The objective of this work is to assess the skill of this specific operational hydrological forecasting system, which to this extend (to the knowledge of the author) has not been reported in literature before. For this purpose 2 years of pre-operational EFAS medium-range deterministic and probabilistic flood forecasts with up to 10 days in advance were analysed statistically for the whole of Europe.
Because of (distributed) nature of EFAS this data intensive skill assessment cannot make use of the traditional hydrological approach as it cannot make use of observed discharges, in fact it is using a proxy (hydrologic simulation with observed meteo data).
Most suitable skill scores and visualizations are reviewed and pros and cons are analysed. This skill assessment is looking at the past performance but is also designed to improve the future performance of such a system by making it possible to incorporate the past experience into the current forecast. By making the past performance easily accessible at each pixel the forecaster can better evaluate the probability of a current forecast. This is especially useful as the study showed that the meteorological assumed equi-probability of forecast members does (in the case of EFAS) not linearly translate into the hydrological probabilistic forecasts and that there are biases.