Influence of boundaries and exchange at interfaces

Regional Climate Modeling

At present the land surface (e.g. orography, lakes, snow cover or land-use) is only poorly represented in the state of the art climate models. The poor representation leads to biases of surface temperature and the hydrological cycle. This strongly influences the large atmospheric circulation systems as the Monsoon, the ITCZ and the mid-latitude cyclone belts. At the moment the development of a hierarchy of climate models is supported. Global atmospheric data sets (observations or simulations) drive the Weather Research and Forecast Model (WRF), which represents the land surface in a more accurate way and generates a regional atmospheric data set at a high spatial (e.g. 3km) and temporal (e.g. 1 hour) scale. This atmospheric data set is then used as input for statistical downscaling approaches and local scale geo-ecosystem models.