Regional climate model simulations as input for hydrological applications: evaluation of uncertainties

S. Kotlarski (1) , A. Block (2) , U. Böhm (3) , D. Jacob (1) , K. Keuler (2) , R. Knoche (4) , D. Rechid (1) , A. Walter (5)

1 MPI-M - Max-Planck-Institut für Meteorologie
2 Brandenburgische Technische Universität Cottbus-Senftenberg
3 PIK - Potsdam-Institut für Klimafolgenforschung
4 Forschungszentrum Karlsruhe
5 DWD - Deutscher Wetterdienst [Offenbach]

The ERA15 Reanalysis (1979-1993) has been dynamically downscaled over Central Europe using 4 different regional climate models. The regional simulations were analysed with respect to 2m temperature and total precipitation, the main input parameters for hydrological applications. Model results were validated against three reference data sets (ERA15, CRU, DWD) and uncertainty ranges were derived. For mean annual 2 m temperature over Germany, the simulation bias lies between -1.1°C and +0.9°C depending on the combination of model and reference data set. The bias of mean annual precipitation varies between -31 and +108 mm/year. Differences between RCM results are of the same magnitude as differences between the reference data sets.

Domaines: Sciences de la Terre

S. Kotlarski, A. Block, U. Böhm, D. Jacob, K. Keuler, et al.. Regional climate model simulations as input for hydrological applications: evaluation of uncertainties. Advances in Geosciences, 2005, 5, pp.119-125. ⟨hal-00296850⟩