The GROWA model was applied to the entire Federal State of North Rhine-Westphalia (ca. 34,000 km(2)) using a grid resolution of 100 m. It conceptually combines distributed meteorological data (winter and summer precipitation and potential evapotranspiration) with distributed site parameters (land use, soil properties, slope gradient, slope exposure, mean depth to groundwater) to facilitate the calculation of long-term annual averages of total runoff. In the GROWA model groundwater recharge is expressed as a constant proportion (baseflow indices) of the total runoff. This portion depends on certain characteristics of the investigated area, e.g. the slope gradient, soil and hydrogeological properties as well as the degree of surface sealing. In this paper special emphasis is put on the influence of geology on groundwater recharge. In this respect, a new calibration procedure for the parameterization of geology-related parameters is described. In previous applications of the GROWA model baseflow indices have been identified on the basis of observed mean monthly low-water runoff values (MoMLR). Since the MoMLR-values significantly overestimate groundwater recharge in solid rock regions due to high interflow proportions, a more sophisticated hydrograph separation method has been applied. In this study runoff data from about 125 gauging stations within the Federal State of North Rhine-Westphalia were used to derive baseflow indices. The raster-based simulation was carried out using a grid resolution of 100 m. The accuracy of the calculated groundwater recharge values for the period 1979-1999 was verified on the basis of data from gauging stations. A good agreement between observed runoff values from the sub-catchments and model results was achieved. (c) 2005 Elsevier B.V. All rights reserved.
