Datensatz

Zusammenfassung

Many sectors, such as hydropower, agriculture, water supply and waterway transport, require information about the possible evolution of meteorological and hydrological conditions within the next weeks and months to optimize their decision-making processes on a long term. Since July 2022, the German Federal Institute of Hydrology (BfG) is providing operational 6-week forecasts for selected gauges at river Rhine and Elbe on the monthly timescale. Due to ongoing research, this forecast for the upper Danube is published in a pre-operational state. Taking the increasing uncertainties with longer lead times into account, user communication and a proper postprocessing is the key to a useful forecast. Hindcast analysis shows, that the forecast has skill at least for the first three weeks during the whole year, even without advanced postprocessing or data assimilation. Recently in addition to an autoregressive and wavelet-based output correction, a postprocessing method called EMOS (ensemble model output statistics) (Gneiting et al., 2005 and Hemri & Klein, 2017) has been implemented to further improve the forecast skill. The results show significant differences between a non-postprocessed and a postprocessed hydrological 6-week forecast. There are also differences in forecast skill depending on the postprocessing method and between the analyzed rivers Rhein, Elbe and Danube, due to their catchment attributes. References Gneiting, Raftery, Westveld, Goldman (2005): Calibrated probabilistic forecasting using ensemble model output statistics and minimum CRPS estimation, Monthly Weather Review 133(5), 1098-1118 Hemri, Klein (2017): Analog-Based Postprocessing of Navigation-Related Hydrological Ensemble Forecasts, Water Resources Research 53(11), 9059-9077