Abstract: Modelling landscapes with GIS provides a powerful tool to analyze and model terrestrial ecosystems for land use planning. In order to gain realistic results, input data has to represent the natural environment as realistically as possible. While some of the fundamental data layers such as topography and soils are more or less static, climatic data is variable both in space and time. In addition, climatic data is measured at discrete locations and depends to a considerable extent on the topography around the sampling site. Most interpolation methods that are used to generate gridded climate data surfaces, however, fail to represent the influence of topography on climate altogether or use overly simplistic models. This paper describes a new method called REGEOTOP that includes the influence of topography on climate in an interpolation procedure. First, relief forms are extracted from a Digital Elevation Model (DEM) with the help of principal component analysis. In general, more than 90% of the DEM variance can be explained by 13~18 principal components. Position (latitude, longitude and altitude) and relief forms around meteorological stations are then related to observed climate values such as monthly precipitation or evaporation observed at these stations by multivariate regression. Inserting position and relief forms for each grid cell of the DEM into the regression equation allow to calculate a predicted climate surface. In the last step, residuals between observed and predicted values of the meteorological stations are interpolated by geostatistical interpolation (“kriging”) and added to the predicted climate surfaces to account for the remaining variance. Examples for high resolution data sets describing monthly temperature, precipitation and evapotranspiration (1951~1990) and their application to modelling landscape properties in Central China are given. These data surfaces can be used as precision input data to analyze and model landscape properties for planning purposes with GIS. Calculating basic ecological properties of wetlands such as soil moisture and plant available water supplies allows to assess the suitability for reconversion of farmland to wetland areas.

Key words: model, climate date, land use dynamics, hydrological plan