Berhan, Gessesse (PhD)Wondimu, Haimanote2022-03-152023-11-282022-03-152023-11-282022-02http://etd.aau.edu.et/handle/12345678/30582The production of global and continental satellite rainfall products at fine spatial and temporal scales for users in reliable and cost-effective methods to overcome the shortcoming of ground-measured rainfall has resulted from advances in remote sensing technology. However, the accuracy of satellite rainfall estimations is influenced by onboard sensors, retrieval algorithms, topography features, and seasonal fluctuations. Thus, the purpose of this study is to evaluate the accuracy of the most recently released satellite rainfall products (ARC-V2, CHIRPS-V2, PERSIANN-CCS, RFE-V2, TAMSAT-V3.1, TRMM-3B42-V7, and IMERG-F-V6) against ground-measured data at dekadal, monthly, and yearly time scales in the Omo-Gibe river basin, from 2003 to 2015. In addition, to this, we evaluated the effect of seasonal and elevation variations on the accuracy of satellite rainfall estimates. A point to pixel evaluation was carried out using continuous (r, MAE, ME, PBias, and Eff), categorical (POD, FAR, and CSI), and volumetric (VHI, VFAR, and VCSI) statistical indices, and ArcGIS and statistical software’s were used to analyze the space-time variability of rainfall. The results showed that IMERG-F-V6 and CHIRPS-V2 have the best performance in detecting and estimating rainfall followed by TRMM-3B42-V7 at all-time scales. These products showed low MAE, slightly overestimates, high correlation (r > 0.71, r ≥ 0.8, and r > 0.75), and good efficiency (Eff > 0.4, Eff ≥ 0.6, and Eff ≥ 0.33) with rain gauge measurements at dekadal, monthly, and annual time scales, respectively. In contrast, ARC-V2 and PERSIANN-CCS products were relatively the worst performance in most validations criteria at all-time scales. Overall, the accuracy of all satellite rainfall estimates is significantly improved at monthly time scale. The accuracy of IMERG-F-V6 and CHIRPS-V2 were more stable than others during the wet season, and the accuracy of all satellite estimates was better in highlands than lowlands for the wet season. Furthermore, the assessments for various elevation categories revealed that IMERG-F-V6 performed better in all statistical indices and was less affected by elevation fluctuations, followed by CHIRPS-V2 and TRMM-3B42-V7. TAMASAT-V3.1 and ARC-V2, on the other hand, had the worst performance in detecting rainfall at various elevations, although they didn't exhibit a clear association with elevation changes. Similarly, in virtually all continuous statistical outcomes, PERSIANN-CCS had the worst performance, and its accuracy was clearly reliant on elevation. As a consequence of this research, the best products are IMERG-F-V6 and CHIRPS-V2, which show good results in predicting and detecting rainfall in the Omo-Gibe river basin. Furthermore, these products are utilized to improve socioeconomic activities, decision-making, climate change, and water-related applications, particularly in data-scarce places and other locations with comparable features, such as the Omo-Gibe river basin.en-USAccuracy evaluationSatellite-based rainfall estimatesRain-gauge measurementsStatistical indicesRugged topographyElevationWet seasonThesis