Physically based Modelling of Climate Change Impact on Snow Cover Dynamics in Alpine Regions using a Stochastic Weather Generator Mauser, W. 1 , Prasch, M. 1 and Strasser, U. 1 1 Department of Geography, Ludwig-Maximilians University, Munich, Germany Email: w.mauser@iggf.geo.uni-muenchen.de Keywords: Climate Change, snow modelling, alpine regions, PROMET, Brahmatwinn, GLOWA Danube EXTENDED ABSTRACT Alpine mountains are of global importance in providing downstream freshwater, much of it stored as snow several months of the year. This function of a temporal storage of precipitation and its delayed release during melt is one of the key characteristics of a mountain snow cover. In a changing climate, the latter process is supposed to be modified in its temporal dynamics. This important objective is investigated in the frameworks of the projects Brahmatwinn (http://www.brahmatwinn.uni-jena.de ) as well as GLOWA Danube (www.glowa-danube.de ) by determining the impact of future climate change on the hydrology of the Upper Danube catchment with its alpine headwaters. For a distributed simulation of the snow cover evolution for both present as well as for future scenarios we apply the model PROMET (Processes of Radiation, Mass and Energy Transfer), the land surface core of the model framework DANUBIA. In its current version it is a distributed, physical, non-calibrated regional Earth System model, which simulates all water fluxes (rainfall, soil water movement, evapotranspiration, direct runoff, interflow, streamflow) on a 1 km grid base. It also includes a dynamic vegetation model. PROMET has interfaces to the outputs of regional climate models as well as to the groundwater model MODFLOW. It also includes a snow module for simulating the energy balance, the water equivalent and the melt rate of a snow cover. To validate the modelled snow water equivalent at both the local as well as at the regional scale, two representative weather stations with distinct characteristics in the Upper Danube catchment are chosen, and model results are compared with measurements. Furthermore, we compare distributed simulation results of snow coverage with NOAA AVHRR satellite data derived snow cover. In a next step, a stochastic, nearest neighbour weather generator to produce climate change scenario data is introduced. It generates a consistent future climate data set for the period 2005-2104 by appropriate stochastic rearrange- ment of historically measured meteorological data. This data set is then used as input to PROMET to determine the impact of future climate change according to the IPCC-B2 scenario with a temperature increase of 2.7K per 100 years. The impacts of the simulated future climate are discussed focussing on changed precipitation. Then, the change in snow water equivalent compared to the past 30 years from 1971 to 2000 is analysed for the future periods 2031 to 2060 and 2071 to 2100. For this purpose the mean evolution of the snow cover over the entire period, as well as the annual courses are considered. The duration of the snow cover for the winter season of the respective periods is analysed as well. Finally, the impact of climate change on the snow cover dynamics and its consequences for runoff generation is discussed. The model is currently adapted to the Upper Brahmaputra basin. In the future, we intend to apply it with properly downscaled output of global circulation models (GCMs).