| # Model Card for chimp_smhi_v3 | |
| Retrieves surface precipitation and radar reflectivity from SEVIRI observations. | |
| ## Model Details | |
| ### Model Description | |
| The chimp_smhi_v3 model is a 2+1D convolutional network that retrieves ground-based radar reflectivity | |
| and surface precipitation estimates from 15-minute MeteoSat SEVIRI observations. It is trained used OPERA | |
| and BALTRAD composite radar reflectivities as well as GPM CMB surface precipitation estimates. | |
| {{ model_description | default("", true) }} | |
| - **Developed by:** Simon Pfreundschuh | |
| - **Model type:** CNN | |
| - **License:** MIT | |
| ### Model Sources [optional] | |
| <!-- Provide the basic links for the model. --> | |
| - **Repository:** https://github.com/simonpf/chimp | |
| ## Uses | |
| THe model is still experimental and should only be used for validation purposes at the current development stage. | |
| ## How to Get Started with the Model | |
| See instruction at [https://github.com/simonpf/chimp_smhi][https://github.com/simonpf/chimp_smhi] | |
| ## Training Details | |
| ### Training Data | |
| Training was performed on two years (2019, 2020) of matched SEVIRI and OPERA/BALTRAD data. | |
| ## Model Card Authors [optional] | |
| @simonpf | |
| ## Model Card Contact | |
| @simonpf |