This dataset contains the pretrained model weights and precomputed results for the paper "Thunderstorm nowcasting with deep learning: a multi-hazard data fusion model" submitted to Geophysical Research Letters. A preprint of the paper will be submitted. The ML code can be found at https://github.com/MeteoSwiss/c4dl-multi. Download all the files here and extract the contents to the following subdirectories in the ML code directory: Results (c4dl-results-lightningdl.zip) -> results/ Pretrained models (c4dl-models-lightningdl.zip) -> models/ Additionally, you will need the datasets from this Zenodo archive. Follow the instructions there for downloading.

The work of JL was supported by the fellowship "Seamless Artificially Intelligent Thunderstorm Nowcasts" from the European Organisation for the Exploitation of Meteorological Satellites (EUMETSAT). The hosting institution of this fellowship is MeteoSwiss in Switzerland.