Future Precipitation Extremes and Urban Flood Risk Under +2°C and +3°C Warming: A Novel Non-Stationary Climate-Hydrodynamic Modeling Chain for Using High-Resolution Radar Data and a Convection-Permitting Climate Model Ensemble

Urban planners and engineers rely on historical climate data to design flood protection infrastructure capable of withstanding extreme flooding events, typically associated with a 1% annual exceedance probability (the 100-year flood). This study examines how hourly precipitation extremes are expected to evolve with rising temperatures and how these changes will influence urban flooding risks. Specifically, we address the often-overlooked impact of short-duration rainfall extremes using a new non-stationary temperature-conditional extreme precipitation scaling method and a novel regional climate convection-permitting model ensemble for +2°C and +3°C global warming scenarios for the whole of Germany. We compare this newly generated non-stationary extreme precipitation dataset with an established dataset, and then assess the implications of the future precipitation changes on flood risks in two pre-alpine communes in Germany using hydrodynamic modeling. Our results reveal that ignoring climate change can lead to significant underestimations of flood risk. Under the +3°C scenario, flood risks increase dramatically, with a 60% rise in the number of buildings affected by high flood levels (water levels of 1 meter or more). These findings suggest that current or recently implemented flood protection infrastructure may be insufficient to address the future challenges posed by climate change, underscoring the need for adaptive planning to mitigate escalating flood risks.