Abstract ID: 3.11481 | Not reviewed | Requested as: Talk | TBA | TBA

High-resolution convection-permitting regional climate models (CPRCMs) have demonstrated improved representation of extreme precipitation compared to coarser regional climate models (RCMs). However, their added value for simulating hydrological extremes, such as floods, remains uncertain, particularly in complex mountainous terrain. This study assesses the performance of a 3-km convection-permitting model (HCLIM3) against a coarser 12-km model (HCLIM12) from the HARMONIE-Climate (HCLIM) system in reproducing precipitation, temperature, and flood characteristics in two hydrologically contrasting basins in Western Norway: the coastal Røykenes basin, dominated by rainfall-induced floods, and the mountainous Bulken basin, where floods are primarily driven by snowmelt. To evaluate the impact of CPRCMs on hydrological extremes, we employ both a physically-based distributed model (WRF-Hydro) and a conceptual lumped model (HBV) for flood simulations. HCLIM3 better captures the spatial distribution of extreme precipitation, particularly for annual maximum 1-day and 1-hour events, compared to HCLIM12. However, both models exhibit a cold bias, which is more pronounced at lower elevations, especially in HCLIM12. Despite improvements in precipitation representation, HCLIM3-driven flood simulations do not consistently outperform those driven by HCLIM12, except for extreme flood peaks. The choice of hydrological model significantly impacts flood simulations. The HBV model underestimates flood peaks and frequencies, whereas WRF-Hydro provides more accurate simulations in Røykenes but overestimates floods in Bulken, likely due to forcing biases, particularly when driven by HCLIM3. These results suggest that while CPRCMs improve the representation of extreme precipitation and temperature, their direct benefit for flood simulations is less evident without adequate bias correction, especially in snowmelt-driven basins. This study highlights the importance of evaluating high-resolution climate models in an integrated atmosphere-hydrology framework to improve the prediction of hydrological extremes over complex mountainous regions.