Evaluating the performance of CCD Satellite Rainfall estimates over the Nile Basin

Abstract

Rainfall is probably the most important input to hydrologic models and the most uncertain of them as well. The high uncertainty in rainfall estimate in general, and over the Nile basin in particular is due to the lack of adequate surface‐based rainfall monitoring networks. Using satellite imagery provides a solution to develop rainfall estimates over the Nile basin in light of the shortage of rain gauge data. In this paper, the rainfall estimates from a 9yr (1998‐2006) continuous stream of realtime satellite data from the European METEOSAT satellite positioned over Africa have been evaluated through comparison with the gauge‐based CRU rainfall data. The satellite estimate is based on the Cold Cloud Duration (CCD) technique utilizing the infra‐red band imagery (i.e. a single channel) only. The common error measures Root Mean Squared Error (RMSE) and Mean Error (Bias) were used in this study as they give together a good picture about the performance. The evaluation is done on a monthly time step at a spatial scale of (50km x 50km).