CAMELS-DEBY: a lumped and gridded dataset for hourly and daily rainfall-runoff modeling in 210 catchments of Bavaria, Germany

CAMELS-DEBY is a hydrometeorological dataset for 210 catchments of the Federal state of Bavaria in Germany. Variables such as precipitation, temperature, potential evapotranspiration, river flow, river flow depth, groundwater table elevations, landuse/landcover and soil properties are included. The time series have hourly and daily temporal resolutions. The hourly time series cover the time period of 2007 till 2023 (17 years) while the daily are from 1970 till 2023 (54 years). For all catchments the data are prepared in a lumped and gridded (1km) format covering a total surface area of 46,800 sq.km. In this region, a special situation was taken advantage of in case of the the meteorological variables i.e., precipitation and temperature. Networks maintained by two independent authorities, the German Weather Service (DWD; Deutscher Wetter Dienst) and the Bavarian Hydrological Service (GKD; Gewässerkundlicher Dienst), were merged to produce a much denser network of observations in the recent two decades. Along with serving the purpose of distributed modeling, this dataset also offers the benefit of modeling quasi-input-uncertainty testing by not only having data derived from observation gauges alone but also having other existing gridded products such as HYRAS, RADOLAN, RADKLIM and E-OBS. River discharge was simulated using a fully-lumped and a semi-distributed HBV with various input products that showed that the performance was, practically, independent of the type of input in almost all cases. All data is prepared in a consistent manner by resampling them to the same 1km grid (or point) and time period(s) so that it is trivial to change model inputs without having to change any other setting. Please read the accompanying dcamels_deby_readme.txt file for proper description. A Paper is submitted to the Copernicus ESSD journal (same place as CAMELS-DE).