juliensimon/aurora-forecast

Aurora Forecast (OVATION Nowcast)

Credit: NASA

Part of a dataset collection on Hugging Face.

Dataset description

A compact daily time series of global auroral activity derived from NOAA SWPC's OVATION Prime aurora nowcast. Updated daily, growing incrementally.

OVATION Prime is an empirical model that converts real-time solar-wind and geomagnetic conditions into the probability of seeing the aurora across a global grid of roughly 65,000 cells, with a lead time of about 30-40 minutes. It is the model behind NOAA's public "aurora forecast" maps. The raw product is a dense spatial grid; this dataset reduces each snapshot to per-hemisphere summary metrics so that long-term trends in auroral activity can be tracked in a single tidy table.

For each daily snapshot the pipeline records, separately for the Northern and Southern hemispheres, the total integrated aurora probability (an index of overall activity), the peak cell probability, and the number of "active" cells above a 10% threshold (a proxy for the area of the auroral oval). These aggregates rise and fall with geomagnetic storms and closely track indices such as Kp and Dst, making the dataset useful for studying aurora visibility, comparing hemispheres, and connecting solar-wind drivers to auroral response. Because the source is a short-lead nowcast rather than an observation, values reflect modeled probability, not confirmed sightings.

This dataset is suitable for time-series forecasting tasks.

Schema

Column	Type	Description	Sample	Null %
observation_time	datetime64[us, UTC]	UTC timestamp of the solar-wind/satellite observations that seed this OVATION nowcast.	2026-06-23 09:52:00+00:00	0.0%
forecast_time	datetime64[us, UTC]	UTC valid time of the nowcast, roughly 30-40 minutes after the observation time (the lead time for aurora to respond).	2026-06-23 11:08:00+00:00	0.0%
north_total_power	float64	Sum of aurora probability (%) over all Northern-Hemisphere grid cells. A unitless index proportional to total Northern auroral activity; rises sharply during geomagnetic storms.	41950.0	0.0%
north_max_intensity	float64	Peak aurora probability (%) in any Northern-Hemisphere cell (0-100). 100 means visible aurora is essentially certain somewhere in the north.	15.0	0.0%
north_active_cells	Int64	Number of Northern-Hemisphere grid cells (0 to ~32,000, up to half the ~65k-cell global grid) with aurora probability >= 10%, a proxy for the spatial extent (area) of the northern auroral oval.	959	0.0%
south_total_power	float64	Sum of aurora probability (%) over all Southern-Hemisphere grid cells. Unitless index proportional to total Southern (aurora australis) activity.	51826.0	0.0%
south_max_intensity	float64	Peak aurora probability (%) in any Southern-Hemisphere cell (0-100).	24.0	0.0%
south_active_cells	Int64	Number of Southern-Hemisphere grid cells (0 to ~32,000, up to half the ~65k-cell global grid) with aurora probability >= 10%, a proxy for the area of the southern auroral oval.	1079	0.0%

Quick stats

• 7 daily auroral snapshots (2026-06-23 to 2026-06-28)
• Peak Northern activity index: 91,542
• Peak Southern activity index: 124,334

Usage

from datasets import load_dataset

ds = load_dataset("juliensimon/aurora-forecast", split="train")
df = ds.to_pandas()

from datasets import load_dataset
import matplotlib.pyplot as plt

ds = load_dataset("juliensimon/aurora-forecast", split="train")
df = ds.to_pandas().sort_values("observation_time")

# Northern vs Southern auroral activity over time
fig, ax = plt.subplots(figsize=(12, 4))
ax.plot(df["observation_time"], df["north_total_power"], label="North")
ax.plot(df["observation_time"], df["south_total_power"], label="South")
ax.set_ylabel("Integrated aurora probability (activity index)")
ax.set_title("Global Auroral Activity (OVATION nowcast)")
ax.legend()
plt.tight_layout()
plt.show()

Data source

https://www.swpc.noaa.gov/products/aurora-30-minute-forecast

Update schedule

Daily at 16:25 UTC

Related datasets

• juliensimon/geomagnetic-kp-index

• juliensimon/dst-index

• juliensimon/substorm-onsets

• juliensimon/donki-space-weather-events

• juliensimon/solar-wind

• juliensimon/parker-solar-probe

• juliensimon/solar-orbiter

• juliensimon/cdaw-cme

• juliensimon/feng-icmes

If you find this dataset useful, please consider giving it a like on Hugging Face. It helps others discover it.

About the author

Created by Julien Simon — AI Operating Partner at Fortino Capital. Part of the Space Datasets collection.

Citation

@dataset{aurora_forecast,
  title = {Aurora Forecast (OVATION Nowcast)},
  author = {juliensimon},
  year = {2026},
  url = {https://huggingface.co/datasets/juliensimon/aurora-forecast},
  publisher = {Hugging Face}
}

License

CC-BY-4.0