Fully differentiable, fully distributed River Discharge Prediction: data sets

This repository contains the data sets used in: Scholz et al. (2025). Fully differentiable, fully distributed River Discharge Prediction. dem_1000.h5 based on EU-DEM v1.1, reprojected to RADOLAN grid: https://sdi.eea.europa.eu/catalogue/srv/api/records/3473589f-0854-4601-919e-2e7dd172ff50 efas.h5 based on EFAS historical: https://ewds.climate.copernicus.eu/datasets/efas-historical?tab=overview era5_ssrd_neckar*.nc based on ERA5 provided by ECMWF, reprojected to RADOLAN grid: https://www.ecmwf.int/en/forecasts/dataset/ecmwf-reanalysis-v5 radolan_neckar_*.h5 based on RADOLAN rw product provided by the Deutsche Wetterdienst: https://opendata.dwd.de/climate_environment/CDC/grids_germany/hourly/radolan/ Due to copyright, the discharge data has to be downloaded manually from the Global Runoff Data Centre (https://grdc.bafg.de/), and then preprocessed with the provided bafg_parser.py python script. We use the following stations in our work: 6335290: STEIN 6335291: GAILDORF 6335565: BAD IMNAU 6335600: ROCKENAU SKA 6335601: LAUFFEN 6335602: PLOCHINGEN 6335603: ROTTWEIL 6335604: KIRCHENTELLINSFURT 6335620: MOSBACH 6335660: PFORZHEIM 6335665: DENKENDORF 6335671: ALTENSTEIG 6335675: MURR 6335676: OPPENWEILER 6335680: SCHWABSBERG 6335681: UNTERGRIESHEIM 6335690: NEUSTADT To preprocess the discharge data, additionally the river network data "Fließgewässer (AWGN)" provided by the Landesanstalt für Umwelt Baden-Württemberg (LUBW) is required: https://rips-metadaten.lubw.de/trefferanzeige?docuuid=7251515f-6aed-4555-8319-ab6314155ab1