Figure 1: Workflow of the modeling and validation procedure.

Observed daily stream flow data was used as the response variable in a weighted linear regression (LMWW). Seasonal accumulative precipitation data was used as the predictor variable. The model was validated spatially using Wilcoxon tests on the RMSE. The IHA metrics were calculated from observed and simulated streamflow and validated through Spearman’s correlation. The model was validated temporally with KGE, RMSE and R², and GLMs were performed on 34 species and compared with IHA metrics calculated from observed and simulated flow data, respectively. The daily stream flow data were extrapolated to the entire stream network of Germany resulting in a time series covering 64 years. The validated IHA indices were then calculated for all grid cells on the stream network. Top right map is Germany’s location in Europe, and underneath, the study area of Germany with distribution of gauging stations. Bottom right map shows the modeled daily stream flow (m³s⁻¹) of 11^th February 1950.