Leveraging reforecasts for flood estimation with long continuous simulation: a proof-of-concept study

Abstract. Flood estimation is critical for risk assessment, but traditional methods are often constrained by the limited length of observation data. This study explores the potential of reforecasts (RFs) to enhance flood estimation through use in long continuous simulation (CS) with a hydrological model. As a proof of concept, we processed individual RFs from the vast database of the European Center for Medium-Range Weather Forecasts (ECMWF) with bias correction, stochastic downscaling and disaggregation with analogs to finally obtain mean areal precipitation and mean areal temperature for a set of test catchments in Switzerland. We subsequently concatenated these RFs into a time series of close to 10 000 years length and used them in long CS to derive flood return levels. Results demonstrate the potential of RFs as a complementary tool in flood estimation, providing insights into extreme event magnitudes and frequencies. Moreover, RFs can provide a relevant alternative view on exceptionally high extremes when compared to flood estimates derived from using other inputs to long CS, such as those generated by a stochastic weather generator. Limitations apply to catchments smaller than approximately 500 km², where the stochastic downscaling becomes increasingly inadequate, especially for resolving convective events. There, dynamical downscaling would be more appropriate, but was not feasible with the data currently available.