Keywords
Ensemble prediction systems| Data Assimilation | Multimodeling | Postprocessing | HydrOlOgical Prediction LAboratory (HOOPLA)
Hydrological ensemble prediction systems (H-EPS) provide information on the most probable forecast and its uncertainty range. We explore the three main sources of uncertainty of H-EPS in order to quantify and reduce them: structural uncertainty pooling multiple hydrological models, initial conditions uncertainty implementing a probabilistic data assimilation scheme, and meteorological uncertainty resorting to meteorological ensemble forecasts. Our team coded a comprehensive MATLAB toolbox for exploring H-EPSs: the HydrOlOgical Prediction LAboratory (HOOPLA), available on Github for download.