This dataset contains the machine learning training data files, pretrained model weights and precomputed results for the paper "Seamless lightning nowcasting with recurrent-convolutional deep learning" published in: Leinonen, J., Hamann, U., & Germann, U. (2022). Seamless Lightning Nowcasting with Recurrent-Convolutional Deep Learning, Artificial Intelligence for the Earth Systems, 1(4), e220043, doi:10.1175/AIES-D-22-0043.1. A preprint of the paper can be found at https://arxiv.org/abs/2203.10114. The ML code can be found at https://github.com/MeteoSwiss/c4dl-lightningdl. Download all the files here and extract the contents to the following subdirectories in the ML code directory: Training data (c4dl-patches-*.zip) -> data/2020/ Results (c4dl-results-lightningdl.zip) -> results/ Pretrained models (c4dl-models-lightningdl.zip) -> models/ Additionally, the file c4dl-randomexamples-lightningdl.zip contains the randomly selected examples complementing Figs. 7–9 of the paper, and the file c4dl-inputsamples-lightningdl.zip contains figures showing samples of all the input variables for the three cases shown in Figs. 7–9.

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.