Climate change impact assessments form the basis for the development of suitable climate change adaptation strategies. For this purpose, ensembles consisting of stepwise coupled models are generally used [emission scenario → global circulation model → downscaling approach (DA) → bias correction → impact model (hydrological model)], in which every item is affected by considerable uncertainty. The aim of the current study is (1) to analyse the uncertainty related to the choice of the DA as well as the hydrological model and its parameterization and (2) to evaluate the vulnerability of the studied catchment, a subcatchment of the highly anthropogenically impacted Spree River catchment, to hydrological change. Four different DAs are used to drive four different model configurations of two conceptually different hydrological models (Water Balance Simulation Model developed at ETH Zürich and HBV-light). In total, 452 simulations are carried out. The results show that all simulations compute an increase in air temperature and potential evapotranspiration. For precipitation, runoff and actual evapotranspiration, opposing trends are computed depending on the DA used to drive the hydrological models. Overall, the largest source of uncertainty can be attributed to the choice of the DA, especially regarding whether it is statistical or dynamical. The choice of the hydrological model and its parameterization is of less importance when long-term mean annual changes are compared. The large bandwidth at the end of the modelling chain may exacerbate the formulation of suitable climate change adaption strategies on the regional scale. Figure 6. Frequency plot for daily (top, precipitation > 10 mm/day not displayed) and monthly (bottom) precipitation for the reference period (CCLM: COSMO model in climate mode; REMO: regional model; WettReg: weather-type regionalization method)
3990A. GÄDEKE ET AL. Figure 8. Comparison between the interannual course of measured and simulated (reference 1963-1992 and scenario period 2031-2060) temperatures for the Weißer Schöps River catchment (interpolation by the inverse distance method) (CCLM: COSMO model in climate mode; REMO: regional model; STAR: Statistical Regional model; WettReg: weather-type regionalization method) 3992 A. GÄDEKE ET AL.