Published March 22, 2023 | Version 1.0.0
Dataset Open

Data archive for Exploiting radar polarimetry for nowcasting thunderstorm hazards using deep learning

  • 1. MeteoSwiss

Description

This dataset contains the machine learning training data files, pretrained model weights and results for the paper Exploiting radar polarimetry for nowcasting thunderstorm hazards using deep learning, submitted to Natural Hazards and Earth System Sciences, 2023.

The radar dataset can be found at the following Zenodo repository: https://doi.org/10.5281/zenodo.6325370

For instructions for using the data, please see the GitHub code repository at https://github.com/meteoswiss/c4dl-polar. Download all the files here and extract the contents to the following subdirectories in the ML code directory:

  • Training data (patches_quality-index_2020.zip or patches_*_2020.nc) -> data/2020/
  • Results: (results.zip) -> runs/run*/results/
  • Pretrained models (models_run*) -> runs/run*/

Files

models_run1.zip

Files (16.1 GB)

Name Size Download all
md5:7023577dd65c903a1f345f24a2b05ac7
439.3 MB Preview Download
md5:5af70bbd034df257f97e2db9b5c73099
439.2 MB Preview Download
md5:e9f3ed79933f36d4c500be4e92172f8d
439.3 MB Preview Download
md5:8bf478f7aaab3166eaf89a96574e3651
3.9 GB Download
md5:78c8147e92ff3d2864b2b1a4d26f8d45
2.7 MB Preview Download
md5:62528f3fec58ad5a9a05b44ee97ce764
3.4 GB Download
md5:c85c2f2747bb2533fc6226a2e2506d15
3.9 GB Download
md5:a38f686a87ae0b4ad0faefb4e2734b10
3.7 GB Download
md5:bb742e5c689131e93a116e40b8a416d6
579.0 kB Preview Download