Discharge projections for the major rivers of Germany based on bias-corrected climate projections and the water balance model LARSIM-ME.

The data are based on water balance simulations conducted within the framework of the BMDV Expert Network (BMDV-ExpN) research program for the period 1971-2099, provided by the DAS-Basisdienst "Klima und Wasser". Results from a multi-model ensemble of climate simulations (mainly CMIP5, EURO-CORDEX, ReKliEs-De) under emission scenarios RCP 2.6, RCP 4.5, and RCP 8.5 were converted into daily discharge values at various gauging stations using the regional water balance model LARSIM-ME. Hydrological model:LARSIM-ME (LSM-ME) is a spatially distributed, process-based model that covers German river basins including upstream foreign catchments. It has a horizontal resolution of 5 km and a daily temporal resolution [exceptions apply, e.g., in tidal areas]. Evapotranspiration is calculated using the Penman-Monteith method (ATV-DVWK-M, 2002), requiring bias-corrected input variables such as air temperature, global radiation, wind speed, relative humidity, and air pressure. The climate projections were bias-corrected by the German Meteorological Service (DWD) using the HYRAS observational dataset (DWD, 2025) for the period 1971-2000, as part of the BMDV Expert Network (BC-EXP). Further information on literature, data sources, processing steps, and additional results and interpretation approaches can be found in Nilson et al. (2020). Additional gauging stations are available upon request. Data origin:The data result from an extensive processing chain consisting of emission scenarios (RCPs), various climate models (GCM-RCM combinations), hydrological models (WHM), statistical processing steps (e.g., bias correction), and evaluation steps:(1) assessment and selection of GCM-RCM combinations based on model bias,(2) spatiotemporal transformation of GCM-RCM data to WHM model domains,(3) bias correction of selected GCM-RCM combinations,(4) hydrological modeling,(5) temporal aggregation and index calculation (Hänsel et al., 2020).