-2)=", "text_with_prefix": "EQC Quality Assessment: \"Seasonal forecasts bias assessment for impact models\"\nDataset: seasonal-monthly-single-levels [CDS]\nAspect: forecast-skill_q02 | Category: Seasonal_Forecasts\nSection: Seasonal forecasts bias assessment for impact models > Analysis and results > 1. Choose the data to use and setup the code > Define the required functions\n---\n- The `regionalised_spatial_weighted_mean` function extracts the regional means over the selected domains. It uses spatial weighting to account for the latitudinal dependence of the grid size in the lon-lat grids used for the reanalysis and for the forecast models. The bias is then calculated later in the notebook by simply subtracting the regional means.\n\n- The `postprocess_dataarray` function performs two important postprocessings operations required to align the arrays containing the reanalsysis and the forecast data.\n\n- The time dimensions are renamed to allow the broadcasting of the dimension `valid_time`, which is used to compute the bias, independently of the corresponding `lead_times` in the forecast.\n\n- Furthermore, the cumulative variables have to be rescaled in order to make ERA5 data and seasonal forecast data comparable on a monthly timescale. A [conversion table](https://confluence.ecmwf.int/pages/viewpage.action?pageId=197702790) for accumulated variables is available in the [Copernicus Knowledge Base](https://confluence.ecmwf.int/pages/viewpage.action?pageId=55116796) repository.\n\n- The function `get_seasonal_maps` is used to derive a sample bias map to illustrate how the analysis is performed.\n\n- The `monthly_mean` function will be used to derive the monthly values at the appropriate `monthly_mean` before calculating the bias.\n\nConvert units\n\n(section-2)="} {"chunk_id": "seasonal_seasonal-monthly-single-levels_forecast-skill_q02__6c23eeb769dc", "report_id": "seasonal_seasonal-monthly-single-levels_forecast-skill_q02", "dataset_id": "seasonal-monthly-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "forecast-skill_q02", "aspect_base": "forecast-skill", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Seasonal forecasts bias assessment for impact models > Analysis and results > 2. ERA5 data retrieval and area average", "title": "Seasonal forecasts bias assessment for impact models", "chunk_index": 7, "token_count": 125, "text_raw": "The `regionalised_spatial_weighted_mean` is applied for all regions and all variables selected for download. All time series of the regionalised means are collected in the same xarray `ds_reanalysis`.\n\nGet the reanalysis data\n\n(section-3)=", "text_with_prefix": "EQC Quality Assessment: \"Seasonal forecasts bias assessment for impact models\"\nDataset: seasonal-monthly-single-levels [CDS]\nAspect: forecast-skill_q02 | Category: Seasonal_Forecasts\nSection: Seasonal forecasts bias assessment for impact models > Analysis and results > 2. ERA5 data retrieval and area average\n---\nThe `regionalised_spatial_weighted_mean` is applied for all regions and all variables selected for download. All time series of the regionalised means are collected in the same xarray `ds_reanalysis`.\n\nGet the reanalysis data\n\n(section-3)="} {"chunk_id": "seasonal_seasonal-monthly-single-levels_forecast-skill_q02__824e57cbbde4", "report_id": "seasonal_seasonal-monthly-single-levels_forecast-skill_q02", "dataset_id": "seasonal-monthly-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "forecast-skill_q02", "aspect_base": "forecast-skill", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Seasonal forecasts bias assessment for impact models > Analysis and results > 3. Seasonal hindcast data retrieval and area average", "title": "Seasonal forecasts bias assessment for impact models", "chunk_index": 8, "token_count": 139, "text_raw": "The `regionalised_spatial_weighted_mean` is applied for all centres, regions and all variables selected for download. Missing variables are handled, so the download and transform process is not interrupted. All time series of the regionalised means are collected in the same xarray `ds_seasonal`.\n\nGet the seasonal forecast data", "text_with_prefix": "EQC Quality Assessment: \"Seasonal forecasts bias assessment for impact models\"\nDataset: seasonal-monthly-single-levels [CDS]\nAspect: forecast-skill_q02 | Category: Seasonal_Forecasts\nSection: Seasonal forecasts bias assessment for impact models > Analysis and results > 3. Seasonal hindcast data retrieval and area average\n---\nThe `regionalised_spatial_weighted_mean` is applied for all centres, regions and all variables selected for download. Missing variables are handled, so the download and transform process is not interrupted. All time series of the regionalised means are collected in the same xarray `ds_seasonal`.\n\nGet the seasonal forecast data"} {"chunk_id": "seasonal_seasonal-monthly-single-levels_forecast-skill_q02__03b2f4fbf6e0", "report_id": "seasonal_seasonal-monthly-single-levels_forecast-skill_q02", "dataset_id": "seasonal-monthly-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "forecast-skill_q02", "aspect_base": "forecast-skill", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Seasonal forecasts bias assessment for impact models > Analysis and results > 3. Seasonal hindcast data retrieval and area average > Show a sample global map of temperature biases", "title": "Seasonal forecasts bias assessment for impact models", "chunk_index": 9, "token_count": 217, "text_raw": "To reduce computation, this notebook simplifies the bias analysis by focusing on mean values of surface variables across the chosen regions. The goal is to offer a broad overview of the bias evolution over lead times. Thus, it is crucial that the selected regions exhibit sufficient homogeneity to minimize the compensation effects from systematic errors of varying sign.\n\nThis section of the notebook offers a simple tool for exploring sample bias maps and confirming that bias sign remain relatively constant within selected regions. Additionally, it serves as an insight on the behavior of individual models summarized in the comprehensive tables provided later in the notebook.\n\nSet parameters\nGet reanalysis map\nGet seasonal map\nCompute bias\n\n(section-4)=", "text_with_prefix": "EQC Quality Assessment: \"Seasonal forecasts bias assessment for impact models\"\nDataset: seasonal-monthly-single-levels [CDS]\nAspect: forecast-skill_q02 | Category: Seasonal_Forecasts\nSection: Seasonal forecasts bias assessment for impact models > Analysis and results > 3. Seasonal hindcast data retrieval and area average > Show a sample global map of temperature biases\n---\nTo reduce computation, this notebook simplifies the bias analysis by focusing on mean values of surface variables across the chosen regions. The goal is to offer a broad overview of the bias evolution over lead times. Thus, it is crucial that the selected regions exhibit sufficient homogeneity to minimize the compensation effects from systematic errors of varying sign.\n\nThis section of the notebook offers a simple tool for exploring sample bias maps and confirming that bias sign remain relatively constant within selected regions. Additionally, it serves as an insight on the behavior of individual models summarized in the comprehensive tables provided later in the notebook.\n\nSet parameters\nGet reanalysis map\nGet seasonal map\nCompute bias\n\n(section-4)="} {"chunk_id": "seasonal_seasonal-monthly-single-levels_forecast-skill_q02__b7363a8b8215", "report_id": "seasonal_seasonal-monthly-single-levels_forecast-skill_q02", "dataset_id": "seasonal-monthly-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "forecast-skill_q02", "aspect_base": "forecast-skill", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Seasonal forecasts bias assessment for impact models > Analysis and results > 4. Compute the hindcast bias", "title": "Seasonal forecasts bias assessment for impact models", "chunk_index": 10, "token_count": 277, "text_raw": "The the seasonal hindcast data are arranged in a four-dimensional array `ds_seasonal[i,j,n,k]`, where:\n> * `i` is the reference start month of the forecast (from January to December)\n> * `j` is the valid time of the forecast (from January to December)\n> * `n` is the originating centre of the forecast (currently a maximum of 8)\n> * `k` is the region (currently 8)\n\nCorrespondingly, ERA5 data are arranged in a three-dimensional array `ds_seasonal[j,n,k]`, where:\n> * `j` is the valid time of the forecast (from January to December)\n> * `n` is the originating centre of the forecast (currently a maximum of 8)\n> * `k` is the region (currently 8)\n\nWith such a shaping and setup of the datasets `ds_seasonal` and `ds_reanalysis` the computation of the bias is straightforward.\n\nAvoid ticks interpolation\n\n(section-5)=", "text_with_prefix": "EQC Quality Assessment: \"Seasonal forecasts bias assessment for impact models\"\nDataset: seasonal-monthly-single-levels [CDS]\nAspect: forecast-skill_q02 | Category: Seasonal_Forecasts\nSection: Seasonal forecasts bias assessment for impact models > Analysis and results > 4. Compute the hindcast bias\n---\nThe the seasonal hindcast data are arranged in a four-dimensional array `ds_seasonal[i,j,n,k]`, where:\n> * `i` is the reference start month of the forecast (from January to December)\n> * `j` is the valid time of the forecast (from January to December)\n> * `n` is the originating centre of the forecast (currently a maximum of 8)\n> * `k` is the region (currently 8)\n\nCorrespondingly, ERA5 data are arranged in a three-dimensional array `ds_seasonal[j,n,k]`, where:\n> * `j` is the valid time of the forecast (from January to December)\n> * `n` is the originating centre of the forecast (currently a maximum of 8)\n> * `k` is the region (currently 8)\n\nWith such a shaping and setup of the datasets `ds_seasonal` and `ds_reanalysis` the computation of the bias is straightforward.\n\nAvoid ticks interpolation\n\n(section-5)="} {"chunk_id": "seasonal_seasonal-monthly-single-levels_forecast-skill_q02__a9913eb61a4a", "report_id": "seasonal_seasonal-monthly-single-levels_forecast-skill_q02", "dataset_id": "seasonal-monthly-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "forecast-skill_q02", "aspect_base": "forecast-skill", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Seasonal forecasts bias assessment for impact models > Analysis and results > 5. Plot and describe results > How to read the results", "title": "Seasonal forecasts bias assessment for impact models", "chunk_index": 11, "token_count": 208, "text_raw": "The following tables summarise the monthly bias for chosen C3S seasonal systems (labelled by producing centre) and selected areas.\n\nIn each panel, the horizontal axis represents the validity period of the forecast, while the vertical axis shows the corresponding start dates. Thus, each horizontal strip of coloured tiles shows the evolution of the mean temperature bias over the six-month period of validity of each forecast.\n\nReading each panel from bottom to top, the strips of coloured tiles are progressively shifted by one month, corresponding to the same shift in the starting date of the forecast. The horizontal axis is periodic: forecasts starting after August have their last months on the left side of the panel.", "text_with_prefix": "EQC Quality Assessment: \"Seasonal forecasts bias assessment for impact models\"\nDataset: seasonal-monthly-single-levels [CDS]\nAspect: forecast-skill_q02 | Category: Seasonal_Forecasts\nSection: Seasonal forecasts bias assessment for impact models > Analysis and results > 5. Plot and describe results > How to read the results\n---\nThe following tables summarise the monthly bias for chosen C3S seasonal systems (labelled by producing centre) and selected areas.\n\nIn each panel, the horizontal axis represents the validity period of the forecast, while the vertical axis shows the corresponding start dates. Thus, each horizontal strip of coloured tiles shows the evolution of the mean temperature bias over the six-month period of validity of each forecast.\n\nReading each panel from bottom to top, the strips of coloured tiles are progressively shifted by one month, corresponding to the same shift in the starting date of the forecast. The horizontal axis is periodic: forecasts starting after August have their last months on the left side of the panel."} {"chunk_id": "seasonal_seasonal-monthly-single-levels_forecast-skill_q02__729c367e6806", "report_id": "seasonal_seasonal-monthly-single-levels_forecast-skill_q02", "dataset_id": "seasonal-monthly-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "forecast-skill_q02", "aspect_base": "forecast-skill", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Seasonal forecasts bias assessment for impact models > Analysis and results > 5. Plot and describe results > Discussion", "title": "Seasonal forecasts bias assessment for impact models", "chunk_index": 12, "token_count": 440, "text_raw": "The tables below summarise the monthly bias for all originating centres (columns) and for all selected areas (rows). Note that only one system (system 3) is shown from the ECCC contributions.\n\n**Temperature** - All seasonal forecast models produce a significant temperature bias over most of the regions selected for analysis. Two centres, DWD and METEOFRANCE, produce a forecast with a predominant warm bias over most regions and in all seasons. The temperature bias is a function of both the start time and the valid time of the forecast, with a slightly stronger dependence on the valid time. Therefore, the bias does not increase with the lead time of the forecast. Instead, it tends to be a characteristic aspect of each model over the specific region and time of year. The seasonal cycle of the bias is particularly evident over western North America.\n\n**Precipitation** - As in the case of temperature, the bias in precipitation tends to be more dependent on the validity period of the forecasts. The regions with the largest systematic bias are northeastern Brazil and southern Asia, where all models tend to overestimate the amplitude of the seasonal cycle.\n\n**Wind speed** - Unlike temperature and precipitation, the bias of wind speed does not show a significant seasonal cycle over most of the selected regions. The only exceptions are northeastern Brazil and southern Asia, where some models (e.g. ECCC and NCEP) show a positive bias during the months of June to September.\n\n**Dew point temperature** - As in the case of wind speed, the dew point temperature bias tends to have a weaker seasonal cycle than temperature and precipitation. However, some regions (southern Asia and western North America) show a slightly stronger seasonal dependence of the bias. NOTE: Dew point temperature is not available for ECCC.\n\nCreate a colorbar for each row", "text_with_prefix": "EQC Quality Assessment: \"Seasonal forecasts bias assessment for impact models\"\nDataset: seasonal-monthly-single-levels [CDS]\nAspect: forecast-skill_q02 | Category: Seasonal_Forecasts\nSection: Seasonal forecasts bias assessment for impact models > Analysis and results > 5. Plot and describe results > Discussion\n---\nThe tables below summarise the monthly bias for all originating centres (columns) and for all selected areas (rows). Note that only one system (system 3) is shown from the ECCC contributions.\n\n**Temperature** - All seasonal forecast models produce a significant temperature bias over most of the regions selected for analysis. Two centres, DWD and METEOFRANCE, produce a forecast with a predominant warm bias over most regions and in all seasons. The temperature bias is a function of both the start time and the valid time of the forecast, with a slightly stronger dependence on the valid time. Therefore, the bias does not increase with the lead time of the forecast. Instead, it tends to be a characteristic aspect of each model over the specific region and time of year. The seasonal cycle of the bias is particularly evident over western North America.\n\n**Precipitation** - As in the case of temperature, the bias in precipitation tends to be more dependent on the validity period of the forecasts. The regions with the largest systematic bias are northeastern Brazil and southern Asia, where all models tend to overestimate the amplitude of the seasonal cycle.\n\n**Wind speed** - Unlike temperature and precipitation, the bias of wind speed does not show a significant seasonal cycle over most of the selected regions. The only exceptions are northeastern Brazil and southern Asia, where some models (e.g. ECCC and NCEP) show a positive bias during the months of June to September.\n\n**Dew point temperature** - As in the case of wind speed, the dew point temperature bias tends to have a weaker seasonal cycle than temperature and precipitation. However, some regions (southern Asia and western North America) show a slightly stronger seasonal dependence of the bias. NOTE: Dew point temperature is not available for ECCC.\n\nCreate a colorbar for each row"} {"chunk_id": "seasonal_seasonal-monthly-single-levels_forecast-skill_q02__9addc477937f", "report_id": "seasonal_seasonal-monthly-single-levels_forecast-skill_q02", "dataset_id": "seasonal-monthly-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "forecast-skill_q02", "aspect_base": "forecast-skill", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Seasonal forecasts bias assessment for impact models > ℹ️ If you want to know more > Key resources", "title": "Seasonal forecasts bias assessment for impact models", "chunk_index": 13, "token_count": 240, "text_raw": "Some key resources and further reading were linked throughout this assessment.\n\nThe CDS catalogue entries for the data used were:\n* Seasonal forecast monthly statistics on single levels: https://cds.climate.copernicus.eu/datasets/seasonal-monthly-single-levels?tab=overview\n* ERA5 monthly averaged data on single levels from 1940 to present: https://cds.climate.copernicus.eu/datasets/reanalysis-era5-single-levels-monthly-means?tab=overview\n\nCode libraries used:\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)", "text_with_prefix": "EQC Quality Assessment: \"Seasonal forecasts bias assessment for impact models\"\nDataset: seasonal-monthly-single-levels [CDS]\nAspect: forecast-skill_q02 | Category: Seasonal_Forecasts\nSection: Seasonal forecasts bias assessment for impact models > ℹ️ If you want to know more > Key resources\n---\nSome key resources and further reading were linked throughout this assessment.\n\nThe CDS catalogue entries for the data used were:\n* Seasonal forecast monthly statistics on single levels: https://cds.climate.copernicus.eu/datasets/seasonal-monthly-single-levels?tab=overview\n* ERA5 monthly averaged data on single levels from 1940 to present: https://cds.climate.copernicus.eu/datasets/reanalysis-era5-single-levels-monthly-means?tab=overview\n\nCode libraries used:\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)"} {"chunk_id": "seasonal_seasonal-monthly-single-levels_forecast-skill_q02__e68953c290d7", "report_id": "seasonal_seasonal-monthly-single-levels_forecast-skill_q02", "dataset_id": "seasonal-monthly-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "forecast-skill_q02", "aspect_base": "forecast-skill", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Seasonal forecasts bias assessment for impact models > ℹ️ If you want to know more > References", "title": "Seasonal forecasts bias assessment for impact models", "chunk_index": 14, "token_count": 1027, "text_raw": "[[1]](https://doi.org/10.1038/s41598-018-20628-2) Rodriguez, D., De Voil, P., Hudson, D., Brown, J. N., Hayman, P., Marrou, H., & Meinke, H. (2018). Predicting optimum crop designs using crop models and seasonal climate forecasts. Scientific reports, 8(1), 2231.\n\n[[2]](https://doi.org/10.1016/j.agrformet.2018.11.029) Jha, P. K., Athanasiadis, P., Gualdi, S., Trabucco, A., Mereu, V., Shelia, V., & Hoogenboom, G. (2019). Using daily data from seasonal forecasts in dynamic crop models for yield prediction: A case study for rice in Nepal’s Terai. Agricultural and forest meteorology, 265, 349-358.\n\n[[3]](https://doi.org/10.1016/j.cliser.2022.100324) Dainelli, R., Calmanti, S., Pasqui, M., Rocchi, L., Di Giuseppe, E., Monotti, C., ... & Toscano, P. (2022). Moving climate seasonal forecasts information from useful to usable for early within-season predictions of durum wheat yield. Climate Services, 28, 100324.\n\n[[4]](https://doi.org/10.1016/j.renene.2019.04.135) Lledó, L., Torralba, V., Soret, A., Ramon, J., & Doblas-Reyes, F. J. (2019). Seasonal forecasts of wind power generation. Renewable Energy, 143, 91-100.\n\n[[5]](https://doi.org/10.1007/s00382-021-05828-3) Prodhomme, C., Materia, S., Ardilouze, C., White, R. H., Batté, L., Guemas, V., ... & García-Serrano, J. (2021). Seasonal prediction of European summer heatwaves. Climate Dynamics, 1-18.\n\n[[6]](https://doi.org/10.1175/JHM-D-19-0042.1) Li, W., Chen, J. I. E., Li, L. U., Chen, H. U. A., Liu, B., Xu, C. Y., & Li, X. (2019). Evaluation and bias correction of S2S precipitation for hydrological extremes. Journal of Hydrometeorology, 20(9), 1887-1906.\n\n[[7]](https://doi.org/10.1007/s00382-019-04640-4) Manzanas, R. (2020). Assessment of model drifts in seasonal forecasting: Sensitivity to ensemble size and implications for bias correction. Journal of Advances in Modeling Earth Systems, 12(3), e2019MS001751.\n\n[[8]](https://doi.org/10.1007/s00382-017-3962-9) Hermanson, L., Ren, H. L., Vellinga, M., Dunstone, N. D., Hyder, P., Ineson, S., ... & Williams, K. D. (2018). Different types of drifts in two seasonal forecast systems and their dependence on ENSO. Climate dynamics, 51, 1411-1426.\n\n[[9]](https://doi.org/10.1007/s00382-020-05314-2) Hemri, S., Bhend, J., Liniger, M. A., Manzanas, R., Siegert, S., Stephenson, D. B., ... & Doblas-Reyes, F. J. (2020). How to create an operational multi-model of seasonal forecasts?. Climate Dynamics, 55, 1141-1157.\n\n[[10]](https://doi.org/10.1007/s00382-019-04640-4) Manzanas, R., Gutiérrez, J. M., Bhend, J., Hemri, S., Doblas-Reyes, F. J., Torralba, V., ... & Brookshaw, A. (2019). Bias adjustment and ensemble recalibration methods for seasonal forecasting: a comprehensive intercomparison using the C3S dataset. Climate Dynamics, 53, 1287-1305.", "text_with_prefix": "EQC Quality Assessment: \"Seasonal forecasts bias assessment for impact models\"\nDataset: seasonal-monthly-single-levels [CDS]\nAspect: forecast-skill_q02 | Category: Seasonal_Forecasts\nSection: Seasonal forecasts bias assessment for impact models > ℹ️ If you want to know more > References\n---\n[[1]](https://doi.org/10.1038/s41598-018-20628-2) Rodriguez, D., De Voil, P., Hudson, D., Brown, J. N., Hayman, P., Marrou, H., & Meinke, H. (2018). Predicting optimum crop designs using crop models and seasonal climate forecasts. Scientific reports, 8(1), 2231.\n\n[[2]](https://doi.org/10.1016/j.agrformet.2018.11.029) Jha, P. K., Athanasiadis, P., Gualdi, S., Trabucco, A., Mereu, V., Shelia, V., & Hoogenboom, G. (2019). Using daily data from seasonal forecasts in dynamic crop models for yield prediction: A case study for rice in Nepal’s Terai. Agricultural and forest meteorology, 265, 349-358.\n\n[[3]](https://doi.org/10.1016/j.cliser.2022.100324) Dainelli, R., Calmanti, S., Pasqui, M., Rocchi, L., Di Giuseppe, E., Monotti, C., ... & Toscano, P. (2022). Moving climate seasonal forecasts information from useful to usable for early within-season predictions of durum wheat yield. Climate Services, 28, 100324.\n\n[[4]](https://doi.org/10.1016/j.renene.2019.04.135) Lledó, L., Torralba, V., Soret, A., Ramon, J., & Doblas-Reyes, F. J. (2019). Seasonal forecasts of wind power generation. Renewable Energy, 143, 91-100.\n\n[[5]](https://doi.org/10.1007/s00382-021-05828-3) Prodhomme, C., Materia, S., Ardilouze, C., White, R. H., Batté, L., Guemas, V., ... & García-Serrano, J. (2021). Seasonal prediction of European summer heatwaves. Climate Dynamics, 1-18.\n\n[[6]](https://doi.org/10.1175/JHM-D-19-0042.1) Li, W., Chen, J. I. E., Li, L. U., Chen, H. U. A., Liu, B., Xu, C. Y., & Li, X. (2019). Evaluation and bias correction of S2S precipitation for hydrological extremes. Journal of Hydrometeorology, 20(9), 1887-1906.\n\n[[7]](https://doi.org/10.1007/s00382-019-04640-4) Manzanas, R. (2020). Assessment of model drifts in seasonal forecasting: Sensitivity to ensemble size and implications for bias correction. Journal of Advances in Modeling Earth Systems, 12(3), e2019MS001751.\n\n[[8]](https://doi.org/10.1007/s00382-017-3962-9) Hermanson, L., Ren, H. L., Vellinga, M., Dunstone, N. D., Hyder, P., Ineson, S., ... & Williams, K. D. (2018). Different types of drifts in two seasonal forecast systems and their dependence on ENSO. Climate dynamics, 51, 1411-1426.\n\n[[9]](https://doi.org/10.1007/s00382-020-05314-2) Hemri, S., Bhend, J., Liniger, M. A., Manzanas, R., Siegert, S., Stephenson, D. B., ... & Doblas-Reyes, F. J. (2020). How to create an operational multi-model of seasonal forecasts?. Climate Dynamics, 55, 1141-1157.\n\n[[10]](https://doi.org/10.1007/s00382-019-04640-4) Manzanas, R., Gutiérrez, J. M., Bhend, J., Hemri, S., Doblas-Reyes, F. J., Torralba, V., ... & Brookshaw, A. (2019). Bias adjustment and ensemble recalibration methods for seasonal forecasting: a comprehensive intercomparison using the C3S dataset. Climate Dynamics, 53, 1287-1305."} {"chunk_id": "seasonal_seasonal-monthly-single-levels_forecast-skill_q04__77ed441a90ce", "report_id": "seasonal_seasonal-monthly-single-levels_forecast-skill_q04", "dataset_id": "seasonal-monthly-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "forecast-skill_q04", "aspect_base": "forecast-skill", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Evaluation of the hit-rate of seasonal forecasts for tercile categories of monthly anomalies", "title": "Evaluation of the hit-rate of seasonal forecasts for tercile categories of monthly anomalies", "chunk_index": 0, "token_count": 106, "text_raw": "Production date: 30/05/2024\n\nProduced by: Sandro Calmanti (ENEA), Alessandro Dell'Aquila (ENEA)", "text_with_prefix": "EQC Quality Assessment: \"Evaluation of the hit-rate of seasonal forecasts for tercile categories of monthly anomalies\"\nDataset: seasonal-monthly-single-levels [CDS]\nAspect: forecast-skill_q04 | Category: Seasonal_Forecasts\nSection: Evaluation of the hit-rate of seasonal forecasts for tercile categories of monthly anomalies\n---\nProduction date: 30/05/2024\n\nProduced by: Sandro Calmanti (ENEA), Alessandro Dell'Aquila (ENEA)"} {"chunk_id": "seasonal_seasonal-monthly-single-levels_forecast-skill_q04__f6c84f0065e1", "report_id": "seasonal_seasonal-monthly-single-levels_forecast-skill_q04", "dataset_id": "seasonal-monthly-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "forecast-skill_q04", "aspect_base": "forecast-skill", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Evaluation of the hit-rate of seasonal forecasts for tercile categories of monthly anomalies > Quality assessment questions", "title": "Evaluation of the hit-rate of seasonal forecasts for tercile categories of monthly anomalies", "chunk_index": 1, "token_count": 712, "text_raw": "* **What are the hit-rates for terciles categories in seasonal forecasts for anomalies above/below normal?** \n* **Can I use information from seasonal forecasts of monthly anomalies to make decisions?**\n\nSeasonal forecasts are not deterministic: they are not expected to provide quantitative information on the value of climate parameters with an attached error bar.\n\nInstead, **seasonal forecasts are probabilistic**: they provide information on how climate parameters are expected to change over the coming months, compared to climatology, in terms of their main statistical aspects, given the current state of the climate system.\n\nA key issue regarding the usability of seasonal forecasts is that end-users do not always have a way to incorporate a probability distribution function (PDF) related to a climate variability in their decision workflow. Many practical decisions, for which seasonal forecast might be beneficial require some form of binary choice [[1]](https://doi.org/10.1016/j.cliser.2024.100496), for example:\n\n- Do I have to plan for an anticipated harvest?\n - Do I have to choose the long-cycle or short-cycle maize variety for the coming season?\n - Will the coming season allow a reduced number of treatments for my crops?\n - Do I have to purchase in advance, and at a lower price, an extra stock of dispatchable energy for air conditioning?\n\nThis notebook focuses on the **hit-rate** for terciles categories in seasonal forecasts, a simple yet fundamental building block for measuring the quality, and potential value, of seasonal forecasts.\n\nThe hit-rate (also mentioned sometimes as _hit-score_) addresses the key question: how often did the forecasted category with the highest probability actually occur? As such, the hit-rate for terciles categories describes a basic attribute of probabilistic forecasts, i.e. that the outcome of a given season is expected to differ from the normal if the forecast differs from the normal. This characteristic of probabilistic forecasts is called **resolution**.\n\nA complementary aspect in describing the quality of seasonal forecasts is **discrimination**, which is the answer to the opposite question: do the forecasts differ when the outcome differ? As an example if, on average, a forecasting system issues always the same forecast, regardless of the expected outcome, there can be some _hits_ in the forecast but, at the same time, the forecast would not be able to discriminate the occurrence of an event from a non-event. Discrimination is best measured by **Relative Operating Characteristic (ROC)** scores.\n\nAs long as there is some resolution or discrimination, there is a potential to extract some valuable information from the forecasts.\n\nIndeed, the WMO [Guidance on Verification of Operational Seasonal Climate Forecasts](https://library.wmo.int/idurl/4/56227) highlights how the use of generalised methodologies for the assessment of seasonal forecasts sometimes represent limited attraction for new users. Therefore, the [Guidance](https://library.wmo.int/idurl/4/56227) recommends a range of methodologies of increasing complexity, including very simple ones like the hit-rate, which are particularly useful for introducing the potential value of seasonal forecasts to non-specialists.", "text_with_prefix": "EQC Quality Assessment: \"Evaluation of the hit-rate of seasonal forecasts for tercile categories of monthly anomalies\"\nDataset: seasonal-monthly-single-levels [CDS]\nAspect: forecast-skill_q04 | Category: Seasonal_Forecasts\nSection: Evaluation of the hit-rate of seasonal forecasts for tercile categories of monthly anomalies > Quality assessment questions\n---\n* **What are the hit-rates for terciles categories in seasonal forecasts for anomalies above/below normal?** \n* **Can I use information from seasonal forecasts of monthly anomalies to make decisions?**\n\nSeasonal forecasts are not deterministic: they are not expected to provide quantitative information on the value of climate parameters with an attached error bar.\n\nInstead, **seasonal forecasts are probabilistic**: they provide information on how climate parameters are expected to change over the coming months, compared to climatology, in terms of their main statistical aspects, given the current state of the climate system.\n\nA key issue regarding the usability of seasonal forecasts is that end-users do not always have a way to incorporate a probability distribution function (PDF) related to a climate variability in their decision workflow. Many practical decisions, for which seasonal forecast might be beneficial require some form of binary choice [[1]](https://doi.org/10.1016/j.cliser.2024.100496), for example:\n\n- Do I have to plan for an anticipated harvest?\n - Do I have to choose the long-cycle or short-cycle maize variety for the coming season?\n - Will the coming season allow a reduced number of treatments for my crops?\n - Do I have to purchase in advance, and at a lower price, an extra stock of dispatchable energy for air conditioning?\n\nThis notebook focuses on the **hit-rate** for terciles categories in seasonal forecasts, a simple yet fundamental building block for measuring the quality, and potential value, of seasonal forecasts.\n\nThe hit-rate (also mentioned sometimes as _hit-score_) addresses the key question: how often did the forecasted category with the highest probability actually occur? As such, the hit-rate for terciles categories describes a basic attribute of probabilistic forecasts, i.e. that the outcome of a given season is expected to differ from the normal if the forecast differs from the normal. This characteristic of probabilistic forecasts is called **resolution**.\n\nA complementary aspect in describing the quality of seasonal forecasts is **discrimination**, which is the answer to the opposite question: do the forecasts differ when the outcome differ? As an example if, on average, a forecasting system issues always the same forecast, regardless of the expected outcome, there can be some _hits_ in the forecast but, at the same time, the forecast would not be able to discriminate the occurrence of an event from a non-event. Discrimination is best measured by **Relative Operating Characteristic (ROC)** scores.\n\nAs long as there is some resolution or discrimination, there is a potential to extract some valuable information from the forecasts.\n\nIndeed, the WMO [Guidance on Verification of Operational Seasonal Climate Forecasts](https://library.wmo.int/idurl/4/56227) highlights how the use of generalised methodologies for the assessment of seasonal forecasts sometimes represent limited attraction for new users. Therefore, the [Guidance](https://library.wmo.int/idurl/4/56227) recommends a range of methodologies of increasing complexity, including very simple ones like the hit-rate, which are particularly useful for introducing the potential value of seasonal forecasts to non-specialists."} {"chunk_id": "seasonal_seasonal-monthly-single-levels_forecast-skill_q04__c42369c791f2", "report_id": "seasonal_seasonal-monthly-single-levels_forecast-skill_q04", "dataset_id": "seasonal-monthly-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "forecast-skill_q04", "aspect_base": "forecast-skill", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Evaluation of the hit-rate of seasonal forecasts for tercile categories of monthly anomalies > Quality assessment statement", "title": "Evaluation of the hit-rate of seasonal forecasts for tercile categories of monthly anomalies", "chunk_index": 2, "token_count": 347, "text_raw": "These are the key outcomes of this assessment\n\n* The hit-rate for terciles cetegories of seasonal forecasts in predicting anomalies at a monthly time-scale depends significantly on the variable and on the region of interest. \n* The hit-rates for terciles of different variables do not depend significantly on the forecasting system. \n* The larger hit-rates for terciles of monthly anomalies (above 70%) are for 2m temperatures in regions surrounding the Pacific, under the direct influence of El Niño [[2]](https://doi.org/10.1038/s41612-023-00519-8).\n* The usability and value of seasonal forecasts must be assessed on a case-by-case basis, by explicitly accounting for the impact on the decision at stake [[3]](https://doi.org/10.1002/wcc.523).\n```\n\nattachment:a684544d-7232-4092-82a5-7ec01184a99a.png\n---\nheight: 500px\n---\nThe figure shows a sample of the hit-rate for terciles of ECMWF monthly anomalies for 2m temperature over South East Asia as a function of the starting month and of the valid month of the forecast. More details on how to interpret the results are available below in the notebook.\n```", "text_with_prefix": "EQC Quality Assessment: \"Evaluation of the hit-rate of seasonal forecasts for tercile categories of monthly anomalies\"\nDataset: seasonal-monthly-single-levels [CDS]\nAspect: forecast-skill_q04 | Category: Seasonal_Forecasts\nSection: Evaluation of the hit-rate of seasonal forecasts for tercile categories of monthly anomalies > Quality assessment statement\n---\nThese are the key outcomes of this assessment\n\n* The hit-rate for terciles cetegories of seasonal forecasts in predicting anomalies at a monthly time-scale depends significantly on the variable and on the region of interest. \n* The hit-rates for terciles of different variables do not depend significantly on the forecasting system. \n* The larger hit-rates for terciles of monthly anomalies (above 70%) are for 2m temperatures in regions surrounding the Pacific, under the direct influence of El Niño [[2]](https://doi.org/10.1038/s41612-023-00519-8).\n* The usability and value of seasonal forecasts must be assessed on a case-by-case basis, by explicitly accounting for the impact on the decision at stake [[3]](https://doi.org/10.1002/wcc.523).\n```\n\nattachment:a684544d-7232-4092-82a5-7ec01184a99a.png\n---\nheight: 500px\n---\nThe figure shows a sample of the hit-rate for terciles of ECMWF monthly anomalies for 2m temperature over South East Asia as a function of the starting month and of the valid month of the forecast. More details on how to interpret the results are available below in the notebook.\n```"} {"chunk_id": "seasonal_seasonal-monthly-single-levels_forecast-skill_q04__eb5339822df9", "report_id": "seasonal_seasonal-monthly-single-levels_forecast-skill_q04", "dataset_id": "seasonal-monthly-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "forecast-skill_q04", "aspect_base": "forecast-skill", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Evaluation of the hit-rate of seasonal forecasts for tercile categories of monthly anomalies > Methodology", "title": "Evaluation of the hit-rate of seasonal forecasts for tercile categories of monthly anomalies", "chunk_index": 3, "token_count": 381, "text_raw": "This notebook provides a comprehensive assessment of the hit-rates for terciles on monthly anomalies of selected ECVs of the forecast systems by comparing the model predictions with the ERA5 reanalysis. The analysis is carried out for eight regions selected from those used in the IPCC-AR6 (see figure below) [[4]](https://doi.org/10.1017/CBO9781139177245.006).\n\nThe computation of the hit-score is straightforward: the monthly values of the selected ECV (e.g. temperature) are firstly divided in terciles (below normal, normal, above normal). Then, the hit-score is computed by counting the number of events when the most populated tercile derived from the distribution of the seasonal forecasts corresponds to the observed tercile of the target month.\n\nThe analysis and results are organised in the following steps, which are detailed in the sections below:\n\n**[](section-1):**\n * Import required packages\n * Define parameters describing regions, variables, and data\n * Plot the chosen regions\n * Define functions for the tercile analysis\n * Define plot function\n\n**[](section-2):**\n * Download and process ERA5\n * Download and process seasonal forecast data\n\n**[](section-3)**\n * Analysis of terciles\n * Compute hit-score and show a sample plot\n\n**[](section-4):** \n * Hit-rate for terciles of 2m temperature \n * Hit-rate for terciles of other variables", "text_with_prefix": "EQC Quality Assessment: \"Evaluation of the hit-rate of seasonal forecasts for tercile categories of monthly anomalies\"\nDataset: seasonal-monthly-single-levels [CDS]\nAspect: forecast-skill_q04 | Category: Seasonal_Forecasts\nSection: Evaluation of the hit-rate of seasonal forecasts for tercile categories of monthly anomalies > Methodology\n---\nThis notebook provides a comprehensive assessment of the hit-rates for terciles on monthly anomalies of selected ECVs of the forecast systems by comparing the model predictions with the ERA5 reanalysis. The analysis is carried out for eight regions selected from those used in the IPCC-AR6 (see figure below) [[4]](https://doi.org/10.1017/CBO9781139177245.006).\n\nThe computation of the hit-score is straightforward: the monthly values of the selected ECV (e.g. temperature) are firstly divided in terciles (below normal, normal, above normal). Then, the hit-score is computed by counting the number of events when the most populated tercile derived from the distribution of the seasonal forecasts corresponds to the observed tercile of the target month.\n\nThe analysis and results are organised in the following steps, which are detailed in the sections below:\n\n**[](section-1):**\n * Import required packages\n * Define parameters describing regions, variables, and data\n * Plot the chosen regions\n * Define functions for the tercile analysis\n * Define plot function\n\n**[](section-2):**\n * Download and process ERA5\n * Download and process seasonal forecast data\n\n**[](section-3)**\n * Analysis of terciles\n * Compute hit-score and show a sample plot\n\n**[](section-4):** \n * Hit-rate for terciles of 2m temperature \n * Hit-rate for terciles of other variables"} {"chunk_id": "seasonal_seasonal-monthly-single-levels_forecast-skill_q04__00f552f3dcff", "report_id": "seasonal_seasonal-monthly-single-levels_forecast-skill_q04", "dataset_id": "seasonal-monthly-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "forecast-skill_q04", "aspect_base": "forecast-skill", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Evaluation of the hit-rate of seasonal forecasts for tercile categories of monthly anomalies > Analysis and results > 1. Choose the data to use and setup the code > Define parameters describing regions, variables, and data", "title": "Evaluation of the hit-rate of seasonal forecasts for tercile categories of monthly anomalies", "chunk_index": 4, "token_count": 270, "text_raw": "The analysis of hit-scores is conducted over 5 different variables for the 8 originating centres. The AR6 regions selected for the analysis are.\n\n- NEAF: Northern-East Africa\n- ENA: Eastern North America\n- MED: Southern Europe/Mediterranean\n- NES: Eastern Coast of South America\n- SAS: Southern Asia\n- SEA: South-Eastern Asia\n- WNA: Western North America\n- NWS: Western Coast of South America\n\nA `common_request` is defined to cover the entire global domain and the entire hindcast period from 1993 to 2016. Specific data requests are then defined for the reanalysis and for the seasonal forecasts, so that all available months and lead times are retrieved for the analysis.\n\nDetrend timeseries\nDefine centres with missing variables\nDefine data requests", "text_with_prefix": "EQC Quality Assessment: \"Evaluation of the hit-rate of seasonal forecasts for tercile categories of monthly anomalies\"\nDataset: seasonal-monthly-single-levels [CDS]\nAspect: forecast-skill_q04 | Category: Seasonal_Forecasts\nSection: Evaluation of the hit-rate of seasonal forecasts for tercile categories of monthly anomalies > Analysis and results > 1. Choose the data to use and setup the code > Define parameters describing regions, variables, and data\n---\nThe analysis of hit-scores is conducted over 5 different variables for the 8 originating centres. The AR6 regions selected for the analysis are.\n\n- NEAF: Northern-East Africa\n- ENA: Eastern North America\n- MED: Southern Europe/Mediterranean\n- NES: Eastern Coast of South America\n- SAS: Southern Asia\n- SEA: South-Eastern Asia\n- WNA: Western North America\n- NWS: Western Coast of South America\n\nA `common_request` is defined to cover the entire global domain and the entire hindcast period from 1993 to 2016. Specific data requests are then defined for the reanalysis and for the seasonal forecasts, so that all available months and lead times are retrieved for the analysis.\n\nDetrend timeseries\nDefine centres with missing variables\nDefine data requests"} {"chunk_id": "seasonal_seasonal-monthly-single-levels_forecast-skill_q04__7b71c5040181", "report_id": "seasonal_seasonal-monthly-single-levels_forecast-skill_q04", "dataset_id": "seasonal-monthly-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "forecast-skill_q04", "aspect_base": "forecast-skill", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Evaluation of the hit-rate of seasonal forecasts for tercile categories of monthly anomalies > Analysis and results > 1. Choose the data to use and setup the code > Plot the chosen regions", "title": "Evaluation of the hit-rate of seasonal forecasts for tercile categories of monthly anomalies", "chunk_index": 5, "token_count": 161, "text_raw": "For the bias analysis we use 8 out of the 46 AR6 regions defined in the `regionmask` package in order to assess the systematic errors of the modelling systems under different climatic regimes (tropical, extratropical, continental, maritime) with a number of cases manageable within the notebook.", "text_with_prefix": "EQC Quality Assessment: \"Evaluation of the hit-rate of seasonal forecasts for tercile categories of monthly anomalies\"\nDataset: seasonal-monthly-single-levels [CDS]\nAspect: forecast-skill_q04 | Category: Seasonal_Forecasts\nSection: Evaluation of the hit-rate of seasonal forecasts for tercile categories of monthly anomalies > Analysis and results > 1. Choose the data to use and setup the code > Plot the chosen regions\n---\nFor the bias analysis we use 8 out of the 46 AR6 regions defined in the `regionmask` package in order to assess the systematic errors of the modelling systems under different climatic regimes (tropical, extratropical, continental, maritime) with a number of cases manageable within the notebook."} {"chunk_id": "seasonal_seasonal-monthly-single-levels_forecast-skill_q04__dff7d3d5ce8f", "report_id": "seasonal_seasonal-monthly-single-levels_forecast-skill_q04", "dataset_id": "seasonal-monthly-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "forecast-skill_q04", "aspect_base": "forecast-skill", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Evaluation of the hit-rate of seasonal forecasts for tercile categories of monthly anomalies > Analysis and results > 1. Choose the data to use and setup the code > Define the required functions", "title": "Evaluation of the hit-rate of seasonal forecasts for tercile categories of monthly anomalies", "chunk_index": 6, "token_count": 725, "text_raw": "- The function `reindex_seasonal_forecast` is necessary in the tercile analysis to align the valid time of the forecast to the corresponding date in the reanalysis.\n\n- The function `tercile_analysis` converts the value of both ERA5 data and seasonal forecast data into the corresponding tercile value, which corresponds to below normal, normal, and above normal. The processing of ERA5 and of seasonal forecasts is detailed below.\n\n**Processing of ERA5**\n\n1.\tCompute the mean monthly temperature from the reanalysis over the selected AR6 regions.\n\n2. If required, remove the linear trend to emphasise hit-scores associated to natural climate variability rather than to long-term tendencies. This step may be of interest for the analysis of hit-scores for temperature. By default, detrending is not applied in the analysis presented in this notebook. However, the code cell shows how to implement the detrending procedure.\n\n3.\tFor each month, and for each region, compute the values associated to the 33-th and 66-th percentile of the corresponding monthly distribution\n\n4.\tConvert the series of mean monthly temperature anomalies over each region into the corresponding terciles, using the following conversion table:\n\n- -1 = below normal, values below the 33-th percentile\n - 0 , values between the 33-th and the 66-th percentile\n - 1 = above normal, values above the 33-th percentile\n\n**Processing of seasonal forecasts**\n\n1.\tApply step 1. as for ERA5 for each seasonal forecasting system. In this case there is a different temperature anomaly for each originating centre, starting time and valid time for the forecast.\n\n2. If required, apply step 2. as for ERA5 to remove linear trends.\n\n3. Apply step 3. as for ERA5 to obtain the position of the terciles for each originating centre and for each starting time and valid time. In this case, the underlying distribution for the computation of the terciles includes all years in the hindcast as well as all ensemble members. The 33-th and 66-th percentiles depend on the lead-time, i.e. the difference between valid time and starting time, so that the analysis accounts for potential model drifts. In the present analysis, we consider a maximum lead-time of six months.\n\n4. Apply step 4. as for ERA5 to convert temperature anomalies into the corresponding terciles for all regions, for all time months and valid time, and for each member of the ensemble forecast.\n\n5. For seasonal forecasts the most likely tercile in the forecast is derived by computing the mode of the terciles (step 3) over all the ensemble members. The most likely terciles is an array whose dimensions are the originating centres, the region, the starting time and the valid time.\n\nStack starting_time and leading_month\nShift valid_time\nReindex: valid_time and starting_month\nSpatial mean\nDetrend timeseries\nCompute anomaly\nReindex using valid year/month month\nGet quantiles\n\n(section-2)=", "text_with_prefix": "EQC Quality Assessment: \"Evaluation of the hit-rate of seasonal forecasts for tercile categories of monthly anomalies\"\nDataset: seasonal-monthly-single-levels [CDS]\nAspect: forecast-skill_q04 | Category: Seasonal_Forecasts\nSection: Evaluation of the hit-rate of seasonal forecasts for tercile categories of monthly anomalies > Analysis and results > 1. Choose the data to use and setup the code > Define the required functions\n---\n- The function `reindex_seasonal_forecast` is necessary in the tercile analysis to align the valid time of the forecast to the corresponding date in the reanalysis.\n\n- The function `tercile_analysis` converts the value of both ERA5 data and seasonal forecast data into the corresponding tercile value, which corresponds to below normal, normal, and above normal. The processing of ERA5 and of seasonal forecasts is detailed below.\n\n**Processing of ERA5**\n\n1.\tCompute the mean monthly temperature from the reanalysis over the selected AR6 regions.\n\n2. If required, remove the linear trend to emphasise hit-scores associated to natural climate variability rather than to long-term tendencies. This step may be of interest for the analysis of hit-scores for temperature. By default, detrending is not applied in the analysis presented in this notebook. However, the code cell shows how to implement the detrending procedure.\n\n3.\tFor each month, and for each region, compute the values associated to the 33-th and 66-th percentile of the corresponding monthly distribution\n\n4.\tConvert the series of mean monthly temperature anomalies over each region into the corresponding terciles, using the following conversion table:\n\n- -1 = below normal, values below the 33-th percentile\n - 0 , values between the 33-th and the 66-th percentile\n - 1 = above normal, values above the 33-th percentile\n\n**Processing of seasonal forecasts**\n\n1.\tApply step 1. as for ERA5 for each seasonal forecasting system. In this case there is a different temperature anomaly for each originating centre, starting time and valid time for the forecast.\n\n2. If required, apply step 2. as for ERA5 to remove linear trends.\n\n3. Apply step 3. as for ERA5 to obtain the position of the terciles for each originating centre and for each starting time and valid time. In this case, the underlying distribution for the computation of the terciles includes all years in the hindcast as well as all ensemble members. The 33-th and 66-th percentiles depend on the lead-time, i.e. the difference between valid time and starting time, so that the analysis accounts for potential model drifts. In the present analysis, we consider a maximum lead-time of six months.\n\n4. Apply step 4. as for ERA5 to convert temperature anomalies into the corresponding terciles for all regions, for all time months and valid time, and for each member of the ensemble forecast.\n\n5. For seasonal forecasts the most likely tercile in the forecast is derived by computing the mode of the terciles (step 3) over all the ensemble members. The most likely terciles is an array whose dimensions are the originating centres, the region, the starting time and the valid time.\n\nStack starting_time and leading_month\nShift valid_time\nReindex: valid_time and starting_month\nSpatial mean\nDetrend timeseries\nCompute anomaly\nReindex using valid year/month month\nGet quantiles\n\n(section-2)="} {"chunk_id": "seasonal_seasonal-monthly-single-levels_forecast-skill_q04__84df13e01a1b", "report_id": "seasonal_seasonal-monthly-single-levels_forecast-skill_q04", "dataset_id": "seasonal-monthly-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "forecast-skill_q04", "aspect_base": "forecast-skill", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Evaluation of the hit-rate of seasonal forecasts for tercile categories of monthly anomalies > Analysis and results > 2. Download and process data > Download and process ERA5", "title": "Evaluation of the hit-rate of seasonal forecasts for tercile categories of monthly anomalies", "chunk_index": 7, "token_count": 263, "text_raw": "ERA5 data are converted to tercile values by applying the function `tercile_analysis`. The outcome of the download and transform process is a three-dimensional array containing tercile values, i.e. (-1, 0, 1) for (\"_Below normal_, _Normal_, _Above Normal_\") respectively, for each of the 8 regions, for 25 years and 12 months each year.\n\nThe final outcome of the tercile analysis for each variable is an array with the following dimensions:\n\n> - valid_year: 25\n > - valid_month: 12\n > - region: 8\n\nA separate array is created for each of the 5 selected variables. All arrays are assembled into a single `xarray` dataset for the following computation and visualisations steps.\n\nGet the reanalysis data", "text_with_prefix": "EQC Quality Assessment: \"Evaluation of the hit-rate of seasonal forecasts for tercile categories of monthly anomalies\"\nDataset: seasonal-monthly-single-levels [CDS]\nAspect: forecast-skill_q04 | Category: Seasonal_Forecasts\nSection: Evaluation of the hit-rate of seasonal forecasts for tercile categories of monthly anomalies > Analysis and results > 2. Download and process data > Download and process ERA5\n---\nERA5 data are converted to tercile values by applying the function `tercile_analysis`. The outcome of the download and transform process is a three-dimensional array containing tercile values, i.e. (-1, 0, 1) for (\"_Below normal_, _Normal_, _Above Normal_\") respectively, for each of the 8 regions, for 25 years and 12 months each year.\n\nThe final outcome of the tercile analysis for each variable is an array with the following dimensions:\n\n> - valid_year: 25\n > - valid_month: 12\n > - region: 8\n\nA separate array is created for each of the 5 selected variables. All arrays are assembled into a single `xarray` dataset for the following computation and visualisations steps.\n\nGet the reanalysis data"} {"chunk_id": "seasonal_seasonal-monthly-single-levels_forecast-skill_q04__8ad4313a4ee8", "report_id": "seasonal_seasonal-monthly-single-levels_forecast-skill_q04", "dataset_id": "seasonal-monthly-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "forecast-skill_q04", "aspect_base": "forecast-skill", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Evaluation of the hit-rate of seasonal forecasts for tercile categories of monthly anomalies > Analysis and results > 2. Download and process data > Download and process seasonal forecasts", "title": "Evaluation of the hit-rate of seasonal forecasts for tercile categories of monthly anomalies", "chunk_index": 8, "token_count": 330, "text_raw": "Similarly to ERA5 data, seasonal forecast data are converted to tercile values by applying the function `tercile_analysis`. However, in this case the outcome of the download and transform process is a five-dimensional array: besides the 8 regions, 25 years and 12 months corresponding to the valid time of the forecasts, two additional dimensions are necessary to account for the 8 originating centres and for the 12 possible starting months of the forecast. For each starting month, there are 6 tercile values that correspond to the length of the forecast taken into account, whereas the tercile values for the other six months are set to Not-a-Number.\n\nThe final outcome of the tercile analysis for each variable is an array with the following dimensions:\n\n> - valid_year: 25\n > - valid_month: 12\n > - starting_month: 12\n > - region: 8\n > - centre: 8\n\nAll arrays are assembled into a single `xarray` dataset containing all variables for the following computation and visualisations steps.\n\nGet the seasonal forecast data - DETREND\n\n(section-3)=", "text_with_prefix": "EQC Quality Assessment: \"Evaluation of the hit-rate of seasonal forecasts for tercile categories of monthly anomalies\"\nDataset: seasonal-monthly-single-levels [CDS]\nAspect: forecast-skill_q04 | Category: Seasonal_Forecasts\nSection: Evaluation of the hit-rate of seasonal forecasts for tercile categories of monthly anomalies > Analysis and results > 2. Download and process data > Download and process seasonal forecasts\n---\nSimilarly to ERA5 data, seasonal forecast data are converted to tercile values by applying the function `tercile_analysis`. However, in this case the outcome of the download and transform process is a five-dimensional array: besides the 8 regions, 25 years and 12 months corresponding to the valid time of the forecasts, two additional dimensions are necessary to account for the 8 originating centres and for the 12 possible starting months of the forecast. For each starting month, there are 6 tercile values that correspond to the length of the forecast taken into account, whereas the tercile values for the other six months are set to Not-a-Number.\n\nThe final outcome of the tercile analysis for each variable is an array with the following dimensions:\n\n> - valid_year: 25\n > - valid_month: 12\n > - starting_month: 12\n > - region: 8\n > - centre: 8\n\nAll arrays are assembled into a single `xarray` dataset containing all variables for the following computation and visualisations steps.\n\nGet the seasonal forecast data - DETREND\n\n(section-3)="} {"chunk_id": "seasonal_seasonal-monthly-single-levels_forecast-skill_q04__a0bc81c7bea9", "report_id": "seasonal_seasonal-monthly-single-levels_forecast-skill_q04", "dataset_id": "seasonal-monthly-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "forecast-skill_q04", "aspect_base": "forecast-skill", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Evaluation of the hit-rate of seasonal forecasts for tercile categories of monthly anomalies > Analysis and results > 3. Hit-score analysis > Analysis of terciles", "title": "Evaluation of the hit-rate of seasonal forecasts for tercile categories of monthly anomalies", "chunk_index": 9, "token_count": 671, "text_raw": "Splitting the distribution of climate parameters into terciles is a simple way of converting probabilistic forecasts into binary information that a user of climate information can consult to make decisions.\n\nFor example, a user may ask whether the coming months will be warmer (or colder) than normal, and the answer would be \"likely yes\" or \"likely no\" depending on which is the most likely anomaly in the forecast. As a working hypothesis, it is reasonable to assume that specific actions are undertaken when a significant deviation from the norm is expected, for example planting a drought resistant but less productive maize variety if less rainfall than normal is expected, or purchasing in advance, and at a lower price, an extra stock of dispatchable energy for air conditioning systems if warmer than normal conditions are expected during summer.\n\nFor the analysis presented in this notebook, the forecasting system scores a hit if the observed monthly anomaly matches the corresponding forecast.\n\nTo illustrate how the hit-score analysis works, the three figures below show an example mean 2m temperature over South East Asia, using ECMWF System 5.1 as a sample seasonal forecast.\n\n- The first figure shows a complete series of the terciles derived from ERA5 as a function of year and month. For example, the year 1998, during which a strong El Niño event has been registered, is characterised by a sequence of temperature anomalies above normal, starting from December 1997 to October 1998. Instead, the year 2008, a strong La Niña year, shows a sequence of below normal anomalies.\n\n- The second figure shows, read from below to the top, the tercile analysis for the seasonal forecasts issued during 1998. Each row in the panel corresponds to a single six-month forecast issued at the corresponding starting month on the vertical axis. For the year 1998 most of the forecast starts with a temperature anomaly above normal and persists in this status until the end of the six-month period. Forecasts starting at the end of the previous year, which are represented in the top-left corner of the panel, also start with a warm anomaly that persists throughout the forecast. Since 1998 was a warm year in this region, most of the monthly anomalies of 1998 count as a hit for the forecasts issued in 1998 at all lead times.\n\nplt.suptitle(\"ERA5\")\n\n```text\n/data/common/miniforge3/envs/wp3/lib/python3.11/site-packages/dask/core.py:127: RuntimeWarning: divide by zero encountered in divide\n return func(*(_execute_task(a, cache) for a in args))\n/data/common/miniforge3/envs/wp3/lib/python3.11/site-packages/dask/core.py:127: RuntimeWarning: invalid value encountered in multiply\n return func(*(_execute_task(a, cache) for a in args))\n```", "text_with_prefix": "EQC Quality Assessment: \"Evaluation of the hit-rate of seasonal forecasts for tercile categories of monthly anomalies\"\nDataset: seasonal-monthly-single-levels [CDS]\nAspect: forecast-skill_q04 | Category: Seasonal_Forecasts\nSection: Evaluation of the hit-rate of seasonal forecasts for tercile categories of monthly anomalies > Analysis and results > 3. Hit-score analysis > Analysis of terciles\n---\nSplitting the distribution of climate parameters into terciles is a simple way of converting probabilistic forecasts into binary information that a user of climate information can consult to make decisions.\n\nFor example, a user may ask whether the coming months will be warmer (or colder) than normal, and the answer would be \"likely yes\" or \"likely no\" depending on which is the most likely anomaly in the forecast. As a working hypothesis, it is reasonable to assume that specific actions are undertaken when a significant deviation from the norm is expected, for example planting a drought resistant but less productive maize variety if less rainfall than normal is expected, or purchasing in advance, and at a lower price, an extra stock of dispatchable energy for air conditioning systems if warmer than normal conditions are expected during summer.\n\nFor the analysis presented in this notebook, the forecasting system scores a hit if the observed monthly anomaly matches the corresponding forecast.\n\nTo illustrate how the hit-score analysis works, the three figures below show an example mean 2m temperature over South East Asia, using ECMWF System 5.1 as a sample seasonal forecast.\n\n- The first figure shows a complete series of the terciles derived from ERA5 as a function of year and month. For example, the year 1998, during which a strong El Niño event has been registered, is characterised by a sequence of temperature anomalies above normal, starting from December 1997 to October 1998. Instead, the year 2008, a strong La Niña year, shows a sequence of below normal anomalies.\n\n- The second figure shows, read from below to the top, the tercile analysis for the seasonal forecasts issued during 1998. Each row in the panel corresponds to a single six-month forecast issued at the corresponding starting month on the vertical axis. For the year 1998 most of the forecast starts with a temperature anomaly above normal and persists in this status until the end of the six-month period. Forecasts starting at the end of the previous year, which are represented in the top-left corner of the panel, also start with a warm anomaly that persists throughout the forecast. Since 1998 was a warm year in this region, most of the monthly anomalies of 1998 count as a hit for the forecasts issued in 1998 at all lead times.\n\nplt.suptitle(\"ERA5\")\n\n```text\n/data/common/miniforge3/envs/wp3/lib/python3.11/site-packages/dask/core.py:127: RuntimeWarning: divide by zero encountered in divide\n return func(*(_execute_task(a, cache) for a in args))\n/data/common/miniforge3/envs/wp3/lib/python3.11/site-packages/dask/core.py:127: RuntimeWarning: invalid value encountered in multiply\n return func(*(_execute_task(a, cache) for a in args))\n```"} {"chunk_id": "seasonal_seasonal-monthly-single-levels_forecast-skill_q04__2a892470e2fc", "report_id": "seasonal_seasonal-monthly-single-levels_forecast-skill_q04", "dataset_id": "seasonal-monthly-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "forecast-skill_q04", "aspect_base": "forecast-skill", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Evaluation of the hit-rate of seasonal forecasts for tercile categories of monthly anomalies > Analysis and results > 3. Hit-score analysis > Compute hit-score and show a sample plot", "title": "Evaluation of the hit-rate of seasonal forecasts for tercile categories of monthly anomalies", "chunk_index": 10, "token_count": 447, "text_raw": "By counting the number of hits for all available hindcasts and considering the hits on anomalies both above and below normal, it is possible to obtain a systematic assessment of the hit-rates for terciles on monthly anomalies. Similar hit-scores may also be considered, but for the categories with second- and third- highest probabilities. Other adjustments to the scores may be considered, as suggested by the [WMO Guidance](https://library.wmo.int/idurl/4/56227), by assigning for example a half hit if one of two categories with tied highest probability verifies.\n\nIn this assessment we consider the simplest possible definition:\n\n$Hit \\: rate = 100 * \\frac{Number \\: of \\: hits}{Number \\: of \\: forecasts}$\n\nThe figure below, which should be read from the bottom to the top, shows in each row the hit rate on monthly anomalies of the forecast starting from the corresponding month indicated on the vertical axis.\n\nFor example, the forecasts starting in June (month 6) have a hit-rate above 90% for the first month of the forecast, which is reduced to a value between 80% and 90% for the second month (July). During the following months (August to November) the hit-rate decreases further but remains mostly above 70%.\n\nNote that the horizontal axis is periodic and the tiles in the top left corner of the panel represent the hit-rates for terciles of the forecasts started in August or later during the year.\n\nCompute hit rates\nDefine the custom colormap\nThis adjustment allow maz hit-rate values to be correctly represented with the color palette\nhit_rate_mask = xr.where(hit_rate == 100., 99.9, hit_rate)\n\n(section-4)=", "text_with_prefix": "EQC Quality Assessment: \"Evaluation of the hit-rate of seasonal forecasts for tercile categories of monthly anomalies\"\nDataset: seasonal-monthly-single-levels [CDS]\nAspect: forecast-skill_q04 | Category: Seasonal_Forecasts\nSection: Evaluation of the hit-rate of seasonal forecasts for tercile categories of monthly anomalies > Analysis and results > 3. Hit-score analysis > Compute hit-score and show a sample plot\n---\nBy counting the number of hits for all available hindcasts and considering the hits on anomalies both above and below normal, it is possible to obtain a systematic assessment of the hit-rates for terciles on monthly anomalies. Similar hit-scores may also be considered, but for the categories with second- and third- highest probabilities. Other adjustments to the scores may be considered, as suggested by the [WMO Guidance](https://library.wmo.int/idurl/4/56227), by assigning for example a half hit if one of two categories with tied highest probability verifies.\n\nIn this assessment we consider the simplest possible definition:\n\n$Hit \\: rate = 100 * \\frac{Number \\: of \\: hits}{Number \\: of \\: forecasts}$\n\nThe figure below, which should be read from the bottom to the top, shows in each row the hit rate on monthly anomalies of the forecast starting from the corresponding month indicated on the vertical axis.\n\nFor example, the forecasts starting in June (month 6) have a hit-rate above 90% for the first month of the forecast, which is reduced to a value between 80% and 90% for the second month (July). During the following months (August to November) the hit-rate decreases further but remains mostly above 70%.\n\nNote that the horizontal axis is periodic and the tiles in the top left corner of the panel represent the hit-rates for terciles of the forecasts started in August or later during the year.\n\nCompute hit rates\nDefine the custom colormap\nThis adjustment allow maz hit-rate values to be correctly represented with the color palette\nhit_rate_mask = xr.where(hit_rate == 100., 99.9, hit_rate)\n\n(section-4)="} {"chunk_id": "seasonal_seasonal-monthly-single-levels_forecast-skill_q04__73f3fd6c11ac", "report_id": "seasonal_seasonal-monthly-single-levels_forecast-skill_q04", "dataset_id": "seasonal-monthly-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "forecast-skill_q04", "aspect_base": "forecast-skill", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Evaluation of the hit-rate of seasonal forecasts for tercile categories of monthly anomalies > Analysis and results > 4. Plot and describe results > Hit-rate for 2 m temperature", "title": "Evaluation of the hit-rate of seasonal forecasts for tercile categories of monthly anomalies", "chunk_index": 11, "token_count": 677, "text_raw": "The same analysis described in section 3 above. can be replicated for all forecasting systems and for all the selected sample regions.\n\nOverall, the seasonal forecasts of 2m temperature have a significant, and potentially usable, hit-rate for terciles in regions closer to the tropics and under the direct influence of El Niño. The further away from the tropics, the lower the expected hit-rate on monthly anomalies.\n\nHit-rates smaller than 50% have been masked out, with the assumption that lower hit-rates would be of limited use for most practical applications. Although this working hypothesis can not be assumed to be strictly true for all purposes, it is a useful assumption for a discussion on the general characteristics of seasonal forecasts [[3]](https://doi.org/10.1002/wcc.523) [[5]](https://doi.org/10.1016/j.crm.2021.100375) [[6]](https://doi.org/10.1016/j.cliser.2023.100347) [[7]](https://doi.org/10.1016/j.cliser.2023.100346).\n\nThe figure below shows that, among the sample regions considered in this analysis, those with the higher and more persistent hit-rate are South East Asia (SEA) and North West South America (NWS). These areas are under the direct influence of ENSO and are expected to benefit from a significantly higher degree of predictability compared to other regions. Other areas with significant, and potentially usable, in practical applications are Northern-East Africa (NEAF) and (North East South America) NES. Other areas like Southern Asia (SAS) and Souther Europe/Mediterranean (MED) have a moderate hit-rate, limited mostly to short lead times, whereas NWS shows only scattered cases of hit-rate above 50% [[2]](https://doi.org/10.1038/s41612-023-00519-8) [[8]](https://doi.org/10.1175/BAMS-D-19-0019.1) [[9]](https://doi.org/10.1175/WAF-D-19-0106.1) [[10]](https://doi.org/10.1007/s00704-018-2421-9).\n\nCompared to the similar assessment conducted ([](./seasonal_seasonal-monthly-single-levels_forecast-skill_q02.ipynb)), a significant aspect of the hit-rates for terciles is their weak dependence on the originating centre. Although minor differences exist in the performance of the different forecasting systems, the overall patterns of significant hit-rates and even their seasonality is very similar, e.g. higher hit-rates in North West South America during the period October-December associated to El Niño.\n\nCreate a colorbar for each row", "text_with_prefix": "EQC Quality Assessment: \"Evaluation of the hit-rate of seasonal forecasts for tercile categories of monthly anomalies\"\nDataset: seasonal-monthly-single-levels [CDS]\nAspect: forecast-skill_q04 | Category: Seasonal_Forecasts\nSection: Evaluation of the hit-rate of seasonal forecasts for tercile categories of monthly anomalies > Analysis and results > 4. Plot and describe results > Hit-rate for 2 m temperature\n---\nThe same analysis described in section 3 above. can be replicated for all forecasting systems and for all the selected sample regions.\n\nOverall, the seasonal forecasts of 2m temperature have a significant, and potentially usable, hit-rate for terciles in regions closer to the tropics and under the direct influence of El Niño. The further away from the tropics, the lower the expected hit-rate on monthly anomalies.\n\nHit-rates smaller than 50% have been masked out, with the assumption that lower hit-rates would be of limited use for most practical applications. Although this working hypothesis can not be assumed to be strictly true for all purposes, it is a useful assumption for a discussion on the general characteristics of seasonal forecasts [[3]](https://doi.org/10.1002/wcc.523) [[5]](https://doi.org/10.1016/j.crm.2021.100375) [[6]](https://doi.org/10.1016/j.cliser.2023.100347) [[7]](https://doi.org/10.1016/j.cliser.2023.100346).\n\nThe figure below shows that, among the sample regions considered in this analysis, those with the higher and more persistent hit-rate are South East Asia (SEA) and North West South America (NWS). These areas are under the direct influence of ENSO and are expected to benefit from a significantly higher degree of predictability compared to other regions. Other areas with significant, and potentially usable, in practical applications are Northern-East Africa (NEAF) and (North East South America) NES. Other areas like Southern Asia (SAS) and Souther Europe/Mediterranean (MED) have a moderate hit-rate, limited mostly to short lead times, whereas NWS shows only scattered cases of hit-rate above 50% [[2]](https://doi.org/10.1038/s41612-023-00519-8) [[8]](https://doi.org/10.1175/BAMS-D-19-0019.1) [[9]](https://doi.org/10.1175/WAF-D-19-0106.1) [[10]](https://doi.org/10.1007/s00704-018-2421-9).\n\nCompared to the similar assessment conducted ([](./seasonal_seasonal-monthly-single-levels_forecast-skill_q02.ipynb)), a significant aspect of the hit-rates for terciles is their weak dependence on the originating centre. Although minor differences exist in the performance of the different forecasting systems, the overall patterns of significant hit-rates and even their seasonality is very similar, e.g. higher hit-rates in North West South America during the period October-December associated to El Niño.\n\nCreate a colorbar for each row"} {"chunk_id": "seasonal_seasonal-monthly-single-levels_forecast-skill_q04__8b8894aa5e8c", "report_id": "seasonal_seasonal-monthly-single-levels_forecast-skill_q04", "dataset_id": "seasonal-monthly-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "forecast-skill_q04", "aspect_base": "forecast-skill", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Evaluation of the hit-rate of seasonal forecasts for tercile categories of monthly anomalies > Analysis and results > 4. Plot and describe results > Hit-rate for other variables", "title": "Evaluation of the hit-rate of seasonal forecasts for tercile categories of monthly anomalies", "chunk_index": 12, "token_count": 680, "text_raw": "A similar analysis can be conducted for other variables such as monthly accumulated rainfall, wind speed, solar radiation and dew point temperature.\n\n* For **total precipitation** the hit-rate for terciles of monthly anomalies is mostly very small, except for the NES and SEA. Notably, the window of predictability is concentrated in specific seasons: March-May for NES and June-November for SEA [[11]](https://doi.org/10.1175/JHM-D-19-0095.1)\n\n* For **surface solar radiation downwards** the only region with potentially exploitables hit-rates is South East Asia (SEA); limited hit-rate reported over Europe. [[12]](https://doi.org/10.5194/essd-13-2701-2021) [[13]](https://doi.org/10.1016/j.renene.2019.03.134).\n\n* For **10m wind speed** seasonal forecasts produce only limited cases of hit-rates above 60% and only at very short lead times. [[14]](https://doi.org/10.1016/j.renene.2019.04.135).\n\n* For **2m dew point temperature** the pattern of significant hit-rates is clearly linked to the hit-rates of **2m temperature** coherently with an approximately constant pattern of relative humidity whereby the dewpoint temperature is directly related to local temperature by the Clausius-Clapeyron relation [[11]](https://doi.org/10.1175/JHM-D-19-0095.1).\n\nSimilarly to **2m temperature** hit-rates for terciles of seasonal forecasts for other variables have a weak dependence on the originating centres.\n\nThe analysis highlights how seasonal forecasts provide a significant hit-rate for terciles of anomalies observed at a monthly time scale mainly for some specific areas and variables.\n\nNevertheless, valuable information can be extracted by conducting tailored analysis for specific applications [[3]](https://doi.org/10.1002/wcc.523).\n\nOverall, this analysis highlights how seasonal forecasts can not be used as an extension of weather predictions to derive information on expected climate anomalies at monthly or shorter time scales. The actual added value of the use of seasonal forecasts should be always assessed on the basis of the specific applications and decision making processes. Other use cases can be sensitive to different indicators of the quality of seasonal forecasts such as the chance above normal of multiple categories (see for example the [Met Office 3-month outlooks](https://www.metoffice.gov.uk/binaries/content/assets/metofficegovuk/pdf/business/public-sector/civil-contingency/3moutlook_jja_v2.pdf)) or the possibility of anticipating the likelihood of extreme events [[16]](https://doi.org/10.1038/s41467-023-42377-1).\n\nCreate a colorbar for each row", "text_with_prefix": "EQC Quality Assessment: \"Evaluation of the hit-rate of seasonal forecasts for tercile categories of monthly anomalies\"\nDataset: seasonal-monthly-single-levels [CDS]\nAspect: forecast-skill_q04 | Category: Seasonal_Forecasts\nSection: Evaluation of the hit-rate of seasonal forecasts for tercile categories of monthly anomalies > Analysis and results > 4. Plot and describe results > Hit-rate for other variables\n---\nA similar analysis can be conducted for other variables such as monthly accumulated rainfall, wind speed, solar radiation and dew point temperature.\n\n* For **total precipitation** the hit-rate for terciles of monthly anomalies is mostly very small, except for the NES and SEA. Notably, the window of predictability is concentrated in specific seasons: March-May for NES and June-November for SEA [[11]](https://doi.org/10.1175/JHM-D-19-0095.1)\n\n* For **surface solar radiation downwards** the only region with potentially exploitables hit-rates is South East Asia (SEA); limited hit-rate reported over Europe. [[12]](https://doi.org/10.5194/essd-13-2701-2021) [[13]](https://doi.org/10.1016/j.renene.2019.03.134).\n\n* For **10m wind speed** seasonal forecasts produce only limited cases of hit-rates above 60% and only at very short lead times. [[14]](https://doi.org/10.1016/j.renene.2019.04.135).\n\n* For **2m dew point temperature** the pattern of significant hit-rates is clearly linked to the hit-rates of **2m temperature** coherently with an approximately constant pattern of relative humidity whereby the dewpoint temperature is directly related to local temperature by the Clausius-Clapeyron relation [[11]](https://doi.org/10.1175/JHM-D-19-0095.1).\n\nSimilarly to **2m temperature** hit-rates for terciles of seasonal forecasts for other variables have a weak dependence on the originating centres.\n\nThe analysis highlights how seasonal forecasts provide a significant hit-rate for terciles of anomalies observed at a monthly time scale mainly for some specific areas and variables.\n\nNevertheless, valuable information can be extracted by conducting tailored analysis for specific applications [[3]](https://doi.org/10.1002/wcc.523).\n\nOverall, this analysis highlights how seasonal forecasts can not be used as an extension of weather predictions to derive information on expected climate anomalies at monthly or shorter time scales. The actual added value of the use of seasonal forecasts should be always assessed on the basis of the specific applications and decision making processes. Other use cases can be sensitive to different indicators of the quality of seasonal forecasts such as the chance above normal of multiple categories (see for example the [Met Office 3-month outlooks](https://www.metoffice.gov.uk/binaries/content/assets/metofficegovuk/pdf/business/public-sector/civil-contingency/3moutlook_jja_v2.pdf)) or the possibility of anticipating the likelihood of extreme events [[16]](https://doi.org/10.1038/s41467-023-42377-1).\n\nCreate a colorbar for each row"} {"chunk_id": "seasonal_seasonal-monthly-single-levels_forecast-skill_q04__3f3aef6b676c", "report_id": "seasonal_seasonal-monthly-single-levels_forecast-skill_q04", "dataset_id": "seasonal-monthly-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "forecast-skill_q04", "aspect_base": "forecast-skill", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Evaluation of the hit-rate of seasonal forecasts for tercile categories of monthly anomalies > ℹ️ If you want to know more > Key resources", "title": "Evaluation of the hit-rate of seasonal forecasts for tercile categories of monthly anomalies", "chunk_index": 13, "token_count": 398, "text_raw": "The CDS catalogue entries for the data used were:\n* Seasonal forecast monthly statistics on single levels: https://cds.climate.copernicus.eu/datasets/seasonal-monthly-single-levels?tab=overview\n* ERA5 monthly averaged data on single levels from 1940 to present: https://cds.climate.copernicus.eu/datasets/reanalysis-era5-single-levels-monthly-means?tab=overview\n\nMore information on the verification of forecasts are available on the [Forecast User Guide](https://confluence.ecmwf.int/display/FUG/Forecast+User+Guide) at ECMWF. In particular, key concepts on the use of contingency tables are discussed in the section dedicated to [Usefulness of the forecast and Cost/Benefit Approach](https://confluence.ecmwf.int/display/FUG/Section+12.A+Statistical+Concepts+-+Deterministic+Data#Section12.AStatisticalConceptsDeterministicData-UsefulnessoftheForecast-ACost/BenefitApproach)\n\nCode libraries used:\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)\n* The `regionmask` package for the computation of regionalised statistics is available [here](https://regionmask.readthedocs.io/en/stable/index.html)", "text_with_prefix": "EQC Quality Assessment: \"Evaluation of the hit-rate of seasonal forecasts for tercile categories of monthly anomalies\"\nDataset: seasonal-monthly-single-levels [CDS]\nAspect: forecast-skill_q04 | Category: Seasonal_Forecasts\nSection: Evaluation of the hit-rate of seasonal forecasts for tercile categories of monthly anomalies > ℹ️ If you want to know more > Key resources\n---\nThe CDS catalogue entries for the data used were:\n* Seasonal forecast monthly statistics on single levels: https://cds.climate.copernicus.eu/datasets/seasonal-monthly-single-levels?tab=overview\n* ERA5 monthly averaged data on single levels from 1940 to present: https://cds.climate.copernicus.eu/datasets/reanalysis-era5-single-levels-monthly-means?tab=overview\n\nMore information on the verification of forecasts are available on the [Forecast User Guide](https://confluence.ecmwf.int/display/FUG/Forecast+User+Guide) at ECMWF. In particular, key concepts on the use of contingency tables are discussed in the section dedicated to [Usefulness of the forecast and Cost/Benefit Approach](https://confluence.ecmwf.int/display/FUG/Section+12.A+Statistical+Concepts+-+Deterministic+Data#Section12.AStatisticalConceptsDeterministicData-UsefulnessoftheForecast-ACost/BenefitApproach)\n\nCode libraries used:\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)\n* The `regionmask` package for the computation of regionalised statistics is available [here](https://regionmask.readthedocs.io/en/stable/index.html)"} {"chunk_id": "seasonal_seasonal-monthly-single-levels_forecast-skill_q04__f0c7912d29c9", "report_id": "seasonal_seasonal-monthly-single-levels_forecast-skill_q04", "dataset_id": "seasonal-monthly-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "forecast-skill_q04", "aspect_base": "forecast-skill", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Evaluation of the hit-rate of seasonal forecasts for tercile categories of monthly anomalies > ℹ️ If you want to know more > References", "title": "Evaluation of the hit-rate of seasonal forecasts for tercile categories of monthly anomalies", "chunk_index": 14, "token_count": 1069, "text_raw": "[[1]](https://doi.org/10.1016/j.cliser.2024.100496) Khosravi F, Bruno Soares M, Teixeira M, Fontes N , Graca Antonio (2024), Assessing the usability and value of a climate service in the wine sector, Climate Services, Volume 34, 100496\n\n[[2]](https://doi.org/10.1038/s41612-023-00519-8) Ludescher, J., Bunde, A. & Schellnhuber, H.J (2023). Forecasting the El Niño type well before the spring predictability barrier. npj Clim Atmos Sci 6, 196.\n\n[[3]](https://doi.org/10.1002/wcc.523) Bruno Soares M., Daly M, Dessai S (2018) Assessing the value of seasonal climate forecasts for decision-making. WIREs Clim Change. 9:e523.\n\n[[4]](https://doi.org/10.1017/CBO9781139177245.006) Seneviratne, S. I., Nicholls, N., Easterling, D., Goodess, C. M., Kanae, S., Kossin, J., Zwiers, F. W. (2012). Changes in Climate Extremes and their Impacts on the Natural Physical Environment. In C. B. Field, V. Barros, T. F. Stocker, & Q. Dahe (Eds.), Managing the Risks of Extreme Events and Disasters to Advance Climate Change Adaptation: Special Report of the Intergovernmental Panel on Climate Change (pp. 109–230). chapter, Cambridge: Cambridge University Press.\n\n[[5]](https://doi.org/10.1016/j.crm.2021.100375) MacLeod D, Kniveton D R, Todd M C,(2021) Playing the long game: Anticipatory action based on seasonal forecasts, Climate Risk Management, Volume 34, 2021, 100375,\n\n[[6]](https://doi.org/10.1016/j.cliser.2023.100347) Terrado M, Marcos R, González-Reviriego N, Vigo I, Nicodemou A, Graça A, Teixeira M, Fontes N, Silva S, Dell'Aquila A, Ponti L, Calmanti S, Bruno Soares M, Khosravi F, Caboni F, (2023) Co-production pathway of an end-to-end climate service for improved decision-making in the wine sector, Climate Services, Volume 30, 100347,\n\n[[7]](https://doi.org/10.1016/j.cliser.2023.100346) Dell'Aquila, A. Graça, A. Teixeira M Fontes N, Gonzalez-Reviriego N, Marcos-Matamoros R, Chou C, Terrado M, Giannakopoulos C, Varotsos K, Caboni F, Locci R, Nanu M, Porru S, Argiolas S, Bruno Soares M, Sanderson, M. Bruno Soares, M Sanderson, M. (2023) Monitoring climate related risk and opportunities for the wine sector: The MED-GOLD pilot service, Climate Services, 2023, 30, 100346\n\n[[8]](https://doi.org/10.1175/BAMS-D-19-0019.1) Weisheimer, A., D. J. Befort, D. MacLeod, T. Palmer, C. O’Reilly, and K. Strømmen (2020) : Seasonal Forecasts of the Twentieth Century. Bull. Amer. Meteor. Soc., 101, E1413–E1426,\n\n[[9]](https://doi.org/10.1175/WAF-D-19-0106.1) Gubler, S., and Coauthors (2020) Assessment of ECMWF SEAS5 Seasonal Forecast Performance over South America. Wea. Forecasting, 35, 561–584,\n\n[[10]](https://doi.org/10.1007/s00704-018-2421-9) Mohanty, U.C., Nageswararao, M.M., Sinha, P. et al. (2019) Evaluation of performance of seasonal precipitation prediction at regional scale over India. Theor Appl Climatol 135, 1123–1142", "text_with_prefix": "EQC Quality Assessment: \"Evaluation of the hit-rate of seasonal forecasts for tercile categories of monthly anomalies\"\nDataset: seasonal-monthly-single-levels [CDS]\nAspect: forecast-skill_q04 | Category: Seasonal_Forecasts\nSection: Evaluation of the hit-rate of seasonal forecasts for tercile categories of monthly anomalies > ℹ️ If you want to know more > References\n---\n[[1]](https://doi.org/10.1016/j.cliser.2024.100496) Khosravi F, Bruno Soares M, Teixeira M, Fontes N , Graca Antonio (2024), Assessing the usability and value of a climate service in the wine sector, Climate Services, Volume 34, 100496\n\n[[2]](https://doi.org/10.1038/s41612-023-00519-8) Ludescher, J., Bunde, A. & Schellnhuber, H.J (2023). Forecasting the El Niño type well before the spring predictability barrier. npj Clim Atmos Sci 6, 196.\n\n[[3]](https://doi.org/10.1002/wcc.523) Bruno Soares M., Daly M, Dessai S (2018) Assessing the value of seasonal climate forecasts for decision-making. WIREs Clim Change. 9:e523.\n\n[[4]](https://doi.org/10.1017/CBO9781139177245.006) Seneviratne, S. I., Nicholls, N., Easterling, D., Goodess, C. M., Kanae, S., Kossin, J., Zwiers, F. W. (2012). Changes in Climate Extremes and their Impacts on the Natural Physical Environment. In C. B. Field, V. Barros, T. F. Stocker, & Q. Dahe (Eds.), Managing the Risks of Extreme Events and Disasters to Advance Climate Change Adaptation: Special Report of the Intergovernmental Panel on Climate Change (pp. 109–230). chapter, Cambridge: Cambridge University Press.\n\n[[5]](https://doi.org/10.1016/j.crm.2021.100375) MacLeod D, Kniveton D R, Todd M C,(2021) Playing the long game: Anticipatory action based on seasonal forecasts, Climate Risk Management, Volume 34, 2021, 100375,\n\n[[6]](https://doi.org/10.1016/j.cliser.2023.100347) Terrado M, Marcos R, González-Reviriego N, Vigo I, Nicodemou A, Graça A, Teixeira M, Fontes N, Silva S, Dell'Aquila A, Ponti L, Calmanti S, Bruno Soares M, Khosravi F, Caboni F, (2023) Co-production pathway of an end-to-end climate service for improved decision-making in the wine sector, Climate Services, Volume 30, 100347,\n\n[[7]](https://doi.org/10.1016/j.cliser.2023.100346) Dell'Aquila, A. Graça, A. Teixeira M Fontes N, Gonzalez-Reviriego N, Marcos-Matamoros R, Chou C, Terrado M, Giannakopoulos C, Varotsos K, Caboni F, Locci R, Nanu M, Porru S, Argiolas S, Bruno Soares M, Sanderson, M. Bruno Soares, M Sanderson, M. (2023) Monitoring climate related risk and opportunities for the wine sector: The MED-GOLD pilot service, Climate Services, 2023, 30, 100346\n\n[[8]](https://doi.org/10.1175/BAMS-D-19-0019.1) Weisheimer, A., D. J. Befort, D. MacLeod, T. Palmer, C. O’Reilly, and K. Strømmen (2020) : Seasonal Forecasts of the Twentieth Century. Bull. Amer. Meteor. Soc., 101, E1413–E1426,\n\n[[9]](https://doi.org/10.1175/WAF-D-19-0106.1) Gubler, S., and Coauthors (2020) Assessment of ECMWF SEAS5 Seasonal Forecast Performance over South America. Wea. Forecasting, 35, 561–584,\n\n[[10]](https://doi.org/10.1007/s00704-018-2421-9) Mohanty, U.C., Nageswararao, M.M., Sinha, P. et al. (2019) Evaluation of performance of seasonal precipitation prediction at regional scale over India. Theor Appl Climatol 135, 1123–1142"} {"chunk_id": "seasonal_seasonal-monthly-single-levels_forecast-skill_q04__d81088a3d0d4", "report_id": "seasonal_seasonal-monthly-single-levels_forecast-skill_q04", "dataset_id": "seasonal-monthly-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "forecast-skill_q04", "aspect_base": "forecast-skill", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Evaluation of the hit-rate of seasonal forecasts for tercile categories of monthly anomalies > ℹ️ If you want to know more > References", "title": "Evaluation of the hit-rate of seasonal forecasts for tercile categories of monthly anomalies", "chunk_index": 15, "token_count": 600, "text_raw": "Nageswararao, M.M., Sinha, P. et al. (2019) Evaluation of performance of seasonal precipitation prediction at regional scale over India. Theor Appl Climatol 135, 1123–1142\n\n[[11]](https://doi.org/10.1175/JHM-D-19-0095.1) Roy, T., X. He, P. Lin, H. E. Beck, C. Castro, and E. F. Wood (2020) Global Evaluation of Seasonal Precipitation and Temperature Forecasts from NMME. J. Hydrometeor., 21, 2473–2486\n\n[[12]](https://doi.org/10.5194/essd-13-2701-2021) Lorenz, C., Portele, T. C., Laux, P., and Kunstmann, H.(2021) Bias-corrected and spatially disaggregated seasonal forecasts: a long-term reference forecast product for the water sector in semi-arid regions, Earth Syst. Sci. Data, 13, 2701–2722,\n\n[[13]](https://doi.org/10.1016/j.renene.2019.03.134) De Felice M, Bruno Soares M, Alessandri A, Troccoli A,(2019) Scoping the potential usefulness of seasonal climate forecasts for solar power management, Renewable Energy, Volume 142, Pages 215-223,\n\n[[14]](https://doi.org/10.1016/j.renene.2019.04.135) Lledó Ll., Torralba V., Soret A, Ramon J, Doblas-Reyes FJ (2019) Seasonal forecasts of wind power generation, Renewable Energy, Volume 143, 91-100\n\n[[15]](https://doi.org/10.1098/rsif.2013.1162 ) Weisheimer A. and Palmer T. N. (2014) On the reliability of seasonal climate forecastsJ. R. Soc. Interface.1120131162\n\n[[16]](https://doi.org/10.1038/s41467-023-42377-1) Dunstone, N., Smith, D.M., Hardiman, S.C. et al. Windows of opportunity for predicting seasonal climate extremes highlighted by the Pakistan floods of 2022. Nat Commun 14, 6544 (2023).", "text_with_prefix": "EQC Quality Assessment: \"Evaluation of the hit-rate of seasonal forecasts for tercile categories of monthly anomalies\"\nDataset: seasonal-monthly-single-levels [CDS]\nAspect: forecast-skill_q04 | Category: Seasonal_Forecasts\nSection: Evaluation of the hit-rate of seasonal forecasts for tercile categories of monthly anomalies > ℹ️ If you want to know more > References\n---\nNageswararao, M.M., Sinha, P. et al. (2019) Evaluation of performance of seasonal precipitation prediction at regional scale over India. Theor Appl Climatol 135, 1123–1142\n\n[[11]](https://doi.org/10.1175/JHM-D-19-0095.1) Roy, T., X. He, P. Lin, H. E. Beck, C. Castro, and E. F. Wood (2020) Global Evaluation of Seasonal Precipitation and Temperature Forecasts from NMME. J. Hydrometeor., 21, 2473–2486\n\n[[12]](https://doi.org/10.5194/essd-13-2701-2021) Lorenz, C., Portele, T. C., Laux, P., and Kunstmann, H.(2021) Bias-corrected and spatially disaggregated seasonal forecasts: a long-term reference forecast product for the water sector in semi-arid regions, Earth Syst. Sci. Data, 13, 2701–2722,\n\n[[13]](https://doi.org/10.1016/j.renene.2019.03.134) De Felice M, Bruno Soares M, Alessandri A, Troccoli A,(2019) Scoping the potential usefulness of seasonal climate forecasts for solar power management, Renewable Energy, Volume 142, Pages 215-223,\n\n[[14]](https://doi.org/10.1016/j.renene.2019.04.135) Lledó Ll., Torralba V., Soret A, Ramon J, Doblas-Reyes FJ (2019) Seasonal forecasts of wind power generation, Renewable Energy, Volume 143, 91-100\n\n[[15]](https://doi.org/10.1098/rsif.2013.1162 ) Weisheimer A. and Palmer T. N. (2014) On the reliability of seasonal climate forecastsJ. R. Soc. Interface.1120131162\n\n[[16]](https://doi.org/10.1038/s41467-023-42377-1) Dunstone, N., Smith, D.M., Hardiman, S.C. et al. Windows of opportunity for predicting seasonal climate extremes highlighted by the Pakistan floods of 2022. Nat Commun 14, 6544 (2023)."} {"chunk_id": "seasonal_seasonal-monthly-single-levels_resolution_q06__ca6bcf3affd7", "report_id": "seasonal_seasonal-monthly-single-levels_resolution_q06", "dataset_id": "seasonal-monthly-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "resolution_q06", "aspect_base": "resolution", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Assessing the impact of spatial scale and temporal trends on seasonal forecast quality", "title": "Assessing the impact of spatial scale and temporal trends on seasonal forecast quality", "chunk_index": 0, "token_count": 104, "text_raw": "Production date: 30.04.2025\n\nProduced by: Johannes Langvatn (METNorway), Johanna Tjernström (METNorway)", "text_with_prefix": "EQC Quality Assessment: \"Assessing the impact of spatial scale and temporal trends on seasonal forecast quality\"\nDataset: seasonal-monthly-single-levels [CDS]\nAspect: resolution_q06 | Category: Seasonal_Forecasts\nSection: Assessing the impact of spatial scale and temporal trends on seasonal forecast quality\n---\nProduction date: 30.04.2025\n\nProduced by: Johannes Langvatn (METNorway), Johanna Tjernström (METNorway)"} {"chunk_id": "seasonal_seasonal-monthly-single-levels_resolution_q06__ad2b2c904a0b", "report_id": "seasonal_seasonal-monthly-single-levels_resolution_q06", "dataset_id": "seasonal-monthly-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "resolution_q06", "aspect_base": "resolution", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Assessing the impact of spatial scale and temporal trends on seasonal forecast quality > Quality assessment question", "title": "Assessing the impact of spatial scale and temporal trends on seasonal forecast quality", "chunk_index": 1, "token_count": 379, "text_raw": "* How does the quality of seasonal forecasts change when looking at different sized areas? What are the associated limitations?\n* How do long term temporal trends impact the quality of seasonal forecasts?\n\nThe effectiveness of seasonal forecasts in predicting the development of key climate variables—such as temperature, precipitation and sea surface temperatures (SSTs)—varies considerably by variable, region, season, lead time and the state of large-scale climate drivers like ENSO (El Niño–Southern Oscillation). This notebook investigates how a potential trend in the data (e.g. as discussed in [Greuell et al. (2019) \\[1\\]](https://doi.org/10.5194/hess-23-371-2019)) along with selection of regions at various spatial scales (e.g. as discussed in [Prodhomme et al. (2021) \\[2\\]](https://doi.org/10.1007/s00382-021-05828-3), [Gubler et al. (2020) \\[3\\]](https://doi.org/10.1175/WAF-D-19-0106.1) [Quaglia et al. (2021) \\[4\\]](https://doi.org/10.1007/s00382-021-05895-6)) affect forecasting quality. For the purpose of this assessment, seasonal forecast quality is measured by the temporal correlation of the ensemble mean with ERA5 reanalysis.", "text_with_prefix": "EQC Quality Assessment: \"Assessing the impact of spatial scale and temporal trends on seasonal forecast quality\"\nDataset: seasonal-monthly-single-levels [CDS]\nAspect: resolution_q06 | Category: Seasonal_Forecasts\nSection: Assessing the impact of spatial scale and temporal trends on seasonal forecast quality > Quality assessment question\n---\n* How does the quality of seasonal forecasts change when looking at different sized areas? What are the associated limitations?\n* How do long term temporal trends impact the quality of seasonal forecasts?\n\nThe effectiveness of seasonal forecasts in predicting the development of key climate variables—such as temperature, precipitation and sea surface temperatures (SSTs)—varies considerably by variable, region, season, lead time and the state of large-scale climate drivers like ENSO (El Niño–Southern Oscillation). This notebook investigates how a potential trend in the data (e.g. as discussed in [Greuell et al. (2019) \\[1\\]](https://doi.org/10.5194/hess-23-371-2019)) along with selection of regions at various spatial scales (e.g. as discussed in [Prodhomme et al. (2021) \\[2\\]](https://doi.org/10.1007/s00382-021-05828-3), [Gubler et al. (2020) \\[3\\]](https://doi.org/10.1175/WAF-D-19-0106.1) [Quaglia et al. (2021) \\[4\\]](https://doi.org/10.1007/s00382-021-05895-6)) affect forecasting quality. For the purpose of this assessment, seasonal forecast quality is measured by the temporal correlation of the ensemble mean with ERA5 reanalysis."} {"chunk_id": "seasonal_seasonal-monthly-single-levels_resolution_q06__990fdb069647", "report_id": "seasonal_seasonal-monthly-single-levels_resolution_q06", "dataset_id": "seasonal-monthly-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "resolution_q06", "aspect_base": "resolution", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Assessing the impact of spatial scale and temporal trends on seasonal forecast quality > Quality assessment statement", "title": "Assessing the impact of spatial scale and temporal trends on seasonal forecast quality", "chunk_index": 2, "token_count": 414, "text_raw": "These are the key outcomes of this assessment\n\n* The forecast quality (ensemble mean correlation) of t2m varies depending on the selected spatial scale and the choice of the regions or locations.\n * There is a clear drop in forecast quality when moving from the global scale to continental scale (Europe), to a national scale (Germany), and finally to a city scale (Bonn).\n * This is not the case when comparing the global scale, to the Pacific and NINO region, where the forecast quality remains largely constant, or even better, the smaller the selected region.\n * The regional mean for NINO3.4 is better correlated to the reanalysis than the global mean.\n * The nearest grid cell to Bonn (an example city in Germany) has low correlation and most of the quality can be attributed to the underlying trend in the dataset.\n * In contrast, the nearest grid cell to Addis Ababa (an example city in Eastern Africa) has some correlation which is retained even after detrending.\n* For regions and locations which have a visible trend in t2m, the trend inflates the correlation of t2m compared to reanalysis.\n * For Europe, most of the visible correlation can be attributed to the underlying trend.\n * This contrasts with the region of greater horn of Africa, where less of the visible correlation can be attributed to the underlying trend.\n* Users are generally advised to avoid selecting small regions or single grid cells when using seasonal forecasts as they are only expected to be skillful in predicting large scale variations in monthly or seasonal deviations. Selecting larger regions based on a common climatology generally results in increased robustness and signal to noise.\n```", "text_with_prefix": "EQC Quality Assessment: \"Assessing the impact of spatial scale and temporal trends on seasonal forecast quality\"\nDataset: seasonal-monthly-single-levels [CDS]\nAspect: resolution_q06 | Category: Seasonal_Forecasts\nSection: Assessing the impact of spatial scale and temporal trends on seasonal forecast quality > Quality assessment statement\n---\nThese are the key outcomes of this assessment\n\n* The forecast quality (ensemble mean correlation) of t2m varies depending on the selected spatial scale and the choice of the regions or locations.\n * There is a clear drop in forecast quality when moving from the global scale to continental scale (Europe), to a national scale (Germany), and finally to a city scale (Bonn).\n * This is not the case when comparing the global scale, to the Pacific and NINO region, where the forecast quality remains largely constant, or even better, the smaller the selected region.\n * The regional mean for NINO3.4 is better correlated to the reanalysis than the global mean.\n * The nearest grid cell to Bonn (an example city in Germany) has low correlation and most of the quality can be attributed to the underlying trend in the dataset.\n * In contrast, the nearest grid cell to Addis Ababa (an example city in Eastern Africa) has some correlation which is retained even after detrending.\n* For regions and locations which have a visible trend in t2m, the trend inflates the correlation of t2m compared to reanalysis.\n * For Europe, most of the visible correlation can be attributed to the underlying trend.\n * This contrasts with the region of greater horn of Africa, where less of the visible correlation can be attributed to the underlying trend.\n* Users are generally advised to avoid selecting small regions or single grid cells when using seasonal forecasts as they are only expected to be skillful in predicting large scale variations in monthly or seasonal deviations. Selecting larger regions based on a common climatology generally results in increased robustness and signal to noise.\n```"} {"chunk_id": "seasonal_seasonal-monthly-single-levels_resolution_q06__78e2dc1b5da5", "report_id": "seasonal_seasonal-monthly-single-levels_resolution_q06", "dataset_id": "seasonal-monthly-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "resolution_q06", "aspect_base": "resolution", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Assessing the impact of spatial scale and temporal trends on seasonal forecast quality > Methodology", "title": "Assessing the impact of spatial scale and temporal trends on seasonal forecast quality", "chunk_index": 3, "token_count": 968, "text_raw": "This notebook provides an assessment of the quality of seasonal monthly forecast temperature anomalies derived from the monthly means (https://cds.climate.copernicus.eu/datasets/seasonal-monthly-single-levels) through comparison with the ERA5 reanalysis (https://cds.climate.copernicus.eu/datasets/reanalysis-era5-single-levels-monthly-means). The hindcast anomalies are calculated from the monthly means as the anomalies catalogue entry only provides the real time forecasts. Data were accessed from the CDS.\n\nUsing this data the anomaly was calculated for the two meter temperature (t2m), against the period 1993-2024, and was detrended assuming a linear trend. Note that this period contains both hindcast and forecast, which are not entirely equivalent with respect to initialization, and contain 25 and 51 ensemble members respectively. A dataset was then generated containing both the original anomaly and the detrended. The analysis was then carried out globally and for three regions chosen based on effects of trend on the correlation.\n\nFirst, the global case was considered both including and excluding trend, plotting anomaly for the ensemble members, ensemble mean, and ERA5. The correlation was also calculated.\n\nTo further consider the correlation, maps were generated to more clearly view the spatial variations of correlation in anomaly (as can seen on the [C3S verification page](https://confluence.ecmwf.int/display/CKB/C3S+seasonal+forecasts+verification+plots)) between the seasonal forecast and ERA5. Based on these maps, a set of regions were selected for further study, based on the magnitude of the correlation.\nPlots were then generated for nested domains. Three spatial scales were considered, a larger, continental scale, a smaller country scale and a much smaller city scale (a single grid box in the 1x1 degree grid). Data were selected for each domain based on a lat/lon box.\n\nThe resulting plots, displaying the ensemble members, ensemble mean, and ERA5, aim to visualise the variation in quality based on the selected regions of interest. This was also done for the detrended anomaly, and correlation was also calculated for each of the domains.\n\n**[](section-1)**\n * Choose a selection of forecast systems and model versions, hindcast period (normally 1993-2016 to align with the C3S common hindcast period), forecast and leadtime months\n * Ensure that the ERA5 and Seasonal forecast data are regridded to the same grid by using the grid-keyword to the cdsapi.\n * Compute the anomaly of the forecast and reanalysis data\n * The data is detrended assuming a linear trend in the data from start year to end year\n * Both the original and detrended data is saved for the following analysis\n\n**[](section-2)**\n * Compute the ensemble mean of the forecasted anomaly\n * Plot the correlation of forecast and reanalysis data, before and after performing the detrending\n\n**[](section-3)**\n * Correlation is plotted for each grid point on a map\n * Based on this, regions are selected to investigate further; Addis Ababa, Bonn, NINO3.4\n\n**[](section-4)**\n * Plots are generated for the regions chosen in the previous section\n\n**Key limitations:**\n * This assessment primarily considers the correlation between the ensemble mean and ERA5, while the full ensemble spread is visualized in some of the plots, the maps of correlation only use the ensemble mean.\n * The method used to detrend the data is very simple, and thus may impact the quality of the detrended data. \n * The usage of ERA5 as ground truth, while convenient, may be considered a limitation as it is a reanalysis and not observation. \n * The region plots use data at scales that may consider only one or a few grid cells, this is generally not a large enough domain to expect skillful prediction of large scale variations in monthly or seasonal deviations. \n * The assessment in this notebook only makes use of one forecasting system, considering one lead time. While the code should work for other models, and lead times this has not been shown explicitly in this assessment. Therefore, the results outlined here are specific to this forecasting system, period and lead time, further study would need to be conducted to see if the conclusions are general.", "text_with_prefix": "EQC Quality Assessment: \"Assessing the impact of spatial scale and temporal trends on seasonal forecast quality\"\nDataset: seasonal-monthly-single-levels [CDS]\nAspect: resolution_q06 | Category: Seasonal_Forecasts\nSection: Assessing the impact of spatial scale and temporal trends on seasonal forecast quality > Methodology\n---\nThis notebook provides an assessment of the quality of seasonal monthly forecast temperature anomalies derived from the monthly means (https://cds.climate.copernicus.eu/datasets/seasonal-monthly-single-levels) through comparison with the ERA5 reanalysis (https://cds.climate.copernicus.eu/datasets/reanalysis-era5-single-levels-monthly-means). The hindcast anomalies are calculated from the monthly means as the anomalies catalogue entry only provides the real time forecasts. Data were accessed from the CDS.\n\nUsing this data the anomaly was calculated for the two meter temperature (t2m), against the period 1993-2024, and was detrended assuming a linear trend. Note that this period contains both hindcast and forecast, which are not entirely equivalent with respect to initialization, and contain 25 and 51 ensemble members respectively. A dataset was then generated containing both the original anomaly and the detrended. The analysis was then carried out globally and for three regions chosen based on effects of trend on the correlation.\n\nFirst, the global case was considered both including and excluding trend, plotting anomaly for the ensemble members, ensemble mean, and ERA5. The correlation was also calculated.\n\nTo further consider the correlation, maps were generated to more clearly view the spatial variations of correlation in anomaly (as can seen on the [C3S verification page](https://confluence.ecmwf.int/display/CKB/C3S+seasonal+forecasts+verification+plots)) between the seasonal forecast and ERA5. Based on these maps, a set of regions were selected for further study, based on the magnitude of the correlation.\nPlots were then generated for nested domains. Three spatial scales were considered, a larger, continental scale, a smaller country scale and a much smaller city scale (a single grid box in the 1x1 degree grid). Data were selected for each domain based on a lat/lon box.\n\nThe resulting plots, displaying the ensemble members, ensemble mean, and ERA5, aim to visualise the variation in quality based on the selected regions of interest. This was also done for the detrended anomaly, and correlation was also calculated for each of the domains.\n\n**[](section-1)**\n * Choose a selection of forecast systems and model versions, hindcast period (normally 1993-2016 to align with the C3S common hindcast period), forecast and leadtime months\n * Ensure that the ERA5 and Seasonal forecast data are regridded to the same grid by using the grid-keyword to the cdsapi.\n * Compute the anomaly of the forecast and reanalysis data\n * The data is detrended assuming a linear trend in the data from start year to end year\n * Both the original and detrended data is saved for the following analysis\n\n**[](section-2)**\n * Compute the ensemble mean of the forecasted anomaly\n * Plot the correlation of forecast and reanalysis data, before and after performing the detrending\n\n**[](section-3)**\n * Correlation is plotted for each grid point on a map\n * Based on this, regions are selected to investigate further; Addis Ababa, Bonn, NINO3.4\n\n**[](section-4)**\n * Plots are generated for the regions chosen in the previous section\n\n**Key limitations:**\n * This assessment primarily considers the correlation between the ensemble mean and ERA5, while the full ensemble spread is visualized in some of the plots, the maps of correlation only use the ensemble mean.\n * The method used to detrend the data is very simple, and thus may impact the quality of the detrended data. \n * The usage of ERA5 as ground truth, while convenient, may be considered a limitation as it is a reanalysis and not observation. \n * The region plots use data at scales that may consider only one or a few grid cells, this is generally not a large enough domain to expect skillful prediction of large scale variations in monthly or seasonal deviations. \n * The assessment in this notebook only makes use of one forecasting system, considering one lead time. While the code should work for other models, and lead times this has not been shown explicitly in this assessment. Therefore, the results outlined here are specific to this forecasting system, period and lead time, further study would need to be conducted to see if the conclusions are general."} {"chunk_id": "seasonal_seasonal-monthly-single-levels_resolution_q06__f753138c0509", "report_id": "seasonal_seasonal-monthly-single-levels_resolution_q06", "dataset_id": "seasonal-monthly-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "resolution_q06", "aspect_base": "resolution", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Assessing the impact of spatial scale and temporal trends on seasonal forecast quality > Analysis and results > 1. Choose data to use and setup code", "title": "Assessing the impact of spatial scale and temporal trends on seasonal forecast quality", "chunk_index": 4, "token_count": 265, "text_raw": "This section contains the setup and data processing needed for performing the analysis.\n\nImport external code libraries needed.\n\ncollapsable code cells - note that the code cell will be collapsed by the addition of a 'hide-input' tag when the Jupyter Book page is built\n\nDefine functions to be used throughout the notebook.\n\nmake the map global rather than have it zoom in to\nthe extents of any plotted data\n\nThis notebook uses the t2m monthly means from one forecasting system, in this case from ECMWF (system code 51), for a forecast produced in May for leadtime months June, July and August. This was requested for both forecast and hindcast periods, thus there is a difference in the number of ensemble members before and after 2016.\n\nSeasonal forecast\nUse dataarrays directly... for now\nda_seasonal = xr.concat(dataarrays, \"leadtime_month\")\n\n(section-2)=", "text_with_prefix": "EQC Quality Assessment: \"Assessing the impact of spatial scale and temporal trends on seasonal forecast quality\"\nDataset: seasonal-monthly-single-levels [CDS]\nAspect: resolution_q06 | Category: Seasonal_Forecasts\nSection: Assessing the impact of spatial scale and temporal trends on seasonal forecast quality > Analysis and results > 1. Choose data to use and setup code\n---\nThis section contains the setup and data processing needed for performing the analysis.\n\nImport external code libraries needed.\n\ncollapsable code cells - note that the code cell will be collapsed by the addition of a 'hide-input' tag when the Jupyter Book page is built\n\nDefine functions to be used throughout the notebook.\n\nmake the map global rather than have it zoom in to\nthe extents of any plotted data\n\nThis notebook uses the t2m monthly means from one forecasting system, in this case from ECMWF (system code 51), for a forecast produced in May for leadtime months June, July and August. This was requested for both forecast and hindcast periods, thus there is a difference in the number of ensemble members before and after 2016.\n\nSeasonal forecast\nUse dataarrays directly... for now\nda_seasonal = xr.concat(dataarrays, \"leadtime_month\")\n\n(section-2)="} {"chunk_id": "seasonal_seasonal-monthly-single-levels_resolution_q06__fe3bf693eac3", "report_id": "seasonal_seasonal-monthly-single-levels_resolution_q06", "dataset_id": "seasonal-monthly-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "resolution_q06", "aspect_base": "resolution", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Assessing the impact of spatial scale and temporal trends on seasonal forecast quality > Analysis and results > 2. Comparing global correlation", "title": "Assessing the impact of spatial scale and temporal trends on seasonal forecast quality", "chunk_index": 5, "token_count": 187, "text_raw": "Compute the ensemble mean of the seasonal forecast anomaly data.\n\nleadtime here maps to index of leadtime_month: 0 is 2, here 0 is forecast for june from may., 2 is forecast for august from may (leadtime_month = 4)\n\nPlot the original anomaly data for each ensemble member, the ensemble mean, and the reanalysis anomaly. \nPlease note that the ensemble member points are coloured by number to highlight the switch from the hindcast configuration (25 members) to the forecasts (51 members).", "text_with_prefix": "EQC Quality Assessment: \"Assessing the impact of spatial scale and temporal trends on seasonal forecast quality\"\nDataset: seasonal-monthly-single-levels [CDS]\nAspect: resolution_q06 | Category: Seasonal_Forecasts\nSection: Assessing the impact of spatial scale and temporal trends on seasonal forecast quality > Analysis and results > 2. Comparing global correlation\n---\nCompute the ensemble mean of the seasonal forecast anomaly data.\n\nleadtime here maps to index of leadtime_month: 0 is 2, here 0 is forecast for june from may., 2 is forecast for august from may (leadtime_month = 4)\n\nPlot the original anomaly data for each ensemble member, the ensemble mean, and the reanalysis anomaly. \nPlease note that the ensemble member points are coloured by number to highlight the switch from the hindcast configuration (25 members) to the forecasts (51 members)."} {"chunk_id": "seasonal_seasonal-monthly-single-levels_resolution_q06__c7af2573ff10", "report_id": "seasonal_seasonal-monthly-single-levels_resolution_q06", "dataset_id": "seasonal-monthly-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "resolution_q06", "aspect_base": "resolution", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Assessing the impact of spatial scale and temporal trends on seasonal forecast quality > Analysis and results > 3. Map of correlation for detrended data", "title": "Assessing the impact of spatial scale and temporal trends on seasonal forecast quality", "chunk_index": 6, "token_count": 166, "text_raw": "The correlation of the detrended data is plotted compared to reanalysis data for each grid cell on a map.\n\nThe hatched area is where the correlation exceeds the critical value.\n\nIt can be seen in the above plots that there is a high correlation over the North American Great Lakes. This is likely due to the fact that they are large enough that the heat retention impacts correlation skill at one month lead time.", "text_with_prefix": "EQC Quality Assessment: \"Assessing the impact of spatial scale and temporal trends on seasonal forecast quality\"\nDataset: seasonal-monthly-single-levels [CDS]\nAspect: resolution_q06 | Category: Seasonal_Forecasts\nSection: Assessing the impact of spatial scale and temporal trends on seasonal forecast quality > Analysis and results > 3. Map of correlation for detrended data\n---\nThe correlation of the detrended data is plotted compared to reanalysis data for each grid cell on a map.\n\nThe hatched area is where the correlation exceeds the critical value.\n\nIt can be seen in the above plots that there is a high correlation over the North American Great Lakes. This is likely due to the fact that they are large enough that the heat retention impacts correlation skill at one month lead time."} {"chunk_id": "seasonal_seasonal-monthly-single-levels_resolution_q06__1e6197094c51", "report_id": "seasonal_seasonal-monthly-single-levels_resolution_q06", "dataset_id": "seasonal-monthly-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "resolution_q06", "aspect_base": "resolution", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Assessing the impact of spatial scale and temporal trends on seasonal forecast quality > Analysis and results > 3. Map of correlation for detrended data > Selecting regions", "title": "Assessing the impact of spatial scale and temporal trends on seasonal forecast quality", "chunk_index": 7, "token_count": 119, "text_raw": "Based on the correlation maps, regions are selected to investigate further; Addis Ababa, Bonn, NINO3.4.\n\n(section-4)=", "text_with_prefix": "EQC Quality Assessment: \"Assessing the impact of spatial scale and temporal trends on seasonal forecast quality\"\nDataset: seasonal-monthly-single-levels [CDS]\nAspect: resolution_q06 | Category: Seasonal_Forecasts\nSection: Assessing the impact of spatial scale and temporal trends on seasonal forecast quality > Analysis and results > 3. Map of correlation for detrended data > Selecting regions\n---\nBased on the correlation maps, regions are selected to investigate further; Addis Ababa, Bonn, NINO3.4.\n\n(section-4)="} {"chunk_id": "seasonal_seasonal-monthly-single-levels_resolution_q06__69ac75eec3f1", "report_id": "seasonal_seasonal-monthly-single-levels_resolution_q06", "dataset_id": "seasonal-monthly-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "resolution_q06", "aspect_base": "resolution", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Assessing the impact of spatial scale and temporal trends on seasonal forecast quality > Analysis and results > 4. Plot and describe results > Discussion", "title": "Assessing the impact of spatial scale and temporal trends on seasonal forecast quality", "chunk_index": 8, "token_count": 498, "text_raw": "Looking at the comparison of different regions at different spatial scales a few conclusions can be drawn.\n\nFor regions with a visible trend, that trend inflates the correlation for t2m compared against analysis. This can be seen for Europe, where comparing the detrended and trended plots, the correlation is greater for the data with trend in Europe. Hence, for Europe, most of the visible correlation can be attributed to the underlying trend. This can be compared with the region of greater horn of Africa. Where comparing the trended and detrended case indicates that less of the visible correlation can be attributed to the underlying trend.\n\nFor the selected regions, forecast quality depending on scale varies. Looking at the Europe, Germany, Bonn case, there is a drop in forecast quality as the scale decreases. This cannot be observed in the NINO3.4 where the quality remains more or less constant, despite the decrease in region. The Greater Horn of Africa, where the drop off in quality with scale can be seen somewhat in the case of Addis Ababa, but not on the same scale as can be seen in the Europe case.\n\nConsidering this, and the anomalies for the different regions at different scales, it can be seen that the region of NINO3.4 has a better correlation with the reanalysis than the global regional mean, while Bonn has a low correlation where most of the quality can be attributed to the underlying trend. Addis Ababa has some correlation, which is retained even after detrending. The high correlations for the NINO3.4 region are partially dependent on the start date and forecast length, which can be seen in the [SST indices plots on the C3S verification page](https://confluence.ecmwf.int/display/CKB/C3S+seasonal+forecasts+verification+plots)), though it should be noted that these plots use sea surface temperature and not t2m. Forecast quality, and seasonal forecast biases, generally vary with start date and lead time, as seen in some of the other assessments.", "text_with_prefix": "EQC Quality Assessment: \"Assessing the impact of spatial scale and temporal trends on seasonal forecast quality\"\nDataset: seasonal-monthly-single-levels [CDS]\nAspect: resolution_q06 | Category: Seasonal_Forecasts\nSection: Assessing the impact of spatial scale and temporal trends on seasonal forecast quality > Analysis and results > 4. Plot and describe results > Discussion\n---\nLooking at the comparison of different regions at different spatial scales a few conclusions can be drawn.\n\nFor regions with a visible trend, that trend inflates the correlation for t2m compared against analysis. This can be seen for Europe, where comparing the detrended and trended plots, the correlation is greater for the data with trend in Europe. Hence, for Europe, most of the visible correlation can be attributed to the underlying trend. This can be compared with the region of greater horn of Africa. Where comparing the trended and detrended case indicates that less of the visible correlation can be attributed to the underlying trend.\n\nFor the selected regions, forecast quality depending on scale varies. Looking at the Europe, Germany, Bonn case, there is a drop in forecast quality as the scale decreases. This cannot be observed in the NINO3.4 where the quality remains more or less constant, despite the decrease in region. The Greater Horn of Africa, where the drop off in quality with scale can be seen somewhat in the case of Addis Ababa, but not on the same scale as can be seen in the Europe case.\n\nConsidering this, and the anomalies for the different regions at different scales, it can be seen that the region of NINO3.4 has a better correlation with the reanalysis than the global regional mean, while Bonn has a low correlation where most of the quality can be attributed to the underlying trend. Addis Ababa has some correlation, which is retained even after detrending. The high correlations for the NINO3.4 region are partially dependent on the start date and forecast length, which can be seen in the [SST indices plots on the C3S verification page](https://confluence.ecmwf.int/display/CKB/C3S+seasonal+forecasts+verification+plots)), though it should be noted that these plots use sea surface temperature and not t2m. Forecast quality, and seasonal forecast biases, generally vary with start date and lead time, as seen in some of the other assessments."} {"chunk_id": "seasonal_seasonal-monthly-single-levels_resolution_q06__3d77e2f79217", "report_id": "seasonal_seasonal-monthly-single-levels_resolution_q06", "dataset_id": "seasonal-monthly-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "resolution_q06", "aspect_base": "resolution", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Assessing the impact of spatial scale and temporal trends on seasonal forecast quality > ℹ️ If you want to know more > Key resources", "title": "Assessing the impact of spatial scale and temporal trends on seasonal forecast quality", "chunk_index": 9, "token_count": 170, "text_raw": "* Seasonal forecast monthly statistics on single levels: [10.24381/cds.68dd14c3](https://doi.org/10.24381/cds.68dd14c3)\n * ERA5 monthly averaged data on single levels from 1940 to present: [10.24381/cds.f17050d7](https://doi.org/10.24381/cds.f17050d7)", "text_with_prefix": "EQC Quality Assessment: \"Assessing the impact of spatial scale and temporal trends on seasonal forecast quality\"\nDataset: seasonal-monthly-single-levels [CDS]\nAspect: resolution_q06 | Category: Seasonal_Forecasts\nSection: Assessing the impact of spatial scale and temporal trends on seasonal forecast quality > ℹ️ If you want to know more > Key resources\n---\n* Seasonal forecast monthly statistics on single levels: [10.24381/cds.68dd14c3](https://doi.org/10.24381/cds.68dd14c3)\n * ERA5 monthly averaged data on single levels from 1940 to present: [10.24381/cds.f17050d7](https://doi.org/10.24381/cds.f17050d7)"} {"chunk_id": "seasonal_seasonal-monthly-single-levels_resolution_q06__2423e34d4df4", "report_id": "seasonal_seasonal-monthly-single-levels_resolution_q06", "dataset_id": "seasonal-monthly-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "resolution_q06", "aspect_base": "resolution", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Assessing the impact of spatial scale and temporal trends on seasonal forecast quality > ℹ️ If you want to know more > Code libraries used:", "title": "Assessing the impact of spatial scale and temporal trends on seasonal forecast quality", "chunk_index": 10, "token_count": 163, "text_raw": "* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)\n * xarray\n * numpy\n * matplotlib\n * cartopy\n * linear_model from sklearn", "text_with_prefix": "EQC Quality Assessment: \"Assessing the impact of spatial scale and temporal trends on seasonal forecast quality\"\nDataset: seasonal-monthly-single-levels [CDS]\nAspect: resolution_q06 | Category: Seasonal_Forecasts\nSection: Assessing the impact of spatial scale and temporal trends on seasonal forecast quality > ℹ️ If you want to know more > Code libraries used:\n---\n* [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)\n * xarray\n * numpy\n * matplotlib\n * cartopy\n * linear_model from sklearn"} {"chunk_id": "seasonal_seasonal-monthly-single-levels_resolution_q06__064ef2c3a830", "report_id": "seasonal_seasonal-monthly-single-levels_resolution_q06", "dataset_id": "seasonal-monthly-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "resolution_q06", "aspect_base": "resolution", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Assessing the impact of spatial scale and temporal trends on seasonal forecast quality > ℹ️ If you want to know more > References", "title": "Assessing the impact of spatial scale and temporal trends on seasonal forecast quality", "chunk_index": 11, "token_count": 376, "text_raw": "[1] Greuell, W., Franssen, W. H. P., and Hutjes, R. W. A.: Seasonal streamflow forecasts for Europe – Part 2: Sources of skill, Hydrol. Earth Syst. Sci., 23, 371–391, https://doi.org/10.5194/hess-23-371-2019, 2019.\n\n[2] Prodhomme, C., Materia, S., Ardilouze, C. et al. Seasonal prediction of European summer heatwaves. Clim Dyn 58, 2149–2166 (2022). https://doi.org/10.1007/s00382-021-05828-3\n\n[3] Gubler, S., and Coauthors, 2020: Assessment of ECMWF SEAS5 Seasonal Forecast Performance over South America. Wea. Forecasting, 35, 561–584, https://doi.org/10.1175/WAF-D-19-0106.1\n\n[4] Calì Quaglia, F., Terzago, S. & von Hardenberg, J. Temperature and precipitation seasonal forecasts over the Mediterranean region: added value compared to simple forecasting methods. Clim Dyn 58, 2167–2191 (2022). https://doi.org/10.1007/s00382-021-05895-6", "text_with_prefix": "EQC Quality Assessment: \"Assessing the impact of spatial scale and temporal trends on seasonal forecast quality\"\nDataset: seasonal-monthly-single-levels [CDS]\nAspect: resolution_q06 | Category: Seasonal_Forecasts\nSection: Assessing the impact of spatial scale and temporal trends on seasonal forecast quality > ℹ️ If you want to know more > References\n---\n[1] Greuell, W., Franssen, W. H. P., and Hutjes, R. W. A.: Seasonal streamflow forecasts for Europe – Part 2: Sources of skill, Hydrol. Earth Syst. Sci., 23, 371–391, https://doi.org/10.5194/hess-23-371-2019, 2019.\n\n[2] Prodhomme, C., Materia, S., Ardilouze, C. et al. Seasonal prediction of European summer heatwaves. Clim Dyn 58, 2149–2166 (2022). https://doi.org/10.1007/s00382-021-05828-3\n\n[3] Gubler, S., and Coauthors, 2020: Assessment of ECMWF SEAS5 Seasonal Forecast Performance over South America. Wea. Forecasting, 35, 561–584, https://doi.org/10.1175/WAF-D-19-0106.1\n\n[4] Calì Quaglia, F., Terzago, S. & von Hardenberg, J. Temperature and precipitation seasonal forecasts over the Mediterranean region: added value compared to simple forecasting methods. Clim Dyn 58, 2167–2191 (2022). https://doi.org/10.1007/s00382-021-05895-6"} {"chunk_id": "seasonal_seasonal-monthly-single-levels_uncertainty-quality-flags_q01__438abd0edf20", "report_id": "seasonal_seasonal-monthly-single-levels_uncertainty-quality-flags_q01", "dataset_id": "seasonal-monthly-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty-quality-flags_q01", "aspect_base": "uncertainty-quality-flags", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Assessing possible outcomes of seasonal temperature forecast > Quality assessment questions", "title": "Assessing possible outcomes of seasonal temperature forecast", "chunk_index": 0, "token_count": 173, "text_raw": "* **How can I assess the forecast temperatures for two months time?**\n* **Do the different forecast systems provide a consistent outlook?**\n\nSeasonal forecast in C3S provides a prediction of the near future. For two months prediction time, emphasis should be put on a probabilistic approach and understanding the uncertainty in the prediction [[1]](https://doi.org/10.1098/rsif.2013.1162). To this end, here we assess important characteristics of seasonal forecasts, how they can be compared, interpreted and used.", "text_with_prefix": "EQC Quality Assessment: \"Assessing possible outcomes of seasonal temperature forecast\"\nDataset: seasonal-monthly-single-levels [CDS]\nAspect: uncertainty-quality-flags_q01 | Category: Seasonal_Forecasts\nSection: Assessing possible outcomes of seasonal temperature forecast > Quality assessment questions\n---\n* **How can I assess the forecast temperatures for two months time?**\n* **Do the different forecast systems provide a consistent outlook?**\n\nSeasonal forecast in C3S provides a prediction of the near future. For two months prediction time, emphasis should be put on a probabilistic approach and understanding the uncertainty in the prediction [[1]](https://doi.org/10.1098/rsif.2013.1162). To this end, here we assess important characteristics of seasonal forecasts, how they can be compared, interpreted and used."} {"chunk_id": "seasonal_seasonal-monthly-single-levels_uncertainty-quality-flags_q01__1de78ced9c6c", "report_id": "seasonal_seasonal-monthly-single-levels_uncertainty-quality-flags_q01", "dataset_id": "seasonal-monthly-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty-quality-flags_q01", "aspect_base": "uncertainty-quality-flags", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Assessing possible outcomes of seasonal temperature forecast > Quality assessment statement", "title": "Assessing possible outcomes of seasonal temperature forecast", "chunk_index": 1, "token_count": 520, "text_raw": "These are the key outcomes of this assessment\n\n* This notebook shows one way to visualise the range of outcomes covered by the ensemble members in seasonal forecast systems from C3S and compare it with the forecast system climatology as a reference.\n* The resulting plots show the uncertainty in the forecast (range of outcomes for each forecast system and between forecast systems), how the seasonal forecast differs from the model climatology, and how likely it is that the coming months will be in the lower, middle or upper terciles of the climatology.\n* For the specific example provided, forecasts issued in June 2023 and valid for 2m air temperature for a region over Southern Norway in July 2023 shows that:\n * All forecast systems provide a consistent outlook that the month will end up in the upper tercile of the climatology (probabilities between 45 and 61% for the different systems).\n * The climatology of the different forecast systems are quite different, hence they need to be bias adjusted before their face-value can be combined in a multi-model forecast. This emphasises the need for always comparing the forecasts with the hindcast climatology.\n* If the forecast systems yield inconsistent outlooks, a multi system summary (see below) can be employed to determine whether there's a broad consensus with only a few models deviating, or if there's a complete disagreement in forecasted probabilities. In the former scenario, the divergent model systems might be seen as outliers highlighting the possibility of different outcome than the majority. While in the latter, the reliability of the forecast is reduced.\n* Seasonal forecasts are inherently probabilistic and must be interpreted considering the chaotic nature of the system and the forecast systems' capabilities and limitations. In general, multi-model system forecasts often provide better estimates of the uncertainty than applying a single system. Due to the long lead times involved and associated uncertainty, forecast skill can only be expected for large-scale features in space and time and using e.g. probabilities for terciles can be useful in decision-making.\n```\n![tercile_distribution_t2m_without_label.png](attachment:94c8182f-0d8a-4b57-bad1-b41d5cf7f1ba.png)", "text_with_prefix": "EQC Quality Assessment: \"Assessing possible outcomes of seasonal temperature forecast\"\nDataset: seasonal-monthly-single-levels [CDS]\nAspect: uncertainty-quality-flags_q01 | Category: Seasonal_Forecasts\nSection: Assessing possible outcomes of seasonal temperature forecast > Quality assessment statement\n---\nThese are the key outcomes of this assessment\n\n* This notebook shows one way to visualise the range of outcomes covered by the ensemble members in seasonal forecast systems from C3S and compare it with the forecast system climatology as a reference.\n* The resulting plots show the uncertainty in the forecast (range of outcomes for each forecast system and between forecast systems), how the seasonal forecast differs from the model climatology, and how likely it is that the coming months will be in the lower, middle or upper terciles of the climatology.\n* For the specific example provided, forecasts issued in June 2023 and valid for 2m air temperature for a region over Southern Norway in July 2023 shows that:\n * All forecast systems provide a consistent outlook that the month will end up in the upper tercile of the climatology (probabilities between 45 and 61% for the different systems).\n * The climatology of the different forecast systems are quite different, hence they need to be bias adjusted before their face-value can be combined in a multi-model forecast. This emphasises the need for always comparing the forecasts with the hindcast climatology.\n* If the forecast systems yield inconsistent outlooks, a multi system summary (see below) can be employed to determine whether there's a broad consensus with only a few models deviating, or if there's a complete disagreement in forecasted probabilities. In the former scenario, the divergent model systems might be seen as outliers highlighting the possibility of different outcome than the majority. While in the latter, the reliability of the forecast is reduced.\n* Seasonal forecasts are inherently probabilistic and must be interpreted considering the chaotic nature of the system and the forecast systems' capabilities and limitations. In general, multi-model system forecasts often provide better estimates of the uncertainty than applying a single system. Due to the long lead times involved and associated uncertainty, forecast skill can only be expected for large-scale features in space and time and using e.g. probabilities for terciles can be useful in decision-making.\n```\n![tercile_distribution_t2m_without_label.png](attachment:94c8182f-0d8a-4b57-bad1-b41d5cf7f1ba.png)"} {"chunk_id": "seasonal_seasonal-monthly-single-levels_uncertainty-quality-flags_q01__29bbf18a7010", "report_id": "seasonal_seasonal-monthly-single-levels_uncertainty-quality-flags_q01", "dataset_id": "seasonal-monthly-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty-quality-flags_q01", "aspect_base": "uncertainty-quality-flags", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Assessing possible outcomes of seasonal temperature forecast > Methodology", "title": "Assessing possible outcomes of seasonal temperature forecast", "chunk_index": 2, "token_count": 437, "text_raw": "This code in this notebook enables an assessment of the range of possible outcomes for two months prediction time of seasonal forecast systems, in the case of near surface temperature. For reference, the forecasts are presented together with the expected (normal) range of temperatures for the month and region of interest. The normal range is based on so-called hindcasts of past weather from the same forecasting system.\n\nThis quality assessment does not assess the historical quality of the seasonal forecast, as the forecasts (e.g. hindcasts) are not compared with historical outcomes (e.g. observations or re-analyses like ERA5). This will be done in another notebook, and is also done [here](https://confluence.ecmwf.int/display/CKB/C3S+seasonal+forecasts+verification+plots), in verification plots for the C3S graphical products.\n\n**[](section-1)**\n - Choose a selection of forecast systems and versions, forecast start time (year and month), hindcast period (normally 1993-2016), region (latitude-longitude box).\n - Define a weighting function for the gridded data, and if needed conversions for the variables used.\n\n**[](section-2)**\n - Retrieve data for monthly means of all ensemble members for the selected parameters above, from the data catalogue ”Seasonal forecast monthly statistics on single levels”, for both forecasts and hindcasts\n - Compute the spatial mean of 2 m temperature.\n\n**[](section-3)**\n - Showing the normal range of each forecast system (hindcast distribution) together with the forecast outcomes. Included in the plots are probabilities of near surface temperature forecast within the lower, middle and upper terciles of the normal range.\n\n**[](section-4)**\n - The forecast probabilities are combined and visualised at the end in a multi-model plot.", "text_with_prefix": "EQC Quality Assessment: \"Assessing possible outcomes of seasonal temperature forecast\"\nDataset: seasonal-monthly-single-levels [CDS]\nAspect: uncertainty-quality-flags_q01 | Category: Seasonal_Forecasts\nSection: Assessing possible outcomes of seasonal temperature forecast > Methodology\n---\nThis code in this notebook enables an assessment of the range of possible outcomes for two months prediction time of seasonal forecast systems, in the case of near surface temperature. For reference, the forecasts are presented together with the expected (normal) range of temperatures for the month and region of interest. The normal range is based on so-called hindcasts of past weather from the same forecasting system.\n\nThis quality assessment does not assess the historical quality of the seasonal forecast, as the forecasts (e.g. hindcasts) are not compared with historical outcomes (e.g. observations or re-analyses like ERA5). This will be done in another notebook, and is also done [here](https://confluence.ecmwf.int/display/CKB/C3S+seasonal+forecasts+verification+plots), in verification plots for the C3S graphical products.\n\n**[](section-1)**\n - Choose a selection of forecast systems and versions, forecast start time (year and month), hindcast period (normally 1993-2016), region (latitude-longitude box).\n - Define a weighting function for the gridded data, and if needed conversions for the variables used.\n\n**[](section-2)**\n - Retrieve data for monthly means of all ensemble members for the selected parameters above, from the data catalogue ”Seasonal forecast monthly statistics on single levels”, for both forecasts and hindcasts\n - Compute the spatial mean of 2 m temperature.\n\n**[](section-3)**\n - Showing the normal range of each forecast system (hindcast distribution) together with the forecast outcomes. Included in the plots are probabilities of near surface temperature forecast within the lower, middle and upper terciles of the normal range.\n\n**[](section-4)**\n - The forecast probabilities are combined and visualised at the end in a multi-model plot."} {"chunk_id": "seasonal_seasonal-monthly-single-levels_uncertainty-quality-flags_q01__0babb73275c4", "report_id": "seasonal_seasonal-monthly-single-levels_uncertainty-quality-flags_q01", "dataset_id": "seasonal-monthly-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty-quality-flags_q01", "aspect_base": "uncertainty-quality-flags", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Assessing possible outcomes of seasonal temperature forecast > Analysis and results > 1. Choose data to use and setup code", "title": "Assessing possible outcomes of seasonal temperature forecast", "chunk_index": 3, "token_count": 699, "text_raw": "In this section, the customisable options of the notebook are set. These variables consists of:\n\n- Model forecast system \n - forecast centre \n - system version\n- Time\n - forecast year\n - first year of the model hindcast\n - last year of the model hindcast\n - forecast month \n- Region of interest\n - name of the region (only used in figure-captions)\n - latitudes given as slice(maxlat, minlat) in degrees (north is positive values)\n - longitudes given as slice(minlon, maxlon) in degrees (east is positive values)\n- Weather parameters\n - name of variable in grib-file\n - name of variable in CADS-api\n - name of the three tercile categories (lower, middle and upper tercile)\n- Download parameters\n - chunk size of the data\n - number of concurrent request for parallel download \n- Combination of originating centre and model system (Operational forecast model systems per March 2024):\n - centre = \"ecmwf\", system = \"51\"\n - centre = \"ukmo\", system = \"602\"\n - centre = \"meteo_france\", system = \"8\"\n - centre = \"dwd\", system = \"21\"\n - centre = \"cmcc\", system = \"35\"\n - centre = \"ncep\", system = \"2\"\n - centre = \"jma\", system = \"3\"\n - centre = \"eccc\", system = \"2\"\n - centre = \"eccc\", system = \"3\"\n\nTo see the different alternatives see the API-request at the bottom of [this page](https://cds.climate.copernicus.eu/datasets/seasonal-monthly-single-levels?tab=form)\n\nNote: if you want to change parameter you might also want to add conversion functions in this section. For example for precipitation you might want to convert from precipitation rate to total precipitation amount\n\nThe data is downloaded to a 1 degree lat/lon grid from 89.5 N to 89.5 S and 179.5 W to 179.5 E which is the grid for the latest model versions of all models except for JMA. The JMA data, which is on a 1.25 x 1.25 grid, is therefore interpolated. This is as a convencience to ensure that the lat-lon slices return the same grids for the different systems, and that we can use similar requests to cds. Otherwise the interpolation would not be needed since we are doing an area average. Also, the 'weights' flag will be set to 'True' in the next section, which means the area average is carried out taking into account the variation of the cell areas with cosine of latitude in the regular lat-lon grid.\n\nDownload parameters\nConstruct the request to be sent to CADS\nDefine the weighting function\nConvert from K to C if variable is 2m air temperature\n\n(section-2)=", "text_with_prefix": "EQC Quality Assessment: \"Assessing possible outcomes of seasonal temperature forecast\"\nDataset: seasonal-monthly-single-levels [CDS]\nAspect: uncertainty-quality-flags_q01 | Category: Seasonal_Forecasts\nSection: Assessing possible outcomes of seasonal temperature forecast > Analysis and results > 1. Choose data to use and setup code\n---\nIn this section, the customisable options of the notebook are set. These variables consists of:\n\n- Model forecast system \n - forecast centre \n - system version\n- Time\n - forecast year\n - first year of the model hindcast\n - last year of the model hindcast\n - forecast month \n- Region of interest\n - name of the region (only used in figure-captions)\n - latitudes given as slice(maxlat, minlat) in degrees (north is positive values)\n - longitudes given as slice(minlon, maxlon) in degrees (east is positive values)\n- Weather parameters\n - name of variable in grib-file\n - name of variable in CADS-api\n - name of the three tercile categories (lower, middle and upper tercile)\n- Download parameters\n - chunk size of the data\n - number of concurrent request for parallel download \n- Combination of originating centre and model system (Operational forecast model systems per March 2024):\n - centre = \"ecmwf\", system = \"51\"\n - centre = \"ukmo\", system = \"602\"\n - centre = \"meteo_france\", system = \"8\"\n - centre = \"dwd\", system = \"21\"\n - centre = \"cmcc\", system = \"35\"\n - centre = \"ncep\", system = \"2\"\n - centre = \"jma\", system = \"3\"\n - centre = \"eccc\", system = \"2\"\n - centre = \"eccc\", system = \"3\"\n\nTo see the different alternatives see the API-request at the bottom of [this page](https://cds.climate.copernicus.eu/datasets/seasonal-monthly-single-levels?tab=form)\n\nNote: if you want to change parameter you might also want to add conversion functions in this section. For example for precipitation you might want to convert from precipitation rate to total precipitation amount\n\nThe data is downloaded to a 1 degree lat/lon grid from 89.5 N to 89.5 S and 179.5 W to 179.5 E which is the grid for the latest model versions of all models except for JMA. The JMA data, which is on a 1.25 x 1.25 grid, is therefore interpolated. This is as a convencience to ensure that the lat-lon slices return the same grids for the different systems, and that we can use similar requests to cds. Otherwise the interpolation would not be needed since we are doing an area average. Also, the 'weights' flag will be set to 'True' in the next section, which means the area average is carried out taking into account the variation of the cell areas with cosine of latitude in the regular lat-lon grid.\n\nDownload parameters\nConstruct the request to be sent to CADS\nDefine the weighting function\nConvert from K to C if variable is 2m air temperature\n\n(section-2)="} {"chunk_id": "seasonal_seasonal-monthly-single-levels_uncertainty-quality-flags_q01__b90ec8cb72d5", "report_id": "seasonal_seasonal-monthly-single-levels_uncertainty-quality-flags_q01", "dataset_id": "seasonal-monthly-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty-quality-flags_q01", "aspect_base": "uncertainty-quality-flags", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Assessing possible outcomes of seasonal temperature forecast > Analysis and results > 2. Seasonal forecast and hindcast data retrieval and area average", "title": "Assessing possible outcomes of seasonal temperature forecast", "chunk_index": 4, "token_count": 768, "text_raw": "In this section the seasonal data are downloaded and transformed into regionalised means. The time is transformed into two variables ‘time/indexing_time’ and ‘forecast month’. For burst ensemble models ‘time’ is the time when all ensembles are initialised. For lagged ensemble models (JMA, NCEP and UK MetOffice) the ‘indexing_time’ must be used instead of ‘time’ in order to get the same variable, since ensemble members have different initialisation time (See more on this [confluence-page](https://confluence.ecmwf.int/display/CKB/Seasonal+forecasts+and+the+Copernicus+Climate+Change+Service#SeasonalforecastsandtheCopernicusClimateChangeService-%22Burst%22vs.%22lagged%22mode)). The ‘forecast month’ is the lead time month from month 1 to 6.\n\nStore the downloaded dataset in the dictionary\n\n```text\n100%|██████████| 24/24 [00:00<00:00, 34.70it/s]\n100%|██████████| 1/1 [00:00<00:00, 47.55it/s]\n100%|██████████| 24/24 [00:00<00:00, 48.03it/s]\n100%|██████████| 1/1 [00:00<00:00, 54.61it/s]\n100%|██████████| 24/24 [00:00<00:00, 63.89it/s]\n100%|██████████| 1/1 [00:00<00:00, 43.95it/s]\n100%|██████████| 24/24 [00:00<00:00, 39.22it/s]\n100%|██████████| 1/1 [00:00<00:00, 51.80it/s]\n100%|██████████| 24/24 [00:00<00:00, 50.26it/s]\n100%|██████████| 1/1 [00:00<00:00, 32.12it/s]\n100%|██████████| 24/24 [00:00<00:00, 33.74it/s]\n100%|██████████| 1/1 [00:00<00:00, 45.89it/s]\n100%|██████████| 24/24 [00:01<00:00, 20.87it/s]\n100%|██████████| 1/1 [00:00<00:00, 3.85it/s]\n100%|██████████| 24/24 [00:00<00:00, 33.76it/s]\n100%|██████████| 1/1 [00:00<00:00, 43.51it/s]\n100%|██████████| 24/24 [00:00<00:00, 47.80it/s]\n100%|██████████| 1/1 [00:00<00:00, 46.99it/s]\n```\n\n(section-3)=", "text_with_prefix": "EQC Quality Assessment: \"Assessing possible outcomes of seasonal temperature forecast\"\nDataset: seasonal-monthly-single-levels [CDS]\nAspect: uncertainty-quality-flags_q01 | Category: Seasonal_Forecasts\nSection: Assessing possible outcomes of seasonal temperature forecast > Analysis and results > 2. Seasonal forecast and hindcast data retrieval and area average\n---\nIn this section the seasonal data are downloaded and transformed into regionalised means. The time is transformed into two variables ‘time/indexing_time’ and ‘forecast month’. For burst ensemble models ‘time’ is the time when all ensembles are initialised. For lagged ensemble models (JMA, NCEP and UK MetOffice) the ‘indexing_time’ must be used instead of ‘time’ in order to get the same variable, since ensemble members have different initialisation time (See more on this [confluence-page](https://confluence.ecmwf.int/display/CKB/Seasonal+forecasts+and+the+Copernicus+Climate+Change+Service#SeasonalforecastsandtheCopernicusClimateChangeService-%22Burst%22vs.%22lagged%22mode)). The ‘forecast month’ is the lead time month from month 1 to 6.\n\nStore the downloaded dataset in the dictionary\n\n```text\n100%|██████████| 24/24 [00:00<00:00, 34.70it/s]\n100%|██████████| 1/1 [00:00<00:00, 47.55it/s]\n100%|██████████| 24/24 [00:00<00:00, 48.03it/s]\n100%|██████████| 1/1 [00:00<00:00, 54.61it/s]\n100%|██████████| 24/24 [00:00<00:00, 63.89it/s]\n100%|██████████| 1/1 [00:00<00:00, 43.95it/s]\n100%|██████████| 24/24 [00:00<00:00, 39.22it/s]\n100%|██████████| 1/1 [00:00<00:00, 51.80it/s]\n100%|██████████| 24/24 [00:00<00:00, 50.26it/s]\n100%|██████████| 1/1 [00:00<00:00, 32.12it/s]\n100%|██████████| 24/24 [00:00<00:00, 33.74it/s]\n100%|██████████| 1/1 [00:00<00:00, 45.89it/s]\n100%|██████████| 24/24 [00:01<00:00, 20.87it/s]\n100%|██████████| 1/1 [00:00<00:00, 3.85it/s]\n100%|██████████| 24/24 [00:00<00:00, 33.76it/s]\n100%|██████████| 1/1 [00:00<00:00, 43.51it/s]\n100%|██████████| 24/24 [00:00<00:00, 47.80it/s]\n100%|██████████| 1/1 [00:00<00:00, 46.99it/s]\n```\n\n(section-3)="} {"chunk_id": "seasonal_seasonal-monthly-single-levels_uncertainty-quality-flags_q01__b3205102fb36", "report_id": "seasonal_seasonal-monthly-single-levels_uncertainty-quality-flags_q01", "dataset_id": "seasonal-monthly-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty-quality-flags_q01", "aspect_base": "uncertainty-quality-flags", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Assessing possible outcomes of seasonal temperature forecast > Analysis and results > 3. Make histogram plots for each forecast system", "title": "Assessing possible outcomes of seasonal temperature forecast", "chunk_index": 5, "token_count": 197, "text_raw": "This section extracts data for the selected region from the monthly data on single levels and compute the spatial mean of the chosen variable. Then density plots are created for each lead time month. The resulting plots show the uncertainty in the forecast (range of outcomes), how the seasonal forecast differs from the model climatology, and how likely it is that the coming months will be in the lower, middle or upper terciles of the climatology.\n\nSet up for for multi-ensemble histogram plot\nHistogram plot for each model system\nAdd background color and text\nText and labels\nText below each plot\n\n(section-4)=", "text_with_prefix": "EQC Quality Assessment: \"Assessing possible outcomes of seasonal temperature forecast\"\nDataset: seasonal-monthly-single-levels [CDS]\nAspect: uncertainty-quality-flags_q01 | Category: Seasonal_Forecasts\nSection: Assessing possible outcomes of seasonal temperature forecast > Analysis and results > 3. Make histogram plots for each forecast system\n---\nThis section extracts data for the selected region from the monthly data on single levels and compute the spatial mean of the chosen variable. Then density plots are created for each lead time month. The resulting plots show the uncertainty in the forecast (range of outcomes), how the seasonal forecast differs from the model climatology, and how likely it is that the coming months will be in the lower, middle or upper terciles of the climatology.\n\nSet up for for multi-ensemble histogram plot\nHistogram plot for each model system\nAdd background color and text\nText and labels\nText below each plot\n\n(section-4)="} {"chunk_id": "seasonal_seasonal-monthly-single-levels_uncertainty-quality-flags_q01__e77fae81ddd1", "report_id": "seasonal_seasonal-monthly-single-levels_uncertainty-quality-flags_q01", "dataset_id": "seasonal-monthly-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty-quality-flags_q01", "aspect_base": "uncertainty-quality-flags", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Assessing possible outcomes of seasonal temperature forecast > Analysis and results > 4. Summarise the probabilities for each system", "title": "Assessing possible outcomes of seasonal temperature forecast", "chunk_index": 6, "token_count": 452, "text_raw": "This section summarises the forecasted probabilites of the temperature being colder, near average or warmer than their respective hindcast climatologies for all systems.\nThe figure produced reinforces the findings from the figures produced in the last section, that it is more likely to be warmer than usual, compared to the model hindcast climatology than it is to be average or colder compared to the hindcast.\n\nPlotting the multi-ensemble histogram\n\nFigure 1 shows the uncertainty, the range of outcomes covered by the ensemble members in the different seasonal forecast systems separately, while Figure 2 shows the probabilities from all the individual forecast systems for the lower, middle, and upper tercile together. While all systems show the same tendency towards that the upper tercile has the highest probability, there is still a substantial chance for the lower and middle tercile to happen as well. In fact, the probability of not ending up in the upper tercile is still around 50%.\n\nTo answer to the questions stated initially; The generated plots assess the forecasted temperatures with a two months lead time and indicate that the most likely tercile is the upper tercile, but that it is almost equally likely that we will not end up in the upper tercile. In particular, Figure 2 shows that the different seasonal forecast systems provide a consistent outlook. In general, multi-model system forecasts often provide better forecasts and estimates of the uncertainty than applying a single model system (e.g. Doblas-Reyes et al., 2009 \\[[2](https://doi.org/10.1002/qj.464)\\]).\n\nThere are multiple ways to visualize seasonal forecasts and their associated uncertainty as probabilistic forecasts. C3S provides more examples of this as in the graphical products: https://climate.copernicus.eu/charts/packages/c3s_seasonal/", "text_with_prefix": "EQC Quality Assessment: \"Assessing possible outcomes of seasonal temperature forecast\"\nDataset: seasonal-monthly-single-levels [CDS]\nAspect: uncertainty-quality-flags_q01 | Category: Seasonal_Forecasts\nSection: Assessing possible outcomes of seasonal temperature forecast > Analysis and results > 4. Summarise the probabilities for each system\n---\nThis section summarises the forecasted probabilites of the temperature being colder, near average or warmer than their respective hindcast climatologies for all systems.\nThe figure produced reinforces the findings from the figures produced in the last section, that it is more likely to be warmer than usual, compared to the model hindcast climatology than it is to be average or colder compared to the hindcast.\n\nPlotting the multi-ensemble histogram\n\nFigure 1 shows the uncertainty, the range of outcomes covered by the ensemble members in the different seasonal forecast systems separately, while Figure 2 shows the probabilities from all the individual forecast systems for the lower, middle, and upper tercile together. While all systems show the same tendency towards that the upper tercile has the highest probability, there is still a substantial chance for the lower and middle tercile to happen as well. In fact, the probability of not ending up in the upper tercile is still around 50%.\n\nTo answer to the questions stated initially; The generated plots assess the forecasted temperatures with a two months lead time and indicate that the most likely tercile is the upper tercile, but that it is almost equally likely that we will not end up in the upper tercile. In particular, Figure 2 shows that the different seasonal forecast systems provide a consistent outlook. In general, multi-model system forecasts often provide better forecasts and estimates of the uncertainty than applying a single model system (e.g. Doblas-Reyes et al., 2009 \\[[2](https://doi.org/10.1002/qj.464)\\]).\n\nThere are multiple ways to visualize seasonal forecasts and their associated uncertainty as probabilistic forecasts. C3S provides more examples of this as in the graphical products: https://climate.copernicus.eu/charts/packages/c3s_seasonal/"} {"chunk_id": "seasonal_seasonal-monthly-single-levels_uncertainty-quality-flags_q01__75b4ab46f215", "report_id": "seasonal_seasonal-monthly-single-levels_uncertainty-quality-flags_q01", "dataset_id": "seasonal-monthly-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty-quality-flags_q01", "aspect_base": "uncertainty-quality-flags", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Assessing possible outcomes of seasonal temperature forecast > ℹ️ If you want to know more > Key resources", "title": "Assessing possible outcomes of seasonal temperature forecast", "chunk_index": 7, "token_count": 157, "text_raw": "- Seasonal forecast single level anomalies: DOI: [10.24381/cds.7e37c951](https://doi.org/10.24381/cds.7e37c951)\n - Seasonal forecast monthly statistics single level: DOI: [10.24381/cds.68dd14c3](https://doi.org/10.24381/cds.68dd14c3)", "text_with_prefix": "EQC Quality Assessment: \"Assessing possible outcomes of seasonal temperature forecast\"\nDataset: seasonal-monthly-single-levels [CDS]\nAspect: uncertainty-quality-flags_q01 | Category: Seasonal_Forecasts\nSection: Assessing possible outcomes of seasonal temperature forecast > ℹ️ If you want to know more > Key resources\n---\n- Seasonal forecast single level anomalies: DOI: [10.24381/cds.7e37c951](https://doi.org/10.24381/cds.7e37c951)\n - Seasonal forecast monthly statistics single level: DOI: [10.24381/cds.68dd14c3](https://doi.org/10.24381/cds.68dd14c3)"} {"chunk_id": "seasonal_seasonal-monthly-single-levels_uncertainty-quality-flags_q01__9c7f4fafeba8", "report_id": "seasonal_seasonal-monthly-single-levels_uncertainty-quality-flags_q01", "dataset_id": "seasonal-monthly-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty-quality-flags_q01", "aspect_base": "uncertainty-quality-flags", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Assessing possible outcomes of seasonal temperature forecast > ℹ️ If you want to know more > Code libraries used", "title": "Assessing possible outcomes of seasonal temperature forecast", "chunk_index": 8, "token_count": 150, "text_raw": "- xarray\n - numpy\n - express and figure_factory from plotly\n - [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)", "text_with_prefix": "EQC Quality Assessment: \"Assessing possible outcomes of seasonal temperature forecast\"\nDataset: seasonal-monthly-single-levels [CDS]\nAspect: uncertainty-quality-flags_q01 | Category: Seasonal_Forecasts\nSection: Assessing possible outcomes of seasonal temperature forecast > ℹ️ If you want to know more > Code libraries used\n---\n- xarray\n - numpy\n - express and figure_factory from plotly\n - [C3S EQC custom functions](https://github.com/bopen/c3s-eqc-automatic-quality-control/tree/main/c3s_eqc_automatic_quality_control), `c3s_eqc_automatic_quality_control`, prepared by [B-Open](https://www.bopen.eu/)"} {"chunk_id": "seasonal_seasonal-monthly-single-levels_uncertainty-quality-flags_q01__733488b83df5", "report_id": "seasonal_seasonal-monthly-single-levels_uncertainty-quality-flags_q01", "dataset_id": "seasonal-monthly-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "uncertainty-quality-flags_q01", "aspect_base": "uncertainty-quality-flags", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Assessing possible outcomes of seasonal temperature forecast > ℹ️ If you want to know more > References", "title": "Assessing possible outcomes of seasonal temperature forecast", "chunk_index": 9, "token_count": 270, "text_raw": "[[1]](https://doi.org/10.1098/rsif.2013.1162) - Weisheimer A. and Palmer T. N.. 2014 On the reliability of seasonal climate forecasts *J. R. Soc.* Interface. **11:** 20131162.\nhttps://doi.org/10.1098/rsif.2013.1162\n\n[[2]](https://doi.org/10.1002/qj.464) - Doblas-Reyes, F.J., Weisheimer, A., Déqué, M., Keenlyside, N., McVean, M., Murphy, J.M., Rogel, P., Smith, D. and Palmer, T.N. (2009), Addressing model uncertainty in seasonal and annual dynamical ensemble forecasts. Q.J.R. Meteorol. Soc., 135: 1538-1559.\nhttps://doi.org/10.1002/qj.464", "text_with_prefix": "EQC Quality Assessment: \"Assessing possible outcomes of seasonal temperature forecast\"\nDataset: seasonal-monthly-single-levels [CDS]\nAspect: uncertainty-quality-flags_q01 | Category: Seasonal_Forecasts\nSection: Assessing possible outcomes of seasonal temperature forecast > ℹ️ If you want to know more > References\n---\n[[1]](https://doi.org/10.1098/rsif.2013.1162) - Weisheimer A. and Palmer T. N.. 2014 On the reliability of seasonal climate forecasts *J. R. Soc.* Interface. **11:** 20131162.\nhttps://doi.org/10.1098/rsif.2013.1162\n\n[[2]](https://doi.org/10.1002/qj.464) - Doblas-Reyes, F.J., Weisheimer, A., Déqué, M., Keenlyside, N., McVean, M., Murphy, J.M., Rogel, P., Smith, D. and Palmer, T.N. (2009), Addressing model uncertainty in seasonal and annual dynamical ensemble forecasts. Q.J.R. Meteorol. Soc., 135: 1538-1559.\nhttps://doi.org/10.1002/qj.464"} {"chunk_id": "seasonal_seasonal-monthly-single-levels_validation_q05__9527b9d9e14e", "report_id": "seasonal_seasonal-monthly-single-levels_validation_q05", "dataset_id": "seasonal-monthly-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q05", "aspect_base": "validation", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Benefit and challenges of a multi-model approach to seasonal forecasts", "title": "Benefit and challenges of a multi-model approach to seasonal forecasts", "chunk_index": 0, "token_count": 92, "text_raw": "Production date: 07-05-2025\n\nProduced by: Sandro Calmanti, Alessandro Dell'Aquila (ENEA)", "text_with_prefix": "EQC Quality Assessment: \"Benefit and challenges of a multi-model approach to seasonal forecasts\"\nDataset: seasonal-monthly-single-levels [CDS]\nAspect: validation_q05 | Category: Seasonal_Forecasts\nSection: Benefit and challenges of a multi-model approach to seasonal forecasts\n---\nProduction date: 07-05-2025\n\nProduced by: Sandro Calmanti, Alessandro Dell'Aquila (ENEA)"} {"chunk_id": "seasonal_seasonal-monthly-single-levels_validation_q05__b6dd3d23f5c1", "report_id": "seasonal_seasonal-monthly-single-levels_validation_q05", "dataset_id": "seasonal-monthly-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q05", "aspect_base": "validation", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Benefit and challenges of a multi-model approach to seasonal forecasts > Quality assessment questions", "title": "Benefit and challenges of a multi-model approach to seasonal forecasts", "chunk_index": 1, "token_count": 452, "text_raw": "* **Can I improve the skill of seasonal forecasts by using a multi-model ensemble of climate predictions?**\n* **Are there robust strategies for creating multi-model ensembles?**\n\nC3S provides a multi-model ensemble of seasonal climate predictions produced by nine forecast centres in Europe, North America, Japan and Australia. The development of a European multi-model ensemble for seasonal forecasting has been supported throughout several research programs ( [DEMETER](https://cordis.europa.eu/article/id/81812-demeter-improves-european-weather-forecasts) / [ENSEMBLES](https://ensembles-eu.metoffice.gov.uk/docs/Ensembles_final_report_Nov09.pdf) / [EUROSIP](https://www.ecmwf.int/en/forecasts/documentation-and-support/long-range/seasonal-forecast-documentation/eurosip-user-guide/eurosip-operational-history)) which have progressively contributed to the improvement of the underpinning climate models (e.g. [[1]](https://doi.org/10.1029/2009GL040896), [[2]](https://doi.org/10.1029/2007GL030740), [[3]](https://doi.org/10.1007/s00382-011-1061-x) ). The multi-model ensemble available through C3S is therefore the outcome of a long lasting endeavour, enriched with similar predictions derived from other non-european forecasting centres.\n\nIn principle, it is possible to use the seasonal predictions from different centers as a single multi-model ensemble to increase the reliability and the overall value of the forecast.\n\nHowever, handling multi-model ensembles to create climate information on a regular basis, is a complex and resource intensive task. This article summarises the significant experiences in working with multi-model ensembles with the objective of highlighting the potential added value, the challenges and limitations in the use of multi-model ensembles", "text_with_prefix": "EQC Quality Assessment: \"Benefit and challenges of a multi-model approach to seasonal forecasts\"\nDataset: seasonal-monthly-single-levels [CDS]\nAspect: validation_q05 | Category: Seasonal_Forecasts\nSection: Benefit and challenges of a multi-model approach to seasonal forecasts > Quality assessment questions\n---\n* **Can I improve the skill of seasonal forecasts by using a multi-model ensemble of climate predictions?**\n* **Are there robust strategies for creating multi-model ensembles?**\n\nC3S provides a multi-model ensemble of seasonal climate predictions produced by nine forecast centres in Europe, North America, Japan and Australia. The development of a European multi-model ensemble for seasonal forecasting has been supported throughout several research programs ( [DEMETER](https://cordis.europa.eu/article/id/81812-demeter-improves-european-weather-forecasts) / [ENSEMBLES](https://ensembles-eu.metoffice.gov.uk/docs/Ensembles_final_report_Nov09.pdf) / [EUROSIP](https://www.ecmwf.int/en/forecasts/documentation-and-support/long-range/seasonal-forecast-documentation/eurosip-user-guide/eurosip-operational-history)) which have progressively contributed to the improvement of the underpinning climate models (e.g. [[1]](https://doi.org/10.1029/2009GL040896), [[2]](https://doi.org/10.1029/2007GL030740), [[3]](https://doi.org/10.1007/s00382-011-1061-x) ). The multi-model ensemble available through C3S is therefore the outcome of a long lasting endeavour, enriched with similar predictions derived from other non-european forecasting centres.\n\nIn principle, it is possible to use the seasonal predictions from different centers as a single multi-model ensemble to increase the reliability and the overall value of the forecast.\n\nHowever, handling multi-model ensembles to create climate information on a regular basis, is a complex and resource intensive task. This article summarises the significant experiences in working with multi-model ensembles with the objective of highlighting the potential added value, the challenges and limitations in the use of multi-model ensembles"} {"chunk_id": "seasonal_seasonal-monthly-single-levels_validation_q05__0c384596563f", "report_id": "seasonal_seasonal-monthly-single-levels_validation_q05", "dataset_id": "seasonal-monthly-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q05", "aspect_base": "validation", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Benefit and challenges of a multi-model approach to seasonal forecasts > Quality assessment statement", "title": "Benefit and challenges of a multi-model approach to seasonal forecasts", "chunk_index": 2, "token_count": 438, "text_raw": "These are the key outcomes of this assessment\n\n* The multi-model ensemble (MME) approach consistently outperforms individual models in terms of skill metrics such as temporal correlation and probabilistic accuracy, particularly when the combined models are developed independently\n* The use of multi-model ensembles allows for a better assessment of forecast uncertainty and has been shown to support more informed decision-making, especially in hydrological and water resource applications.\n* The MME approach shows the greatest value in tropical regions, where model errors dominate over initial condition uncertainties, enabling effective bias compensation.\n* Sectoral applications can benefit from filtering or weighting the most skillful models for a specific region or variable of interest, such as in national wheat yield forecasting or malaria risk.\n* Bias adjustment and model recalibration are necessary post-processing steps that significantly enhance the reliability of multi-model forecasts, especially when outputs are used as input to impact models.\n```\n\nattachment:cd427a52-cbfe-49fb-9865-21858a87ac2e.png\n---\n\n---\nThe C3S multimodel June 2026 forecast of the of the NINO3.4 index highlighted the likelihood of a large El Nino event developing through the latter part of the year. 75% of members of the grand ensemble exceed 2.5°C amplitude in the Nino3.4 index at the end of the forecast period (November)\n\nPlume charts of NINO predictions are available on the [C3S multi-model SST indices page](https://climate.copernicus.eu/charts/packages/c3s_seasonal/products/c3s_seasonal_plume_mm). Correlation heatmaps for the SST indices from individual models are available [here](https://confluence.ecmwf.int/display/CKB/C3S+seasonal+forecasts+verification+plots)\n```", "text_with_prefix": "EQC Quality Assessment: \"Benefit and challenges of a multi-model approach to seasonal forecasts\"\nDataset: seasonal-monthly-single-levels [CDS]\nAspect: validation_q05 | Category: Seasonal_Forecasts\nSection: Benefit and challenges of a multi-model approach to seasonal forecasts > Quality assessment statement\n---\nThese are the key outcomes of this assessment\n\n* The multi-model ensemble (MME) approach consistently outperforms individual models in terms of skill metrics such as temporal correlation and probabilistic accuracy, particularly when the combined models are developed independently\n* The use of multi-model ensembles allows for a better assessment of forecast uncertainty and has been shown to support more informed decision-making, especially in hydrological and water resource applications.\n* The MME approach shows the greatest value in tropical regions, where model errors dominate over initial condition uncertainties, enabling effective bias compensation.\n* Sectoral applications can benefit from filtering or weighting the most skillful models for a specific region or variable of interest, such as in national wheat yield forecasting or malaria risk.\n* Bias adjustment and model recalibration are necessary post-processing steps that significantly enhance the reliability of multi-model forecasts, especially when outputs are used as input to impact models.\n```\n\nattachment:cd427a52-cbfe-49fb-9865-21858a87ac2e.png\n---\n\n---\nThe C3S multimodel June 2026 forecast of the of the NINO3.4 index highlighted the likelihood of a large El Nino event developing through the latter part of the year. 75% of members of the grand ensemble exceed 2.5°C amplitude in the Nino3.4 index at the end of the forecast period (November)\n\nPlume charts of NINO predictions are available on the [C3S multi-model SST indices page](https://climate.copernicus.eu/charts/packages/c3s_seasonal/products/c3s_seasonal_plume_mm). Correlation heatmaps for the SST indices from individual models are available [here](https://confluence.ecmwf.int/display/CKB/C3S+seasonal+forecasts+verification+plots)\n```"} {"chunk_id": "seasonal_seasonal-monthly-single-levels_validation_q05__3fa691b6e2da", "report_id": "seasonal_seasonal-monthly-single-levels_validation_q05", "dataset_id": "seasonal-monthly-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q05", "aspect_base": "validation", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Benefit and challenges of a multi-model approach to seasonal forecasts > Methodology", "title": "Benefit and challenges of a multi-model approach to seasonal forecasts", "chunk_index": 3, "token_count": 405, "text_raw": "This notebook provides information on the use of seasonal forecast data available under different C3S catalogues, such as monthly statistics on [single levels](https://cds.climate.copernicus.eu/datasets/seasonal-monthly-single-levels?tab=overview) and and on [pressure levels](https://cds.climate.copernicus.eu/datasets/seasonal-monthly-pressure-levels?tab=overview), subdaily data on [single levels](https://cds.climate.copernicus.eu/datasets/seasonal-original-single-levels?tab=overview) and on [pressure levels](https://cds.climate.copernicus.eu/datasets/seasonal-original-pressure-levels?tab=overview), monthly anomalies on [single levels](https://cds.climate.copernicus.eu/datasets/seasonal-postprocessed-single-levels?tab=overview) and on [pressure levels](https://cds.climate.copernicus.eu/datasets/seasonal-postprocessed-pressure-levels?tab=overview).\n\nThe notebook focuses on the review of selected scientific literature on the state-of-the art knowledge and practice in the use of multi-model ensembles.\n\nThe discussion is organized in the following sections:\n\n**[](multimodel:section-1)**\n\n**[](multimodel:section-2)**\n * Health\n * Agriculture\n * Water resource management\n * Energy\n * Disaster preparedness\n * High-impact events\n\n**[](multimodel:section-3)**\n * Bias adjustment\n * Combination\n\n**[](multimodel:section-4)**\n\n**[](multimodel:section-5)**", "text_with_prefix": "EQC Quality Assessment: \"Benefit and challenges of a multi-model approach to seasonal forecasts\"\nDataset: seasonal-monthly-single-levels [CDS]\nAspect: validation_q05 | Category: Seasonal_Forecasts\nSection: Benefit and challenges of a multi-model approach to seasonal forecasts > Methodology\n---\nThis notebook provides information on the use of seasonal forecast data available under different C3S catalogues, such as monthly statistics on [single levels](https://cds.climate.copernicus.eu/datasets/seasonal-monthly-single-levels?tab=overview) and and on [pressure levels](https://cds.climate.copernicus.eu/datasets/seasonal-monthly-pressure-levels?tab=overview), subdaily data on [single levels](https://cds.climate.copernicus.eu/datasets/seasonal-original-single-levels?tab=overview) and on [pressure levels](https://cds.climate.copernicus.eu/datasets/seasonal-original-pressure-levels?tab=overview), monthly anomalies on [single levels](https://cds.climate.copernicus.eu/datasets/seasonal-postprocessed-single-levels?tab=overview) and on [pressure levels](https://cds.climate.copernicus.eu/datasets/seasonal-postprocessed-pressure-levels?tab=overview).\n\nThe notebook focuses on the review of selected scientific literature on the state-of-the art knowledge and practice in the use of multi-model ensembles.\n\nThe discussion is organized in the following sections:\n\n**[](multimodel:section-1)**\n\n**[](multimodel:section-2)**\n * Health\n * Agriculture\n * Water resource management\n * Energy\n * Disaster preparedness\n * High-impact events\n\n**[](multimodel:section-3)**\n * Bias adjustment\n * Combination\n\n**[](multimodel:section-4)**\n\n**[](multimodel:section-5)**"} {"chunk_id": "seasonal_seasonal-monthly-single-levels_validation_q05__2270786a465d", "report_id": "seasonal_seasonal-monthly-single-levels_validation_q05", "dataset_id": "seasonal-monthly-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q05", "aspect_base": "validation", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Benefit and challenges of a multi-model approach to seasonal forecasts > Analysis and results > 1. Why the multi-model ensemble (MME)", "title": "Benefit and challenges of a multi-model approach to seasonal forecasts", "chunk_index": 4, "token_count": 508, "text_raw": "There are two main sources of error which affect the performance of seasonal climate predictions in terms of reliability, accuracy and ultimately their usability and value for sectoral applications: \n * the initial condition uncertainties associated to the knowledge of the current state of the climate system, which is the starting point of a forecast [[4]](https://doi.org/10.1002/qj.49712252905); \n * the errors introduced by climate models, which describe the climate system by means of a finite number - albeit an incredibly large one - of interacting elements and by a set of empirical parametrizations and approximations of physical processes [[5]](https://doi.org/10.1002/qj.49712757202).\n\nThe first source of error is, to a large extent, unavoidable because it is associated with the limits of the global observing system and to the techniques adopted to extract information from observations. This limit is generally addressed by producing the largest possible ensembles of forecasts with similar initial conditions. The objective of producing large ensembles of forecasts is not necessarily, and only, the elimination or reduction of the forecast uncertainty. Instead, large ensembles aim at providing more robust statistics and produce the largest range of possible outcomes associated with uncertainty of the state of the climate system.\n\nThe second source of errors depends on the overall modelling approach, on the parameterizations adopted for some physical processes in the climate system, and on the specific solutions adopted for implementation of each climate modelling system. A fundamental goal of climate modelers is to reduce this kind of error as much as possible, in order to avoid the systematic misrepresentation of key features of the climate system (the so-called model bias), and the associated impact on essential climate variables such as precipitation intensity, temperature, or wind patterns.\n\nThe MME approach addresses the second source of error by leveraging the underlying assumption that seasonal predictions produced by different forecasting centers can be considered as the members of a single statistical ensemble. A further assumption is that the systematic errors produced by different climate models can partially compensate each other if the technical choices made for their development are sufficiently independent.\n\n(multimodel:section-2)=", "text_with_prefix": "EQC Quality Assessment: \"Benefit and challenges of a multi-model approach to seasonal forecasts\"\nDataset: seasonal-monthly-single-levels [CDS]\nAspect: validation_q05 | Category: Seasonal_Forecasts\nSection: Benefit and challenges of a multi-model approach to seasonal forecasts > Analysis and results > 1. Why the multi-model ensemble (MME)\n---\nThere are two main sources of error which affect the performance of seasonal climate predictions in terms of reliability, accuracy and ultimately their usability and value for sectoral applications: \n * the initial condition uncertainties associated to the knowledge of the current state of the climate system, which is the starting point of a forecast [[4]](https://doi.org/10.1002/qj.49712252905); \n * the errors introduced by climate models, which describe the climate system by means of a finite number - albeit an incredibly large one - of interacting elements and by a set of empirical parametrizations and approximations of physical processes [[5]](https://doi.org/10.1002/qj.49712757202).\n\nThe first source of error is, to a large extent, unavoidable because it is associated with the limits of the global observing system and to the techniques adopted to extract information from observations. This limit is generally addressed by producing the largest possible ensembles of forecasts with similar initial conditions. The objective of producing large ensembles of forecasts is not necessarily, and only, the elimination or reduction of the forecast uncertainty. Instead, large ensembles aim at providing more robust statistics and produce the largest range of possible outcomes associated with uncertainty of the state of the climate system.\n\nThe second source of errors depends on the overall modelling approach, on the parameterizations adopted for some physical processes in the climate system, and on the specific solutions adopted for implementation of each climate modelling system. A fundamental goal of climate modelers is to reduce this kind of error as much as possible, in order to avoid the systematic misrepresentation of key features of the climate system (the so-called model bias), and the associated impact on essential climate variables such as precipitation intensity, temperature, or wind patterns.\n\nThe MME approach addresses the second source of error by leveraging the underlying assumption that seasonal predictions produced by different forecasting centers can be considered as the members of a single statistical ensemble. A further assumption is that the systematic errors produced by different climate models can partially compensate each other if the technical choices made for their development are sufficiently independent.\n\n(multimodel:section-2)="} {"chunk_id": "seasonal_seasonal-monthly-single-levels_validation_q05__4ec467de529a", "report_id": "seasonal_seasonal-monthly-single-levels_validation_q05", "dataset_id": "seasonal-monthly-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q05", "aspect_base": "validation", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Benefit and challenges of a multi-model approach to seasonal forecasts > Analysis and results > 2. Use cases and user needs", "title": "Benefit and challenges of a multi-model approach to seasonal forecasts", "chunk_index": 5, "token_count": 177, "text_raw": "The MME approach to seasonal forecasting has already been tested in a number of applications to demonstrate the advantages over using a single modelling system.\n\nClimate information is designed to meet specific user needs and a one-size-fits-all solution, rather than a systematic collection of existing applications, would be of limited use, if at all possible.\n\nTherefore, a few examples of sectoral applications of the multi-model ensemble approach are provided with the purpose of illustrating real use cases and describe the potential added value of the MME approach.", "text_with_prefix": "EQC Quality Assessment: \"Benefit and challenges of a multi-model approach to seasonal forecasts\"\nDataset: seasonal-monthly-single-levels [CDS]\nAspect: validation_q05 | Category: Seasonal_Forecasts\nSection: Benefit and challenges of a multi-model approach to seasonal forecasts > Analysis and results > 2. Use cases and user needs\n---\nThe MME approach to seasonal forecasting has already been tested in a number of applications to demonstrate the advantages over using a single modelling system.\n\nClimate information is designed to meet specific user needs and a one-size-fits-all solution, rather than a systematic collection of existing applications, would be of limited use, if at all possible.\n\nTherefore, a few examples of sectoral applications of the multi-model ensemble approach are provided with the purpose of illustrating real use cases and describe the potential added value of the MME approach."} {"chunk_id": "seasonal_seasonal-monthly-single-levels_validation_q05__2a96ada619b1", "report_id": "seasonal_seasonal-monthly-single-levels_validation_q05", "dataset_id": "seasonal-monthly-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q05", "aspect_base": "validation", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Benefit and challenges of a multi-model approach to seasonal forecasts > Analysis and results > 2. Use cases and user needs > Health", "title": "Benefit and challenges of a multi-model approach to seasonal forecasts", "chunk_index": 6, "token_count": 254, "text_raw": "An early assessment of the added value of the multi-model approach has been conducted by focusing on malaria early warnings for southern Africa, using the models participating in the DEMETER project which still constitute the core of the C3S multi-model [[6]](https://doi.org/10.1038/nature04503) [[7]](https://doi.org/10.3402/tellusa.v57i3.14668). To generate the MME the difference in means and ratio of variances between each model and a reference dataset is computed and applied to the predictions of the target year.\nIn the DEMETER multi system, the fact that the sub-grid parameterizations in the component models have, to a large extent, been developed independently, is explicitly mentioned as a varied source of model uncertainty associated with the numerical approximation to the underlying partial differential equations that govern climate models.", "text_with_prefix": "EQC Quality Assessment: \"Benefit and challenges of a multi-model approach to seasonal forecasts\"\nDataset: seasonal-monthly-single-levels [CDS]\nAspect: validation_q05 | Category: Seasonal_Forecasts\nSection: Benefit and challenges of a multi-model approach to seasonal forecasts > Analysis and results > 2. Use cases and user needs > Health\n---\nAn early assessment of the added value of the multi-model approach has been conducted by focusing on malaria early warnings for southern Africa, using the models participating in the DEMETER project which still constitute the core of the C3S multi-model [[6]](https://doi.org/10.1038/nature04503) [[7]](https://doi.org/10.3402/tellusa.v57i3.14668). To generate the MME the difference in means and ratio of variances between each model and a reference dataset is computed and applied to the predictions of the target year.\nIn the DEMETER multi system, the fact that the sub-grid parameterizations in the component models have, to a large extent, been developed independently, is explicitly mentioned as a varied source of model uncertainty associated with the numerical approximation to the underlying partial differential equations that govern climate models."} {"chunk_id": "seasonal_seasonal-monthly-single-levels_validation_q05__4182c5bf35b1", "report_id": "seasonal_seasonal-monthly-single-levels_validation_q05", "dataset_id": "seasonal-monthly-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q05", "aspect_base": "validation", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Benefit and challenges of a multi-model approach to seasonal forecasts > Analysis and results > 2. Use cases and user needs > Agriculture", "title": "Benefit and challenges of a multi-model approach to seasonal forecasts", "chunk_index": 7, "token_count": 691, "text_raw": "At the global scale, two alternative approaches (average method and mosaic method) have been assessed for the forecasting of year-to-year variations in the global yield of key crop commodities such as maize, rice, wheat and soybean. The ROC score analysis of within-season national yield predictions for maize, rice, wheat, and soybean based on mosaic method indicate skillful within-season predictions, especially for maize and wheat [[8]](https://doi.org/10.1016/j.cliser.2018.06.003).\n\nThe average method considers a simple average, with equal weights, over multiple forecasting systems for each location and cropping season. Within the mosaic method, the single best-performing forecasting system is selected for each location and cropping season, based on the corresponding grid-based skill score for yield variability. This study uses only one of the models available in C3S (NCEP) along with other SF available from other forecasting centres (APEC Climate Center, Korea; Meteorological Service of Canada; NASA; Pusan National University, Korea). For this application the mosaic method outperforms the average method and, by definition, it also outperforms the performance of individual forecasting systems. However, it may not be suitable for applications requiring large-scale consistency in forecasts, such as hydrological modeling (see below), where maintaining water mass conservation is essential for accurate streamflow predictions. In this case, the simple averaging of multi-model members has been adopted as a more relevant approach [[9]](https://doi.org/10.5194/egusphere-2023-569) .\n\nOn a more regional scale, a tailored multi-model ensemble has been developed to forecast national wheat yield in Argentina [[10]](https://doi.org/10.1088/1748-9326/ad627c).\n\nattachment:08031f31-1537-434a-a742-1d056f529742.png\n---\nheight: 400px\n---\nExample of mean absolute error between forecasted and reanalysis climate indicators for different forecasting systems and month of initialization. Column-wise for each month of initialization, the mean absolute error is expressed in units of standard deviation (σ), given that all features underwent standardization during preprocessing. Reproduced from [[Zachow et al. (2024]](https://doi.org/10.1088/1748-9326/ad627c) under [[CC BY 4.0]](https://creativecommons.org/licenses/by/4.0/).\n```\n\nThis study considers all but one of the forecasting systems available at C3S and those available through the North-American-Multi-Model Ensemble (NMME). In this case, the multi-model ensemble is built by filtering out the group of three best-performing models for the region and for crop modelling application of interest. Interestingly, out of the initial set of 10 forecasting systems, the final best-performing subset of three systems is a mix of elements coming from both the C3S and of the NMME ensembles.", "text_with_prefix": "EQC Quality Assessment: \"Benefit and challenges of a multi-model approach to seasonal forecasts\"\nDataset: seasonal-monthly-single-levels [CDS]\nAspect: validation_q05 | Category: Seasonal_Forecasts\nSection: Benefit and challenges of a multi-model approach to seasonal forecasts > Analysis and results > 2. Use cases and user needs > Agriculture\n---\nAt the global scale, two alternative approaches (average method and mosaic method) have been assessed for the forecasting of year-to-year variations in the global yield of key crop commodities such as maize, rice, wheat and soybean. The ROC score analysis of within-season national yield predictions for maize, rice, wheat, and soybean based on mosaic method indicate skillful within-season predictions, especially for maize and wheat [[8]](https://doi.org/10.1016/j.cliser.2018.06.003).\n\nThe average method considers a simple average, with equal weights, over multiple forecasting systems for each location and cropping season. Within the mosaic method, the single best-performing forecasting system is selected for each location and cropping season, based on the corresponding grid-based skill score for yield variability. This study uses only one of the models available in C3S (NCEP) along with other SF available from other forecasting centres (APEC Climate Center, Korea; Meteorological Service of Canada; NASA; Pusan National University, Korea). For this application the mosaic method outperforms the average method and, by definition, it also outperforms the performance of individual forecasting systems. However, it may not be suitable for applications requiring large-scale consistency in forecasts, such as hydrological modeling (see below), where maintaining water mass conservation is essential for accurate streamflow predictions. In this case, the simple averaging of multi-model members has been adopted as a more relevant approach [[9]](https://doi.org/10.5194/egusphere-2023-569) .\n\nOn a more regional scale, a tailored multi-model ensemble has been developed to forecast national wheat yield in Argentina [[10]](https://doi.org/10.1088/1748-9326/ad627c).\n\nattachment:08031f31-1537-434a-a742-1d056f529742.png\n---\nheight: 400px\n---\nExample of mean absolute error between forecasted and reanalysis climate indicators for different forecasting systems and month of initialization. Column-wise for each month of initialization, the mean absolute error is expressed in units of standard deviation (σ), given that all features underwent standardization during preprocessing. Reproduced from [[Zachow et al. (2024]](https://doi.org/10.1088/1748-9326/ad627c) under [[CC BY 4.0]](https://creativecommons.org/licenses/by/4.0/).\n```\n\nThis study considers all but one of the forecasting systems available at C3S and those available through the North-American-Multi-Model Ensemble (NMME). In this case, the multi-model ensemble is built by filtering out the group of three best-performing models for the region and for crop modelling application of interest. Interestingly, out of the initial set of 10 forecasting systems, the final best-performing subset of three systems is a mix of elements coming from both the C3S and of the NMME ensembles."} {"chunk_id": "seasonal_seasonal-monthly-single-levels_validation_q05__ac8017ae1155", "report_id": "seasonal_seasonal-monthly-single-levels_validation_q05", "dataset_id": "seasonal-monthly-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q05", "aspect_base": "validation", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Benefit and challenges of a multi-model approach to seasonal forecasts > Analysis and results > 2. Use cases and user needs > Water resource management", "title": "Benefit and challenges of a multi-model approach to seasonal forecasts", "chunk_index": 8, "token_count": 217, "text_raw": "C3S has supported the development of an End-to-end Demonstrator for Improved Decision Making in the water sector in Europe (EdgE, [[11]](https://doi.org/10.1175/JHM-D-18-0040.1) ) based on the implementation of a multi-model prediction of streamflow at the European scale. In this case, four different climate modelling systems have been considered, in combination with four different approaches for the computation of the streamflow associated with the seasonal forecasts. This study emphasises how the multi-model approach allows for a better assessment of the uncertainty in the forecast and therefore a better framing of the value of the prediction for decision making.", "text_with_prefix": "EQC Quality Assessment: \"Benefit and challenges of a multi-model approach to seasonal forecasts\"\nDataset: seasonal-monthly-single-levels [CDS]\nAspect: validation_q05 | Category: Seasonal_Forecasts\nSection: Benefit and challenges of a multi-model approach to seasonal forecasts > Analysis and results > 2. Use cases and user needs > Water resource management\n---\nC3S has supported the development of an End-to-end Demonstrator for Improved Decision Making in the water sector in Europe (EdgE, [[11]](https://doi.org/10.1175/JHM-D-18-0040.1) ) based on the implementation of a multi-model prediction of streamflow at the European scale. In this case, four different climate modelling systems have been considered, in combination with four different approaches for the computation of the streamflow associated with the seasonal forecasts. This study emphasises how the multi-model approach allows for a better assessment of the uncertainty in the forecast and therefore a better framing of the value of the prediction for decision making."} {"chunk_id": "seasonal_seasonal-monthly-single-levels_validation_q05__f847d6bf357c", "report_id": "seasonal_seasonal-monthly-single-levels_validation_q05", "dataset_id": "seasonal-monthly-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q05", "aspect_base": "validation", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Benefit and challenges of a multi-model approach to seasonal forecasts > Analysis and results > 2. Use cases and user needs > Energy", "title": "Benefit and challenges of a multi-model approach to seasonal forecasts", "chunk_index": 9, "token_count": 287, "text_raw": "The multi-model seasonal forecast approach has been tested for applications in the energy sector, both for energy production from renewable energy [[12]](https://doi.org/10.1007/s00382-019-04654-y) and to forecast energy demand [[13]](https://doi.org/10.1007/s00382-017-3766-y) at the national level.\nIt has been demonstrated that:\n * MME predictions indicate consistently higher performance than individual models in terms of different skill metrics such as temporal correlation coefficient (TCC) [[14]](https://doi.org/10.1002/2014WR015426) and fair ranked probability skill score (FRPSS) [[12]](https://doi.org/10.1007/s00382-019-04654-y);\n * the performance of a multi-model ensemble increases when using climate models which are as independent as possible from each other [[13]](https://doi.org/10.1007/s00382-017-3766-y).", "text_with_prefix": "EQC Quality Assessment: \"Benefit and challenges of a multi-model approach to seasonal forecasts\"\nDataset: seasonal-monthly-single-levels [CDS]\nAspect: validation_q05 | Category: Seasonal_Forecasts\nSection: Benefit and challenges of a multi-model approach to seasonal forecasts > Analysis and results > 2. Use cases and user needs > Energy\n---\nThe multi-model seasonal forecast approach has been tested for applications in the energy sector, both for energy production from renewable energy [[12]](https://doi.org/10.1007/s00382-019-04654-y) and to forecast energy demand [[13]](https://doi.org/10.1007/s00382-017-3766-y) at the national level.\nIt has been demonstrated that:\n * MME predictions indicate consistently higher performance than individual models in terms of different skill metrics such as temporal correlation coefficient (TCC) [[14]](https://doi.org/10.1002/2014WR015426) and fair ranked probability skill score (FRPSS) [[12]](https://doi.org/10.1007/s00382-019-04654-y);\n * the performance of a multi-model ensemble increases when using climate models which are as independent as possible from each other [[13]](https://doi.org/10.1007/s00382-017-3766-y)."} {"chunk_id": "seasonal_seasonal-monthly-single-levels_validation_q05__d19430f92ba0", "report_id": "seasonal_seasonal-monthly-single-levels_validation_q05", "dataset_id": "seasonal-monthly-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q05", "aspect_base": "validation", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Benefit and challenges of a multi-model approach to seasonal forecasts > Analysis and results > 2. Use cases and user needs > Disaster preparedness", "title": "Benefit and challenges of a multi-model approach to seasonal forecasts", "chunk_index": 10, "token_count": 556, "text_raw": "The use of multi-model seasonal forecasts has also emerged as a promising approach to enhance disaster preparedness and management efforts.\n\nFor example, advanced flood preparedness in Perù has leveraged multi-model seasonal climate predictions [[15]](https://doi.org/10.5194/nhess-21-2215-2021) by adopting a mix of best-performing climate model selection and model combination, starting from the North-America Multi-Model Ensemble NMME. The correlation between spring (FMA) precipitation and streamflow (the key impact indicator for this use case) is around 0.76 for individual models whereas it increases to 0.84 when averaging the outcome of the best-performing models from NMME.\n\nA multi-model ensemble composed of 7 out of the 13 systems included in the NMME has been calibrated to enhance climate services in Ethiopia [[16]](https://doi.org/10.1016/j.cliser.2021.100272), whose National Meteorological Agency is particularly focused on issuing drought early warnings and supporting disaster preparedness and management. In this case, the multi-model ensemble is built by first correcting the bias in each model (i.e. calibration) and then creating a super-ensemble where each model is assigned the same weight independently of the predictive skill.\n\nattachment:4c2f6447-b2e3-47d7-9e36-7492a59958f4.png\n---\nheight: 400px\n---\nJJAS (Kiremt rainy season) precipitation anomalies in Ethiopia for 2020 (a) forecasted (b) observed. Unit of precipitation anomaly is mm/season. Reproduced from [Acharya et al. (2021)](https://doi.org/10.1016/j.cliser.2021.100272) under [CC BY-NC-ND 4.0](https://creativecommons.org/licenses/by-nc-nd/4.0/).\n```\n\nThe analysis of the added value of the multi-model prediction compared to individual models is not explicitly reported. However the study emphasises the main source of predictability in the region is ENSO, whose prediction is known to benefit significantly from the multi-model approach [[17]](https://doi.org/10.1175/2008MWR2431.1).", "text_with_prefix": "EQC Quality Assessment: \"Benefit and challenges of a multi-model approach to seasonal forecasts\"\nDataset: seasonal-monthly-single-levels [CDS]\nAspect: validation_q05 | Category: Seasonal_Forecasts\nSection: Benefit and challenges of a multi-model approach to seasonal forecasts > Analysis and results > 2. Use cases and user needs > Disaster preparedness\n---\nThe use of multi-model seasonal forecasts has also emerged as a promising approach to enhance disaster preparedness and management efforts.\n\nFor example, advanced flood preparedness in Perù has leveraged multi-model seasonal climate predictions [[15]](https://doi.org/10.5194/nhess-21-2215-2021) by adopting a mix of best-performing climate model selection and model combination, starting from the North-America Multi-Model Ensemble NMME. The correlation between spring (FMA) precipitation and streamflow (the key impact indicator for this use case) is around 0.76 for individual models whereas it increases to 0.84 when averaging the outcome of the best-performing models from NMME.\n\nA multi-model ensemble composed of 7 out of the 13 systems included in the NMME has been calibrated to enhance climate services in Ethiopia [[16]](https://doi.org/10.1016/j.cliser.2021.100272), whose National Meteorological Agency is particularly focused on issuing drought early warnings and supporting disaster preparedness and management. In this case, the multi-model ensemble is built by first correcting the bias in each model (i.e. calibration) and then creating a super-ensemble where each model is assigned the same weight independently of the predictive skill.\n\nattachment:4c2f6447-b2e3-47d7-9e36-7492a59958f4.png\n---\nheight: 400px\n---\nJJAS (Kiremt rainy season) precipitation anomalies in Ethiopia for 2020 (a) forecasted (b) observed. Unit of precipitation anomaly is mm/season. Reproduced from [Acharya et al. (2021)](https://doi.org/10.1016/j.cliser.2021.100272) under [CC BY-NC-ND 4.0](https://creativecommons.org/licenses/by-nc-nd/4.0/).\n```\n\nThe analysis of the added value of the multi-model prediction compared to individual models is not explicitly reported. However the study emphasises the main source of predictability in the region is ENSO, whose prediction is known to benefit significantly from the multi-model approach [[17]](https://doi.org/10.1175/2008MWR2431.1)."} {"chunk_id": "seasonal_seasonal-monthly-single-levels_validation_q05__91867f836d61", "report_id": "seasonal_seasonal-monthly-single-levels_validation_q05", "dataset_id": "seasonal-monthly-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q05", "aspect_base": "validation", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Benefit and challenges of a multi-model approach to seasonal forecasts > Analysis and results > 2. Use cases and user needs > High-impact events", "title": "Benefit and challenges of a multi-model approach to seasonal forecasts", "chunk_index": 11, "token_count": 317, "text_raw": "The prediction of tropical storm frequency also benefits from the adoption of the multi-model approach as demonstrated using seven models participating in the DEMETER project [[18]](https://doi.org/10.1256/qj.05.65). By adding all ensemble forecasts after calibration, this analysis demonstrates that over specific regions, combining several models leads to better forecasts than the best individual model.\n\nThe analysis conducted during the ENSEMBLES project has demonstrated that the skill in forecasting different indices for temperature and rainfall extremes improves with a multi-model approach, compared to any individual model [[19]](https://doi.org/10.1016/j.wace.2015.06.005). This study makes limited, if any, reference to the added value for specific sectoral applications.\n\nAlso, the most significant improvements are mainly detected over the tropical ocean in the ENSO region. However, the study provides a key support for the development of applications that rely on the extracting information from the teleconnection between ENSO and local climate variables [[17]](https://doi.org/10.1175/2008MWR2431.1).\n\n(multimodel:section-3)=", "text_with_prefix": "EQC Quality Assessment: \"Benefit and challenges of a multi-model approach to seasonal forecasts\"\nDataset: seasonal-monthly-single-levels [CDS]\nAspect: validation_q05 | Category: Seasonal_Forecasts\nSection: Benefit and challenges of a multi-model approach to seasonal forecasts > Analysis and results > 2. Use cases and user needs > High-impact events\n---\nThe prediction of tropical storm frequency also benefits from the adoption of the multi-model approach as demonstrated using seven models participating in the DEMETER project [[18]](https://doi.org/10.1256/qj.05.65). By adding all ensemble forecasts after calibration, this analysis demonstrates that over specific regions, combining several models leads to better forecasts than the best individual model.\n\nThe analysis conducted during the ENSEMBLES project has demonstrated that the skill in forecasting different indices for temperature and rainfall extremes improves with a multi-model approach, compared to any individual model [[19]](https://doi.org/10.1016/j.wace.2015.06.005). This study makes limited, if any, reference to the added value for specific sectoral applications.\n\nAlso, the most significant improvements are mainly detected over the tropical ocean in the ENSO region. However, the study provides a key support for the development of applications that rely on the extracting information from the teleconnection between ENSO and local climate variables [[17]](https://doi.org/10.1175/2008MWR2431.1).\n\n(multimodel:section-3)="} {"chunk_id": "seasonal_seasonal-monthly-single-levels_validation_q05__da7a41d1927a", "report_id": "seasonal_seasonal-monthly-single-levels_validation_q05", "dataset_id": "seasonal-monthly-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q05", "aspect_base": "validation", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Benefit and challenges of a multi-model approach to seasonal forecasts > Analysis and results > 3. Technical aspects in using the MME approach", "title": "Benefit and challenges of a multi-model approach to seasonal forecasts", "chunk_index": 12, "token_count": 273, "text_raw": "The MME approach is designed to address multiple sources of climate model uncertainty, including not only differences in the numerical approximations to the governing physical equations, but also variations arising from ensemble size, sampling of initial condition uncertainty, data assimilation techniques, and other aspects of the modeling systems. By combining outputs from multiple models, the MME approach aims to provide more robust and reliable projections than single models alone.\n\nTwo main groups of post-processing approaches are adopted to implement the MME approach:\n\n* the bias adjustment and recalibration methods applied to the output of individual forecasting systems [[20]](https://doi.org/10.1007/s00382-019-04640-4);\n * the combination of predictions issued with different forecasting systems [[21]](https://doi.org/10.1126/science.285.5433.1548) [[22]](https://doi.org/10.3402/tellusa.v57i3.14657).", "text_with_prefix": "EQC Quality Assessment: \"Benefit and challenges of a multi-model approach to seasonal forecasts\"\nDataset: seasonal-monthly-single-levels [CDS]\nAspect: validation_q05 | Category: Seasonal_Forecasts\nSection: Benefit and challenges of a multi-model approach to seasonal forecasts > Analysis and results > 3. Technical aspects in using the MME approach\n---\nThe MME approach is designed to address multiple sources of climate model uncertainty, including not only differences in the numerical approximations to the governing physical equations, but also variations arising from ensemble size, sampling of initial condition uncertainty, data assimilation techniques, and other aspects of the modeling systems. By combining outputs from multiple models, the MME approach aims to provide more robust and reliable projections than single models alone.\n\nTwo main groups of post-processing approaches are adopted to implement the MME approach:\n\n* the bias adjustment and recalibration methods applied to the output of individual forecasting systems [[20]](https://doi.org/10.1007/s00382-019-04640-4);\n * the combination of predictions issued with different forecasting systems [[21]](https://doi.org/10.1126/science.285.5433.1548) [[22]](https://doi.org/10.3402/tellusa.v57i3.14657)."} {"chunk_id": "seasonal_seasonal-monthly-single-levels_validation_q05__27c42644985e", "report_id": "seasonal_seasonal-monthly-single-levels_validation_q05", "dataset_id": "seasonal-monthly-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q05", "aspect_base": "validation", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Benefit and challenges of a multi-model approach to seasonal forecasts > Analysis and results > 3. Technical aspects in using the MME approach > Bias adjustment", "title": "Benefit and challenges of a multi-model approach to seasonal forecasts", "chunk_index": 13, "token_count": 289, "text_raw": "Bias adjustment and recalibration methods have been tested systematically on the datasets available on the Copernicus Data Store [[21]](https://doi.org/10.1126/science.285.5433.1548).\n\nOn the other hand, the multi-model approach based on combining predictions from different systems assumes that climate models are sufficiently independent to improve or at least partially compensate for the respective errors.\n\nIt has been demonstrated, using both toy-models and actual climate model simulations, that multi-model ensembles can outperform a ‘best-model’ approach because multi-model combinations reduce the average ensemble mean error at the cost of widening the spread of the overall ensemble [[23]](https://doi.org/10.1002/qj.210).\n\nA systematic, effective approach for the creation of multimodel ensemble, tested on C3S includes, recalibration and the equally weighting of all members of the multi-model ensemble [[24]](https://doi.org/10.1007/s00382-020-05314-2).", "text_with_prefix": "EQC Quality Assessment: \"Benefit and challenges of a multi-model approach to seasonal forecasts\"\nDataset: seasonal-monthly-single-levels [CDS]\nAspect: validation_q05 | Category: Seasonal_Forecasts\nSection: Benefit and challenges of a multi-model approach to seasonal forecasts > Analysis and results > 3. Technical aspects in using the MME approach > Bias adjustment\n---\nBias adjustment and recalibration methods have been tested systematically on the datasets available on the Copernicus Data Store [[21]](https://doi.org/10.1126/science.285.5433.1548).\n\nOn the other hand, the multi-model approach based on combining predictions from different systems assumes that climate models are sufficiently independent to improve or at least partially compensate for the respective errors.\n\nIt has been demonstrated, using both toy-models and actual climate model simulations, that multi-model ensembles can outperform a ‘best-model’ approach because multi-model combinations reduce the average ensemble mean error at the cost of widening the spread of the overall ensemble [[23]](https://doi.org/10.1002/qj.210).\n\nA systematic, effective approach for the creation of multimodel ensemble, tested on C3S includes, recalibration and the equally weighting of all members of the multi-model ensemble [[24]](https://doi.org/10.1007/s00382-020-05314-2)."} {"chunk_id": "seasonal_seasonal-monthly-single-levels_validation_q05__616961b6a433", "report_id": "seasonal_seasonal-monthly-single-levels_validation_q05", "dataset_id": "seasonal-monthly-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q05", "aspect_base": "validation", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Benefit and challenges of a multi-model approach to seasonal forecasts > Analysis and results > 3. Technical aspects in using the MME approach > Combination", "title": "Benefit and challenges of a multi-model approach to seasonal forecasts", "chunk_index": 14, "token_count": 409, "text_raw": "In general, the multi-model approach is expected to bring much of its added value in the tropics, where the combination of models in a single grand-ensemble can offset errors affecting the predictable components of the climate system [[25]](https://doi.org/10.3402/tellusa.v57i3.14658).\n\nIn extratropical regions, e.g. over Europe, the multi-model does not always perform better than the best single models, therefore it is advisable to test the single models for each application [[26]](https://doi.org/10.1007/s00382-018-4404-z). The figure below shows an example of how different models can be more skillful over different areas.\n\nSeasonal forecasting studies employ different approaches to generate multi‐model ensembles. As examples:\n\n- simple equal‐weight pooling appears in four studies [[25]](https://doi.org/10.3402/tellusa.v57i3.14658) [[24]](https://doi.org/10.1007/s00382-020-05314-2);\n\n- regression‐based optimal weighting [[26]](https://doi.org/10.1007/s00382-018-4404-z) [[25]](https://doi.org/10.3402/tellusa.v57i3.14658);\n\n- bayesian methods [[27]](https://doi.org/10.1175/MWR2818.1);\n\n- performance-based objective weighting approaches [[28]](https://doi.org/10.1175/BAMS-84-12-1783).", "text_with_prefix": "EQC Quality Assessment: \"Benefit and challenges of a multi-model approach to seasonal forecasts\"\nDataset: seasonal-monthly-single-levels [CDS]\nAspect: validation_q05 | Category: Seasonal_Forecasts\nSection: Benefit and challenges of a multi-model approach to seasonal forecasts > Analysis and results > 3. Technical aspects in using the MME approach > Combination\n---\nIn general, the multi-model approach is expected to bring much of its added value in the tropics, where the combination of models in a single grand-ensemble can offset errors affecting the predictable components of the climate system [[25]](https://doi.org/10.3402/tellusa.v57i3.14658).\n\nIn extratropical regions, e.g. over Europe, the multi-model does not always perform better than the best single models, therefore it is advisable to test the single models for each application [[26]](https://doi.org/10.1007/s00382-018-4404-z). The figure below shows an example of how different models can be more skillful over different areas.\n\nSeasonal forecasting studies employ different approaches to generate multi‐model ensembles. As examples:\n\n- simple equal‐weight pooling appears in four studies [[25]](https://doi.org/10.3402/tellusa.v57i3.14658) [[24]](https://doi.org/10.1007/s00382-020-05314-2);\n\n- regression‐based optimal weighting [[26]](https://doi.org/10.1007/s00382-018-4404-z) [[25]](https://doi.org/10.3402/tellusa.v57i3.14658);\n\n- bayesian methods [[27]](https://doi.org/10.1175/MWR2818.1);\n\n- performance-based objective weighting approaches [[28]](https://doi.org/10.1175/BAMS-84-12-1783)."} {"chunk_id": "seasonal_seasonal-monthly-single-levels_validation_q05__ae8fb89137f8", "report_id": "seasonal_seasonal-monthly-single-levels_validation_q05", "dataset_id": "seasonal-monthly-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q05", "aspect_base": "validation", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Benefit and challenges of a multi-model approach to seasonal forecasts > Analysis and results > 3. Technical aspects in using the MME approach > Complementarity of bias adjustment and combination", "title": "Benefit and challenges of a multi-model approach to seasonal forecasts", "chunk_index": 15, "token_count": 327, "text_raw": "While bias correction and model combination are sometimes discussed separately, they are in fact complementary components of the multi-model ensemble framework rather than competing alternatives. Bias correction improves the performance of individual models, ensuring that each contributes more reliable information to the ensemble. Model combination, in turn, integrates the corrected forecasts from multiple systems to capture a broader range of plausible outcomes and reduce the influence of any single model’s systematic deficiencies.\n\nThis complementarity between bias correction and model combination is also supported by comparison of the predictive skill of individual seasonal forecasting systems, which varies substantially across regions, variables, and seasons, with different models achieving the highest anomaly correlation in different areas [[26]](https://doi.org/10.1007/s00382-018-4404-z). Such results highlight that no single forecasting system consistently outperforms all others. Consequently, combining forecasts from multiple models provides an opportunity to exploit the strengths of each system while mitigating their individual weaknesses. Within this framework, bias correction enhances the reliability of the forecasts produced by each model, whereas model combination synthesizes the corrected predictions to generate a more robust and skillful ensemble forecast. Together, these approaches may contribute to improving both forecast calibration and predictive performance.", "text_with_prefix": "EQC Quality Assessment: \"Benefit and challenges of a multi-model approach to seasonal forecasts\"\nDataset: seasonal-monthly-single-levels [CDS]\nAspect: validation_q05 | Category: Seasonal_Forecasts\nSection: Benefit and challenges of a multi-model approach to seasonal forecasts > Analysis and results > 3. Technical aspects in using the MME approach > Complementarity of bias adjustment and combination\n---\nWhile bias correction and model combination are sometimes discussed separately, they are in fact complementary components of the multi-model ensemble framework rather than competing alternatives. Bias correction improves the performance of individual models, ensuring that each contributes more reliable information to the ensemble. Model combination, in turn, integrates the corrected forecasts from multiple systems to capture a broader range of plausible outcomes and reduce the influence of any single model’s systematic deficiencies.\n\nThis complementarity between bias correction and model combination is also supported by comparison of the predictive skill of individual seasonal forecasting systems, which varies substantially across regions, variables, and seasons, with different models achieving the highest anomaly correlation in different areas [[26]](https://doi.org/10.1007/s00382-018-4404-z). Such results highlight that no single forecasting system consistently outperforms all others. Consequently, combining forecasts from multiple models provides an opportunity to exploit the strengths of each system while mitigating their individual weaknesses. Within this framework, bias correction enhances the reliability of the forecasts produced by each model, whereas model combination synthesizes the corrected predictions to generate a more robust and skillful ensemble forecast. Together, these approaches may contribute to improving both forecast calibration and predictive performance."} {"chunk_id": "seasonal_seasonal-monthly-single-levels_validation_q05__bd9dccd37b5e", "report_id": "seasonal_seasonal-monthly-single-levels_validation_q05", "dataset_id": "seasonal-monthly-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q05", "aspect_base": "validation", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Benefit and challenges of a multi-model approach to seasonal forecasts > Analysis and results > 3. Technical aspects in using the MME approach > The Signal to Noise Paradox", "title": "Benefit and challenges of a multi-model approach to seasonal forecasts", "chunk_index": 16, "token_count": 497, "text_raw": "The systematic analysis of climate forecasts has revealed the existence of the so-called Signal-to-Noise Paradox (SNP). This paradox refers to the unexpected situation in which seasonal forecast models can predict the observed climate anomalies more accurately than they can predict their own internal variability.\n\nFor example, in winter NAO forecasts, the ensemble mean can anticipate whether the season will be dominated by a positive or negative NAO phase (i.e. stronger or weaker westerly winds over the North Atlantic) even though the individual ensemble members within the same model vary so much from each other that, taken separately, they appear to have very little predictive power. The mismatch between good prediction of reality despite weak internal consistency defines the Signal-to-Noise Paradox. [[29]](https://doi.org/10.1038/s41612-018-0038-4)\n\nThis is paradoxical because, in principle, if the model is realistic, the predictability within the model’s ensemble should be similar to that of the real world. The existance of a Signal-to-Noise Paradox suggests that the models are underestimating amplitude of their own predictable signal or overestimating internal noise, pointing to systematic deficiencies in the forecasting systems, which are still unclear. [[30]](https://doi.org/10.1175/BAMS-D-24-0019.1).\n\nNonetheless, it has been shown that, in a multi-model ensemble, the signal-to-noise ratio (SNR) can offer valuable insights into forecast reliability. In particular, years characterised by a high SNR tend to exhibit, on average, larger observed deviations from the mean than years with a low SNR, for both near-surface temperature (T2m) and precipitation. This suggests that forecast systems might be more reliable in predicting large anomalies (e.g., extremes) when there is greater coherence among ensemble members [[31]](https://doi.org/10.5194/wcd-4-823-2023).\n\n(multimodel:section-4)=", "text_with_prefix": "EQC Quality Assessment: \"Benefit and challenges of a multi-model approach to seasonal forecasts\"\nDataset: seasonal-monthly-single-levels [CDS]\nAspect: validation_q05 | Category: Seasonal_Forecasts\nSection: Benefit and challenges of a multi-model approach to seasonal forecasts > Analysis and results > 3. Technical aspects in using the MME approach > The Signal to Noise Paradox\n---\nThe systematic analysis of climate forecasts has revealed the existence of the so-called Signal-to-Noise Paradox (SNP). This paradox refers to the unexpected situation in which seasonal forecast models can predict the observed climate anomalies more accurately than they can predict their own internal variability.\n\nFor example, in winter NAO forecasts, the ensemble mean can anticipate whether the season will be dominated by a positive or negative NAO phase (i.e. stronger or weaker westerly winds over the North Atlantic) even though the individual ensemble members within the same model vary so much from each other that, taken separately, they appear to have very little predictive power. The mismatch between good prediction of reality despite weak internal consistency defines the Signal-to-Noise Paradox. [[29]](https://doi.org/10.1038/s41612-018-0038-4)\n\nThis is paradoxical because, in principle, if the model is realistic, the predictability within the model’s ensemble should be similar to that of the real world. The existance of a Signal-to-Noise Paradox suggests that the models are underestimating amplitude of their own predictable signal or overestimating internal noise, pointing to systematic deficiencies in the forecasting systems, which are still unclear. [[30]](https://doi.org/10.1175/BAMS-D-24-0019.1).\n\nNonetheless, it has been shown that, in a multi-model ensemble, the signal-to-noise ratio (SNR) can offer valuable insights into forecast reliability. In particular, years characterised by a high SNR tend to exhibit, on average, larger observed deviations from the mean than years with a low SNR, for both near-surface temperature (T2m) and precipitation. This suggests that forecast systems might be more reliable in predicting large anomalies (e.g., extremes) when there is greater coherence among ensemble members [[31]](https://doi.org/10.5194/wcd-4-823-2023).\n\n(multimodel:section-4)="} {"chunk_id": "seasonal_seasonal-monthly-single-levels_validation_q05__de1f4e3766ef", "report_id": "seasonal_seasonal-monthly-single-levels_validation_q05", "dataset_id": "seasonal-monthly-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q05", "aspect_base": "validation", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Benefit and challenges of a multi-model approach to seasonal forecasts > Analysis and results > 4. The MME approach at work for the C3S multi-model", "title": "Benefit and challenges of a multi-model approach to seasonal forecasts", "chunk_index": 17, "token_count": 516, "text_raw": "On a monthly basis, C3S issues [seasonal forecast charts](https://climate.copernicus.eu/seasonal-forecasts) based on the multi-model data available [here](https://cds.climate.copernicus.eu/datasets/seasonal-original-single-levels?tab=overview).\n\nIn this case, the seasonal outlook is not designed for specific applications. The multi-model ensemble (MME) forecast is calculated as a weighted average of the ensemble means from nine component models: ECMWF, Met Office, Météo-France, CMCC, DWD, NCEP, JMA, BOM and ECCC.\n\nThe weighting ensures that each component contributes equally to the overall variance of the multi-model during the common hindcast period.\nFor each component model, ensemble mean anomalies are calculated relative to its own model climate.\n\nWhen computing the multi-model mean, the weight assigned to each component is determined by dividing the square root of the average variance across all systems by the square root of the variance of the respective component.\n\nA slightly different approach is adopted for the [North American Multi-Model Ensemble (NMME)](https://www.cpc.ncep.noaa.gov/products/NMME/). In this case, to calculate anomalies the forecast bias is removed and is calculated separately for each model using all ensemble members for that particular model. The grand ensemble\nmean, and other diagnostics such as tercile probabilities, are defined as by assuming\nassuming that each ensemble member of each model is equally probable .\n\nattachment:ad6f1cd6-07a6-445f-8542-6287f198b380.jpg\n---\nheight: 600px\n---\nComparison of the July-August-Septmeber 2025 probabilities of above/below/neutral sea surface temperature anomalies forecasted in June 2025, according to the C3S MME (left) and to the NMME (right) methodologies. Note the differences in the respective probabilities as well as the similarities in terms of the overall patterns of global anomalies.\n```\n\n(multimodel:section-5)=", "text_with_prefix": "EQC Quality Assessment: \"Benefit and challenges of a multi-model approach to seasonal forecasts\"\nDataset: seasonal-monthly-single-levels [CDS]\nAspect: validation_q05 | Category: Seasonal_Forecasts\nSection: Benefit and challenges of a multi-model approach to seasonal forecasts > Analysis and results > 4. The MME approach at work for the C3S multi-model\n---\nOn a monthly basis, C3S issues [seasonal forecast charts](https://climate.copernicus.eu/seasonal-forecasts) based on the multi-model data available [here](https://cds.climate.copernicus.eu/datasets/seasonal-original-single-levels?tab=overview).\n\nIn this case, the seasonal outlook is not designed for specific applications. The multi-model ensemble (MME) forecast is calculated as a weighted average of the ensemble means from nine component models: ECMWF, Met Office, Météo-France, CMCC, DWD, NCEP, JMA, BOM and ECCC.\n\nThe weighting ensures that each component contributes equally to the overall variance of the multi-model during the common hindcast period.\nFor each component model, ensemble mean anomalies are calculated relative to its own model climate.\n\nWhen computing the multi-model mean, the weight assigned to each component is determined by dividing the square root of the average variance across all systems by the square root of the variance of the respective component.\n\nA slightly different approach is adopted for the [North American Multi-Model Ensemble (NMME)](https://www.cpc.ncep.noaa.gov/products/NMME/). In this case, to calculate anomalies the forecast bias is removed and is calculated separately for each model using all ensemble members for that particular model. The grand ensemble\nmean, and other diagnostics such as tercile probabilities, are defined as by assuming\nassuming that each ensemble member of each model is equally probable .\n\nattachment:ad6f1cd6-07a6-445f-8542-6287f198b380.jpg\n---\nheight: 600px\n---\nComparison of the July-August-Septmeber 2025 probabilities of above/below/neutral sea surface temperature anomalies forecasted in June 2025, according to the C3S MME (left) and to the NMME (right) methodologies. Note the differences in the respective probabilities as well as the similarities in terms of the overall patterns of global anomalies.\n```\n\n(multimodel:section-5)="} {"chunk_id": "seasonal_seasonal-monthly-single-levels_validation_q05__cf0344ac6fbb", "report_id": "seasonal_seasonal-monthly-single-levels_validation_q05", "dataset_id": "seasonal-monthly-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q05", "aspect_base": "validation", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Benefit and challenges of a multi-model approach to seasonal forecasts > Analysis and results > 5. Challenges and Limitations", "title": "Benefit and challenges of a multi-model approach to seasonal forecasts", "chunk_index": 18, "token_count": 212, "text_raw": "The use of a multi-model approach is case specific and there is no evidence of a single standard approach to be adopted as an all-purpose solution.\n\nFor example, bias adjustment is essential when preparing the (usually daily based) input for downstream impact models, such as crop models or energy production models. In such circumstances, the added value of a multi-model approach may be off-set by the computational demand of the processing chain. On the other hand, recalibration approaches improve the overall reliability of seasonal indicators by building on the temporal correspondence between the ensemble mean predictions and the corresponding observations [[21]](https://doi.org/10.1126/science.285.5433.1548).", "text_with_prefix": "EQC Quality Assessment: \"Benefit and challenges of a multi-model approach to seasonal forecasts\"\nDataset: seasonal-monthly-single-levels [CDS]\nAspect: validation_q05 | Category: Seasonal_Forecasts\nSection: Benefit and challenges of a multi-model approach to seasonal forecasts > Analysis and results > 5. Challenges and Limitations\n---\nThe use of a multi-model approach is case specific and there is no evidence of a single standard approach to be adopted as an all-purpose solution.\n\nFor example, bias adjustment is essential when preparing the (usually daily based) input for downstream impact models, such as crop models or energy production models. In such circumstances, the added value of a multi-model approach may be off-set by the computational demand of the processing chain. On the other hand, recalibration approaches improve the overall reliability of seasonal indicators by building on the temporal correspondence between the ensemble mean predictions and the corresponding observations [[21]](https://doi.org/10.1126/science.285.5433.1548)."} {"chunk_id": "seasonal_seasonal-monthly-single-levels_validation_q05__e943992cdfad", "report_id": "seasonal_seasonal-monthly-single-levels_validation_q05", "dataset_id": "seasonal-monthly-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q05", "aspect_base": "validation", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Benefit and challenges of a multi-model approach to seasonal forecasts > ℹ️ If you want to know more > Key resources", "title": "Benefit and challenges of a multi-model approach to seasonal forecasts", "chunk_index": 19, "token_count": 208, "text_raw": "Explore the [C3S Seasonal Forecast Products](https://climate.copernicus.eu/seasonal-forecasts)\n\nExplore the [North American Multi-Model Ensemble products](https://nmme.earth.miami.edu/forecasts/nmme_prob_forecast_all_var.html)\n\n[PyCPT](https://iri-pycpt.github.io/PyCPT2-Seasonal-Forecast-User-Guide/intro.html) - a tool calibrate and verify multi-model seasonal forecasts of precipitation based on the NOAA North American Multi-Model Ensemble (NMME) and European Copernicus Climate Change Service (C3S) databases.", "text_with_prefix": "EQC Quality Assessment: \"Benefit and challenges of a multi-model approach to seasonal forecasts\"\nDataset: seasonal-monthly-single-levels [CDS]\nAspect: validation_q05 | Category: Seasonal_Forecasts\nSection: Benefit and challenges of a multi-model approach to seasonal forecasts > ℹ️ If you want to know more > Key resources\n---\nExplore the [C3S Seasonal Forecast Products](https://climate.copernicus.eu/seasonal-forecasts)\n\nExplore the [North American Multi-Model Ensemble products](https://nmme.earth.miami.edu/forecasts/nmme_prob_forecast_all_var.html)\n\n[PyCPT](https://iri-pycpt.github.io/PyCPT2-Seasonal-Forecast-User-Guide/intro.html) - a tool calibrate and verify multi-model seasonal forecasts of precipitation based on the NOAA North American Multi-Model Ensemble (NMME) and European Copernicus Climate Change Service (C3S) databases."} {"chunk_id": "seasonal_seasonal-monthly-single-levels_validation_q05__b54ec2f94f02", "report_id": "seasonal_seasonal-monthly-single-levels_validation_q05", "dataset_id": "seasonal-monthly-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q05", "aspect_base": "validation", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Benefit and challenges of a multi-model approach to seasonal forecasts > ℹ️ If you want to know more > References", "title": "Benefit and challenges of a multi-model approach to seasonal forecasts", "chunk_index": 20, "token_count": 1015, "text_raw": "[[1]](https://doi.org/10.1029/2009GL040896) Hagedorn, R., Doblas-Reyes, F. J., & Palmer, T. N. (2006). DEMETER and the application of seasonal forecasts. Predictability of weather and climate, 674-692.\n\n[[2]](https://doi.org/10.1029/2007GL030740) Vitart, F., Huddleston, M. R., Déqué, M., Peake, D., Palmer, T. N., Stockdale, T. N., ... & Weisheimer, A. (2007). Dynamically‐based seasonal forecasts of Atlantic tropical storm activity issued in June by EUROSIP. Geophysical Research Letters, 34(16).\n\n[[3]](https://doi.org/10.1007/s00382-011-1061-x) Rajeevan, M., Unnikrishnan, C.K. & Preethi, B. Evaluation of the ENSEMBLES multi-model seasonal forecasts of Indian summer monsoon variability. Clim Dyn 38, 2257–2274 (2012).\n\n[[4]](https://doi.org/10.1002/qj.49712252905) Molteni, F., Buizza, R., Palmer, T. N., & Petroliagis, T. (1996). The ECMWF ensemble prediction system: Methodology and validation. Quarterly journal of the royal meteorological society, 122(529), 73-119.\n\n[[5]](https://doi.org/10.1002/qj.49712757202) Palmer, T. N. (2001). A nonlinear dynamical perspective on model error: A proposal for non‐local stochastic‐dynamic parametrization in weather and climate prediction models. Quarterly Journal of the Royal Meteorological Society, 127(572), 279-304.\n\n[[6]](https://doi.org/10.1038/nature04503) Thomson, M., Doblas-Reyes, F., Mason, S. et al. Malaria early warnings based on seasonal climate forecasts from multi-model ensembles. Nature 439, 576–579\n\n[[7]](https://doi.org/10.3402/tellusa.v57i3.14668) Morse, A. P., Doblas-Reyes, F. J., Hoshen, M. B., Hagedorn, R., & Palmer, T. N. (2005). A forecast quality assessment of an end-to-end probabilistic multi-model seasonal forecast system using a malaria model. Tellus A: Dynamic Meteorology and Oceanography, 57(3), 464–475. https://doi.org/10.3402/tellusa.v57i3.14668\n\n[[8]](https://doi.org/10.1016/j.cliser.2018.06.003) Iizumi, T., Shin, Y., Kim, W., Kim, M., & Choi, J. (2018). Global crop yield forecasting using seasonal climate information from a multi-model ensemble. Climate Services, 11, 13-23.\n\n[[9]](https://doi.org/10.5194/egusphere-2023-569) Thébault, C., Perrin, C., Andréassian, V., Thirel, G., Legrand, S., & Delaigue, O. (2023). Multi-model approach in a variable spatial framework for streamflow simulation. EGUsphere, 2023, 1-34.\n\n[[10]](https://doi.org/10.1088/1748-9326/ad627c) , Maximilian Zachow, Harald Kunstmann, Daniel Julio Miralles and Senthold Asseng (2024) Multi-model ensembles for regional and national wheat yield forecasts in Argentina Environ. Res. Lett. 19 084037\n\n[[11]](https://doi.org/10.1175/JHM-D-18-0040.1)Wanders, N., S. Thober, R. Kumar, M. Pan, J. Sheffield, L. Samaniego, and E. F. Wood, 2019: Development and Evaluation of a Pan-European Multimodel Seasonal Hydrological Forecasting System. J. Hydrometeor., 20, 99–115, https://doi.org/10.1175/JHM-D-18-0040.1 .", "text_with_prefix": "EQC Quality Assessment: \"Benefit and challenges of a multi-model approach to seasonal forecasts\"\nDataset: seasonal-monthly-single-levels [CDS]\nAspect: validation_q05 | Category: Seasonal_Forecasts\nSection: Benefit and challenges of a multi-model approach to seasonal forecasts > ℹ️ If you want to know more > References\n---\n[[1]](https://doi.org/10.1029/2009GL040896) Hagedorn, R., Doblas-Reyes, F. J., & Palmer, T. N. (2006). DEMETER and the application of seasonal forecasts. Predictability of weather and climate, 674-692.\n\n[[2]](https://doi.org/10.1029/2007GL030740) Vitart, F., Huddleston, M. R., Déqué, M., Peake, D., Palmer, T. N., Stockdale, T. N., ... & Weisheimer, A. (2007). Dynamically‐based seasonal forecasts of Atlantic tropical storm activity issued in June by EUROSIP. Geophysical Research Letters, 34(16).\n\n[[3]](https://doi.org/10.1007/s00382-011-1061-x) Rajeevan, M., Unnikrishnan, C.K. & Preethi, B. Evaluation of the ENSEMBLES multi-model seasonal forecasts of Indian summer monsoon variability. Clim Dyn 38, 2257–2274 (2012).\n\n[[4]](https://doi.org/10.1002/qj.49712252905) Molteni, F., Buizza, R., Palmer, T. N., & Petroliagis, T. (1996). The ECMWF ensemble prediction system: Methodology and validation. Quarterly journal of the royal meteorological society, 122(529), 73-119.\n\n[[5]](https://doi.org/10.1002/qj.49712757202) Palmer, T. N. (2001). A nonlinear dynamical perspective on model error: A proposal for non‐local stochastic‐dynamic parametrization in weather and climate prediction models. Quarterly Journal of the Royal Meteorological Society, 127(572), 279-304.\n\n[[6]](https://doi.org/10.1038/nature04503) Thomson, M., Doblas-Reyes, F., Mason, S. et al. Malaria early warnings based on seasonal climate forecasts from multi-model ensembles. Nature 439, 576–579\n\n[[7]](https://doi.org/10.3402/tellusa.v57i3.14668) Morse, A. P., Doblas-Reyes, F. J., Hoshen, M. B., Hagedorn, R., & Palmer, T. N. (2005). A forecast quality assessment of an end-to-end probabilistic multi-model seasonal forecast system using a malaria model. Tellus A: Dynamic Meteorology and Oceanography, 57(3), 464–475. https://doi.org/10.3402/tellusa.v57i3.14668\n\n[[8]](https://doi.org/10.1016/j.cliser.2018.06.003) Iizumi, T., Shin, Y., Kim, W., Kim, M., & Choi, J. (2018). Global crop yield forecasting using seasonal climate information from a multi-model ensemble. Climate Services, 11, 13-23.\n\n[[9]](https://doi.org/10.5194/egusphere-2023-569) Thébault, C., Perrin, C., Andréassian, V., Thirel, G., Legrand, S., & Delaigue, O. (2023). Multi-model approach in a variable spatial framework for streamflow simulation. EGUsphere, 2023, 1-34.\n\n[[10]](https://doi.org/10.1088/1748-9326/ad627c) , Maximilian Zachow, Harald Kunstmann, Daniel Julio Miralles and Senthold Asseng (2024) Multi-model ensembles for regional and national wheat yield forecasts in Argentina Environ. Res. Lett. 19 084037\n\n[[11]](https://doi.org/10.1175/JHM-D-18-0040.1)Wanders, N., S. Thober, R. Kumar, M. Pan, J. Sheffield, L. Samaniego, and E. F. Wood, 2019: Development and Evaluation of a Pan-European Multimodel Seasonal Hydrological Forecasting System. J. Hydrometeor., 20, 99–115, https://doi.org/10.1175/JHM-D-18-0040.1 ."} {"chunk_id": "seasonal_seasonal-monthly-single-levels_validation_q05__9e2cb2307d37", "report_id": "seasonal_seasonal-monthly-single-levels_validation_q05", "dataset_id": "seasonal-monthly-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q05", "aspect_base": "validation", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Benefit and challenges of a multi-model approach to seasonal forecasts > ℹ️ If you want to know more > References", "title": "Benefit and challenges of a multi-model approach to seasonal forecasts", "chunk_index": 21, "token_count": 1123, "text_raw": "-European Multimodel Seasonal Hydrological Forecasting System. J. Hydrometeor., 20, 99–115, https://doi.org/10.1175/JHM-D-18-0040.1 .\n\n[[12]](https://doi.org/10.1007/s00382-019-04654-y) Lee, D. Y., Doblas-Reyes, F. J., Torralba, V., & Gonzalez-Reviriego, N. (2019). Multi-model seasonal forecasts for the wind energy sector. Climate Dynamics, 53, 2715-2729.\n\n[[13]](https://doi.org/10.1007/s00382-017-3766-y) Alessandri, A., Felice, M.D., Catalano, F. et al. Grand European and Asian-Pacific multi-model seasonal forecasts: maximization of skill and of potential economical value to end-users . Clim Dyn 50, 2719–2738 (2018).\n\n[[14]](https://doi.org/10.1002/2014WR015426) Mendoza, P. A., Rajagopalan, B., Clark, M. P., Cortés, G., & McPhee, J. (2014). A robust multimodel framework for ensemble seasonal hydroclimatic forecasts. Water Resources Research, 50(7), 6030-6052.\n\n[[15]](https://doi.org/10.5194/nhess-21-2215-2021) Keating, C., Lee, D., Bazo, J., and Block, P.: Leveraging multi-model season-ahead streamflow forecasts to trigger advanced flood preparedness in Peru, Nat. Hazards Earth Syst. Sci., 21, 2215–2231,\n\n[[16]](https://doi.org/10.1016/j.cliser.2021.100272) N. Acharya, M.A. Ehsan, A. Admasu, A. Teshome, K.J.C. Hall (2021) On the next generation (NextGen) seasonal prediction system to enhance climate services over Ethiopia Clim. Serv., 24 (2021), 10.1016/j.cliser.2021.100272\n\n[[17]](https://doi.org/10.1175/2008MWR2431.1) Tippett, M. K., & Barnston, A. G. (2008). Skill of multimodel ENSO probability forecasts. Monthly Weather Review, 136(10), 3933-3946.\n\n[[18]](https://doi.org/10.1256/qj.05.65) Vitart, F. (2006), Seasonal forecasting of tropical storm frequency using a multi-model ensemble. Q.J.R. Meteorol. Soc., 132: 647-666. https://doi.org/10.1256/qj.05.65\n\n[[19]](https://doi.org/10.1016/j.wace.2015.06.005) Acacia S. Pepler, Leandro B. Díaz, Chloé Prodhomme, Francisco J. Doblas-Reyes, Arun Kumar, The ability of a multi-model seasonal forecasting ensemble to forecast the frequency of warm, cold and wet extremes (2015) , Weather and Climate Extremes, https://doi.org/10.1016/j.wace.2015.06.005 .\n\n[[20]](https://doi.org/10.1007/s00382-019-04640-4) Manzanas, R., Gutiérrez, J.M., Bhend, J. et al. Bias adjustment and ensemble recalibration methods for seasonal forecasting: a comprehensive intercomparison using the C3S dataset. Clim Dyn 53, 1287–1305 (2019). https://doi.org/10.1007/s00382-019-04640-4\n\n[[21]](https://doi.org/10.1126/science.285.5433.1548) Krishnamurti, T. N., Kishtawal, C. M., LaRow, T. E., Bachiochi, D. R., Zhang, Z., Williford, C. E., ... & Surendran, S. (1999). Improved weather and seasonal climate forecasts from multimodel superensemble. Science, 285(5433), 1548-1550.\n\n[[22]](https://doi.org/10.3402/tellusa.v57i3.14657) Hagedorn, R., Doblas-Reyes, F. J., & Palmer, T. N. (2005). The rationale behind the success of multi-model ensembles in seasonal forecasting — I. Basic concept. Tellus A: Dynamic Meteorology and Oceanography, 57(3), 219–233. https://doi.org/10.3402/tellusa.v57i3.14657", "text_with_prefix": "EQC Quality Assessment: \"Benefit and challenges of a multi-model approach to seasonal forecasts\"\nDataset: seasonal-monthly-single-levels [CDS]\nAspect: validation_q05 | Category: Seasonal_Forecasts\nSection: Benefit and challenges of a multi-model approach to seasonal forecasts > ℹ️ If you want to know more > References\n---\n-European Multimodel Seasonal Hydrological Forecasting System. J. Hydrometeor., 20, 99–115, https://doi.org/10.1175/JHM-D-18-0040.1 .\n\n[[12]](https://doi.org/10.1007/s00382-019-04654-y) Lee, D. Y., Doblas-Reyes, F. J., Torralba, V., & Gonzalez-Reviriego, N. (2019). Multi-model seasonal forecasts for the wind energy sector. Climate Dynamics, 53, 2715-2729.\n\n[[13]](https://doi.org/10.1007/s00382-017-3766-y) Alessandri, A., Felice, M.D., Catalano, F. et al. Grand European and Asian-Pacific multi-model seasonal forecasts: maximization of skill and of potential economical value to end-users . Clim Dyn 50, 2719–2738 (2018).\n\n[[14]](https://doi.org/10.1002/2014WR015426) Mendoza, P. A., Rajagopalan, B., Clark, M. P., Cortés, G., & McPhee, J. (2014). A robust multimodel framework for ensemble seasonal hydroclimatic forecasts. Water Resources Research, 50(7), 6030-6052.\n\n[[15]](https://doi.org/10.5194/nhess-21-2215-2021) Keating, C., Lee, D., Bazo, J., and Block, P.: Leveraging multi-model season-ahead streamflow forecasts to trigger advanced flood preparedness in Peru, Nat. Hazards Earth Syst. Sci., 21, 2215–2231,\n\n[[16]](https://doi.org/10.1016/j.cliser.2021.100272) N. Acharya, M.A. Ehsan, A. Admasu, A. Teshome, K.J.C. Hall (2021) On the next generation (NextGen) seasonal prediction system to enhance climate services over Ethiopia Clim. Serv., 24 (2021), 10.1016/j.cliser.2021.100272\n\n[[17]](https://doi.org/10.1175/2008MWR2431.1) Tippett, M. K., & Barnston, A. G. (2008). Skill of multimodel ENSO probability forecasts. Monthly Weather Review, 136(10), 3933-3946.\n\n[[18]](https://doi.org/10.1256/qj.05.65) Vitart, F. (2006), Seasonal forecasting of tropical storm frequency using a multi-model ensemble. Q.J.R. Meteorol. Soc., 132: 647-666. https://doi.org/10.1256/qj.05.65\n\n[[19]](https://doi.org/10.1016/j.wace.2015.06.005) Acacia S. Pepler, Leandro B. Díaz, Chloé Prodhomme, Francisco J. Doblas-Reyes, Arun Kumar, The ability of a multi-model seasonal forecasting ensemble to forecast the frequency of warm, cold and wet extremes (2015) , Weather and Climate Extremes, https://doi.org/10.1016/j.wace.2015.06.005 .\n\n[[20]](https://doi.org/10.1007/s00382-019-04640-4) Manzanas, R., Gutiérrez, J.M., Bhend, J. et al. Bias adjustment and ensemble recalibration methods for seasonal forecasting: a comprehensive intercomparison using the C3S dataset. Clim Dyn 53, 1287–1305 (2019). https://doi.org/10.1007/s00382-019-04640-4\n\n[[21]](https://doi.org/10.1126/science.285.5433.1548) Krishnamurti, T. N., Kishtawal, C. M., LaRow, T. E., Bachiochi, D. R., Zhang, Z., Williford, C. E., ... & Surendran, S. (1999). Improved weather and seasonal climate forecasts from multimodel superensemble. Science, 285(5433), 1548-1550.\n\n[[22]](https://doi.org/10.3402/tellusa.v57i3.14657) Hagedorn, R., Doblas-Reyes, F. J., & Palmer, T. N. (2005). The rationale behind the success of multi-model ensembles in seasonal forecasting — I. Basic concept. Tellus A: Dynamic Meteorology and Oceanography, 57(3), 219–233. https://doi.org/10.3402/tellusa.v57i3.14657"} {"chunk_id": "seasonal_seasonal-monthly-single-levels_validation_q05__a090fb889e84", "report_id": "seasonal_seasonal-monthly-single-levels_validation_q05", "dataset_id": "seasonal-monthly-single-levels", "store": "CDS", "doc_type": "EQC_QA", "aspect": "validation_q05", "aspect_base": "validation", "category": "Seasonal_Forecasts", "match_confidence": "exact", "section": "Benefit and challenges of a multi-model approach to seasonal forecasts > ℹ️ If you want to know more > References", "title": "Benefit and challenges of a multi-model approach to seasonal forecasts", "chunk_index": 22, "token_count": 890, "text_raw": "sembles in seasonal forecasting — I. Basic concept. Tellus A: Dynamic Meteorology and Oceanography, 57(3), 219–233. https://doi.org/10.3402/tellusa.v57i3.14657\n\n[[23]](https://doi.org/10.1002/qj.210) Weigel, A. P., Liniger, M. A., & Appenzeller, C. (2008). Can multi‐model combination really enhance the prediction skill of probabilistic ensemble forecasts?. Quarterly Journal of the Royal Meteorological Society: A journal of the atmospheric sciences, applied meteorology and physical oceanography, 134(630), 241-260.\n\n[[24]](https://doi.org/10.1007/s00382-020-05314-2) Hemri, S., Bhend, J., Liniger, M.A. et al. How to create an operational multi-model of seasonal forecasts?. Clim Dyn 55, 1141–1157 (2020). https://doi.org/10.1007/s00382-020-05314-2\n\n[[25]](https://doi.org/10.3402/tellusa.v57i3.14658) Doblas-Reyes, F. J., Hagedorn, R., & Palmer, T. N. (2005). The rationale behind the success of multi-model ensembles in seasonal forecasting – II. Calibration and combination. Tellus A: Dynamic Meteorology and Oceanography, 57(3), 234–252.\n\n[[26]](https://doi.org/10.1007/s00382-018-4404-z) Mishra N, Prodhomme C, Guemas V (2018) Multi-model skill assessment of seasonal temperature and precipitation forecasts over Europe. Clim Dyn.\n\n[[27]](https://doi.org/10.1175/MWR2818.1) Robertson, A. W., Lall, U., Zebiak, S. E., & Goddard, L. (2004). Improved combination of multiple atmospheric GCM ensembles for seasonal prediction. Monthly Weather Review, 132(12), 2732-2744.\n\n[[28]](https://doi.org/10.1175/BAMS-84-12-1783) Barnston, A. G., Mason, S. J., Goddard, L., DeWitt, D. G., & Zebiak, S. E. (2003). Multimodel ensembling in seasonal climate forecasting at IRI. Bulletin of the American Meteorological Society, 84(12), 1783-1796.\n\n[[29]](https://doi.org/10.1038/s41612-018-0038-4) Scaife, A. A., & Smith, D. (2018). A signal-to-noise paradox in climate science. npj Climate and Atmospheric Science, 1(1), 28.\n\n[[30]](https://doi.org/10.1175/BAMS-D-24-0019.1) Weisheimer, A., Baker, L. H., Bröcker, J., Garfinkel, C. I., Hardiman, S. C., Hodson, D. L., ... & Sutton, R. T. (2024). The signal-to-noise paradox in climate forecasts: revisiting our understanding and identifying future priorities. Bulletin of the American Meteorological Society, 105(3), E651-E659.\n\n[[31]](https://doi.org/10.5194/wcd-4-823-2023) Acosta Navarro, J. C., & Toreti, A. (2023). Exploiting the signal-to-noise ratio in multi-system predictions of boreal summer precipitation and temperature. Weather and Climate Dynamics, 4(3), 823-831.", "text_with_prefix": "EQC Quality Assessment: \"Benefit and challenges of a multi-model approach to seasonal forecasts\"\nDataset: seasonal-monthly-single-levels [CDS]\nAspect: validation_q05 | Category: Seasonal_Forecasts\nSection: Benefit and challenges of a multi-model approach to seasonal forecasts > ℹ️ If you want to know more > References\n---\nsembles in seasonal forecasting — I. Basic concept. Tellus A: Dynamic Meteorology and Oceanography, 57(3), 219–233. https://doi.org/10.3402/tellusa.v57i3.14657\n\n[[23]](https://doi.org/10.1002/qj.210) Weigel, A. P., Liniger, M. A., & Appenzeller, C. (2008). Can multi‐model combination really enhance the prediction skill of probabilistic ensemble forecasts?. Quarterly Journal of the Royal Meteorological Society: A journal of the atmospheric sciences, applied meteorology and physical oceanography, 134(630), 241-260.\n\n[[24]](https://doi.org/10.1007/s00382-020-05314-2) Hemri, S., Bhend, J., Liniger, M.A. et al. How to create an operational multi-model of seasonal forecasts?. Clim Dyn 55, 1141–1157 (2020). https://doi.org/10.1007/s00382-020-05314-2\n\n[[25]](https://doi.org/10.3402/tellusa.v57i3.14658) Doblas-Reyes, F. J., Hagedorn, R., & Palmer, T. N. (2005). The rationale behind the success of multi-model ensembles in seasonal forecasting – II. Calibration and combination. Tellus A: Dynamic Meteorology and Oceanography, 57(3), 234–252.\n\n[[26]](https://doi.org/10.1007/s00382-018-4404-z) Mishra N, Prodhomme C, Guemas V (2018) Multi-model skill assessment of seasonal temperature and precipitation forecasts over Europe. Clim Dyn.\n\n[[27]](https://doi.org/10.1175/MWR2818.1) Robertson, A. W., Lall, U., Zebiak, S. E., & Goddard, L. (2004). Improved combination of multiple atmospheric GCM ensembles for seasonal prediction. Monthly Weather Review, 132(12), 2732-2744.\n\n[[28]](https://doi.org/10.1175/BAMS-84-12-1783) Barnston, A. G., Mason, S. J., Goddard, L., DeWitt, D. G., & Zebiak, S. E. (2003). Multimodel ensembling in seasonal climate forecasting at IRI. Bulletin of the American Meteorological Society, 84(12), 1783-1796.\n\n[[29]](https://doi.org/10.1038/s41612-018-0038-4) Scaife, A. A., & Smith, D. (2018). A signal-to-noise paradox in climate science. npj Climate and Atmospheric Science, 1(1), 28.\n\n[[30]](https://doi.org/10.1175/BAMS-D-24-0019.1) Weisheimer, A., Baker, L. H., Bröcker, J., Garfinkel, C. I., Hardiman, S. C., Hodson, D. L., ... & Sutton, R. T. (2024). The signal-to-noise paradox in climate forecasts: revisiting our understanding and identifying future priorities. Bulletin of the American Meteorological Society, 105(3), E651-E659.\n\n[[31]](https://doi.org/10.5194/wcd-4-823-2023) Acosta Navarro, J. C., & Toreti, A. (2023). Exploiting the signal-to-noise ratio in multi-system predictions of boreal summer precipitation and temperature. Weather and Climate Dynamics, 4(3), 823-831."}