A comparison of rainfall estimation techniques for sub-Saharan Africa

Interpolated rain-gauge data were compared to Meteosat-based precipitation estimates for sub-Saharan Africa. Validation was carried out using a dataset from a very dense gauge network in South Africa, on a point-to-pixel (PO–PI) as well as on a pixel-to-pixel (PI–PI) basis. Error criteria computed at the gauged pixels indicate that overall the interpolated estimates perform similarly to the satellite-based data: they provide good estimates of lower but underestimate larger precipitation amounts. It is concluded that the satellite estimates are more fitted for the operational modelling of processes such as surface runoff and soil erosion, especially in the developing areas where resources are scarce.