Climate simulation models as they are used to estimate global climate changes, have limitations with respect to regional features. The reason for this is the fact that the regional climate and its possible changes are strongly influenced by the regional topography (mountains, coasts) that cannot be sufficiently resolved by global climate models. In order to describe regional effects, it is necessary to nest a highly resolved regional climate model in the global one. However, this requires a rather high computational effort.
The studies at IPA aim at the development and improvement of regionalisation (downscaling) techniques for results of global climate simulations. In particular, the statistical-dynamical downscaling (SDD) is subject to further improvements. It is based on the results of regional climate models in combination with weather-type statistics. It allows to limit the computationally expensive regional climate simulations to a smaller number of multi-day episodes which are representative of the full regional climate. With the help of SDD it is also possible to extrapolate the results of nested regional climate models to a longer period provided the results of a global climate simulation extends over this period. Moreover, it is possible to use the statistical-dynamical method as an analysis tool for the interpretation of regional climate changes. In particular it allows to estimate the role of a changed importance of specific circulation patterns (large-scale weather types).
The studies at IPA are embedded in two projects, namely QUIRCS (funded by the German Ministry of Education and Research within the framework of DEKLIM) and the European project PRUDENCE (funded by the European Comission).