Daily vs. hourly simulation for estimating future flood peaks in mesoscale catchments

Daily hydrological models are commonly used to study changes in flood peaks due to climate change. Although they often lead to an underestimation of absolute floods, it is assumed that future flood peaks in smaller mesoscale catchments are less underestimated when examining the relative change signal of floods. In this study, the applicability of this hypothesis is investigated by comparing the results of a daily hydrological model set, calibrated on runoff hydrographs, with an hourly model set calibrated on flood peak distributions. For analysis, a daily RCP8.5 climate model ensemble is disaggregated to hourly values and the runoff is simulated on a daily and hourly basis for six mesoscale catchments in Central Germany. Absolute floods and relative flood changes are compared between both model sets. The results show significant differences between the absolute floods of both model sets, in most cases caused by underestimations due to the daily modeling process. In contrast, the differences between the two model sets are not significant for the relative change signal of the floods, especially for higher return periods. To improve results in climate studies with coarse modeling time step, the use of relative change signal of floods instead of absolute values is recommended.