Satellite Rainfall Products and Their Reliability in the Blue Nile Basin

Abstract

In the Upper Blue Nile (UBN) basin, there is very sparse and uneven distribution of ground-based meteorological stations which constrain assessments on rainfall distributions and representation. To assess the diurnal cycle of rainfall across the UBN basin, satellite observations from Tropical Rainfall Measuring Mission (TRMM) were used in this study. Data of 7 years (2002–2008) of Precipitation Radar (PR) and TRMM Microwave Imager (TMI) were processed, with analyses based on geographic information system (GIS) operations, statistical techniques, and harmonic analysis. Diurnal cycle patterns of rainfall occurrence and rain rate from three in-situ weather stations are well represented by the satellite observations. Harmonic analysis depicts large differences in the mean of the diurnal cycle, amplitude, and time of the amplitude across the study area. Diurnal cycle of rainfall occurrence has a single peak in Lake Tana, Gilgel Abbay, and Jemma subbasins and double peaks in Belles, Dabus, and Muger subbasins. Maximum rain rate occurs in the morning (Gilgel Abbay, Dabus, and Jemma), afternoon (Belles, Beshilo, and Muger), and evening (Lake Tana and along the river gorges). Results of this study indicate that satellite observations provide an alternative source of data to characterize diurnal cycle of rainfall in data-scarce regions. We noticed, however, that there are a number of constraints to the use of satellite observations. For more accurate assessments, satellite products require validation by a network of well-distributed ground stations. Also, we advocate bias correction.