Performance evaluation of multi-satellite rainfall products in the Gidabo catchment, Rift Valley Basin, Ethiopia

Abstract

Abstract Satellite rainfalls are good options to overcome shorter records, record challenges and inconsistencies with rain-gauges. However, satellites' rainfall retrieval algorithms are region-time scale specific; hence, the key concern is selection of appropriate satellite products. Accordingly, this study evaluates the performance of five high-resolution satellites' rainfall using multiple-metrics at daily and monthly scales. The result showed that Climate Prediction Center (CPC) Morphing Algorithm (CMORPH.CPC) performed better by scoring: qualitatively; Critical Success Index (CSI = 0.856), Probability of Detection (POD = 0.911), Frequency Bias Index (FBI = 0.974), and quantitatively; correlation coefficient (CC = 0.375), Root Mean Square Error (RMSE ≈ 575), and Volumetric Critical Success Index (VCSI = 0.958) at a daily scale. At a monthly scale, Climate Hazards Group Infrared Precipitation with Stations (CHIRPS.v2) performed better by scoring CSI = 0.983, POD = 1 and FBI = 0.975 qualitatively, and quantitatively, CC = 0.836 with strong VCSI = 0.981 and better RMSE (≈125) than daily. These satellites' daily rainfall needs value-improving techniques before using in place of Gidabo's rain-gauge rainfall, while at monthly scale CHIRPS.v2's rainfall can be an alternative source of rainfall data. Finally, it ensured that for Gidabo catchment, the performance of satellite rainfall was more effective at monthly than daily scale.