Evaluation of Satellite Precipitation Estimates Over Omo-Gibe River Basin in Ethiopia

In this study, the accuracy of CMORPH, CMORPH-CRT, PERSIANN and PERSIANN-CDR satellite precipitation estimates (SPEs) were evaluated over the Omo-Gibe River Basin in Ethiopia across 10-year period from 2007 to 2016. The evaluation was performed at various temporal (daily, monthly, seasonal and annual) and spatial scales (point to pixel and basin scale aerial averaged precipitation) against data from 39 irregularly distributed ground rain gauge stations, using continuous and categorical statistical indices. The evaluation results show that at point to pixel scale PERSIANN-CDR estimate showed the highest performance on daily, monthly, Bega season and annual time scales whereas CMORPH-CRT outperformed on Kiremt and Belg seasonal time scales. At a basin scale, CMORPH-CRT showed the highest performance on daily, monthly, Kiremt season, Belg season, and annual time scales, whereas PERSIANN-CDR outperformed all estimates on Bega seasonal time scale. CMORPH exhibited the weakest performance on all time scales and at both spatial scales (point to pixel and basin-scale). Based on the categorical analysis PERSIANN was well able to characterize the local precipitation than others (PSS = 0.71 and POD = 0.90). The finding of this study indicates that CMORPH-CRT and PERSIANN-CDR estimates presented consistently good performance with observed precipitation at different spatiotemporal scales in the Omo-Gibe River Basin.