Statistical and hydrological evaluation of remotely sensed rainfall products in the Upper Blue Nile basin, Ethiopia

Satellite-based rainfall products (SRPs) have a wide range of applications, but their accuracy and reliability need to be assessed. This study evaluated the performance of three SRPs: Climate Hazards Group InfraRed Precipitation with Stations (CHIRPS), the Integrated Multi-satellitE Retrieval for Global Precipitation Measurement (IMERG), and Precipitation Estimation from Remotely Sensed Information using Artificial Neural Networks-Climate Data Record (PERSIANN-CDR) over the Upper Blue Nile Basin (UBNB) on daily and monthly time scales by employing three distinct approaches. First, the evaluation was carried out by comparing the SRPs with rainfall measurements obtained from 25 rainfall gauging stations using statistical indicators. Following this, the Triple Collocation (TC) was applied. The study continued to investigate the hydrological utility of the SRPs at three selected watersheds in the basin using the GR4J (Génie Rural à 4 paramètres Journalier) hydrological model. Results indicated a statistically significant strong monthly correlation between SRPs and gauge observations, but a weak daily correlation. IMERG showed higher performance on a daily scale, while CHIRPS outperformed on a monthly scale based on statistical metrics. TC-based results also revealed the superior performance of IMERG on the daily scale, with SRPs’ performance declining from west to east and exhibiting higher performance at lower elevations. GR4J modeling results indicated SRPs’ potential for hydrological modeling applications, but challenges in simulating the high flow conditions were noted. Overall, the study underscores the critical role of SRPs in enhancing hydrological modeling, streamflow simulations, and water resource management for policy decision-making, especially in data-scarce regions. However, further research is needed to enhance the reliability and applicability of SRPs for more accurate generalization.