Remote sensing based estimation of evaporation among different land cover types in the Upper-Wami River Basin, Tanzania

Abstract

In Semi-arid Africa, evaporation is the dominant hydrological flux and hence an important component in water resources management. However, it is a challenge to fairly estimate the spatial variation of evaporation especially on composite terrains due to heterogeneity of the landscape. This is further compounded by the uncertainties associated with the various approaches used for estimating the flux, most of which are based on extrapolating point observations. In the recent past, there has been progress in developing and validating various remote sensing algorithms for computing evaporation over large spatial extents. One such algorithm, the Surface Energy Balance System (SEBS) was applied in the Mkindo catchment, located in the upper-Wami River basin in Tanzania using the Moderate Resolution Imaging Spectroradiometer (MODIS) satellite images. Six images covering June 2008 to September 2009 were analyzed together with a land use map of the same area. The SEBS evaporation estimates were compared with the potential evaporation over the Mkindo catchment. SEBS results showed good correlation with the computed Potential evaporation computed from climatological parameter in the catchment, with different land uses/cover types having different evaporative water use signature, on both daily and monthly time scale. Forest and irrigated agriculture land use, located on the lower parts of the catchment, had relatively higher evaporative water use compared to the other land uses in the catchment.