Medium range hydrological forecasts in mesoscale catchments are only possible with the use or hydrological models driven by meteorological forecasts, which in particular contribute quantitative precipitation forecasts (QPF). QPFs are accompanied by large uncertainties, especially for longer lead times, which are propagated wihin the hydrometeorological model system. To deal with this limitation of predictability, a probabilistic forecasting system is tested, which is based on a hydrological-meteorological ensemble prediction syste. The meteorological component of the system is the operational limited-area ensemble to a horizontal resoltuion of 10km, while the hydrological component is based on the semi-distributed hydrological model PREVAH with a spatial resoltuion of 500m. Earlier studies have mostly addressed the potential benefits of hydrometeorological ensemble systems in short case studies. Here we present an analysis of hydrological ensemble hindcasts for two years (2005 and 2006). It is shown that the ensemble covers the uncertainty during different weather situations with an appropriate spread-skill relationship. The ensemble also shows advantages over a corresponding deterministic forecast, even under consideration of an artificial spread.
Jaun, S., & Ahrens, B. (2009). Evaluation of a probabilistic hydrometeorological forecast system. Hydrology and Earth System Sciences, 13(7), 1031–1043. https://doi.org/10.5194/hess-13-1031-2009