Search

Abstract: Scientifically sound decisions in sustainable water management are usually based on hydrological modeling which can only be accomplished by meteorological driving information. Especially in regions with weak infrastructure this task is hampered by limited hydro-meteorological information in sufficient spatial and temporal resolution. We investigated three approaches to provide required meteorological fields driving the distributed hydrological model: the results of the mesoscale meteorological model MM5 which are available near real time, the TRMM product 3B42 available with approximately one month delay, and station data available with a delay of one year or more. The study site is the White Volta catchment in the semi-arid environment of West Africa. The results for 2004 show that the meteorological model is able to provide meteorological input data for near real time water balance estimations. In this study the TRMM product does not improve the simulation results. Besides missing important meteorological data, also gridded information on land surface properties (albedo, LAI, etc.) is usually difficult to obtain, albeit it is an essential input for distributed hydrological models. This information is commonly taken from “static” tables depending on the land use. Satellite remote sensing provides worldwide spatially detailed information on land surface properties which is particularly suitable for large regions in remote settings. Therefore the MODIS products for albedo and LAI were processed to annual time series including the identification and replacement of low quality observations by interpolation. The impact using MODIS data on the spatial distribution of water balance variables occurs mainly on local scale. The hydrological simulations using MODIS LAI and albedo values result in higher annual evapotranspiration and lower total discharge sums for 2004. Altogether it is concluded that hydrological decision support systems in regions with weak infrastructure can benefit significantly from the integration of atmospheric modeling and satellite-derived land surface data. [Copyright &y& Elsevier]