EVALUATION OF MULTI-SATELLITE RAINFALL ESTIMATE FOR DROUGHT MONITORING IN GENALE-DAWA RIVER BASIN, ETHIOPIA

Abstract

Droughts pose significant challenges to Ethiopia, impacting agriculture, food security, hydroelectricity production, and public safety. However, in regions like the Genale Dawa River Basin (GDRB), where ground observations are limited, accurate drought monitoring remains challenging. To address this, this study evaluates the performance of multi-satellite rainfall products for drought monitoring in the GDRB. The major data used for the study were Satellite rainfall products, including (CHIRPSv2.0, RFE 2.0, TAMSATv3.1, PERSIANN-CDR, and ARC2) compared them with historical rainfall data (2001–2020) gathered from 12 weather stations, hydrological, Earth Observation Satellite(eMODIS) and sea surface temperature data. Bias correction was applied using Quantile Mapping with Empirical Distribution (QMED) and utilized statistical metrics such as MAE, RMSE, R, R2, POD, FAR, and CSI were used to evaluate their performance in identifying actual events. The evaluation used a pixel-to-point station locations scale at a daily, dekadal and monthly timescale. Furthermore, drought monitoring indices, including the Standardized Precipitation Index (SPI), Stream Drought Index (SDI), and Vegetation Condition Index (VCI), are utilized to characterize drought conditions in the basin and employed the Modified Mann-Kendall trend test to identify patterns in rainfall. This finding reveals variations in the performance of satellite rainfall products across different stations in the GDRB. The correlation between TAMSAT and Aroresa station was the strongest (r = 0.88), whereas the correlation between ARC2 and Sinana station was the weakest (r = 0.38) at daily timescale. CHIRPS demonstrates robust performance with (r = 0.94) at Arero station, particularly at a monthly temporal scale, indicating its suitability for drought monitoring in the region. Conversely, ARC2 exhibits comparatively poorer performance, underscoring the importance of selecting appropriate satellite products for accurate monitoring. as a result, CHIPRS has been chosen for further drought analysis in this study. For the seasonal and annual trend analysis of rainfall, the MMK trend test indicated majority of the stations has been increasing trend. Moreover, the 3,6- and 12-month SPI values were analyzed based on the selected satellite product (CHIRPS) to assess meteorological drought. the study identifies regions within the basin with heightened drought frequency and severity, emphasizing regions with extreme drought conditions and frequent drought events. Analysis of drought patterns across various temporal scales reveals distinct spatial and temporal distributions, providing valuable insights into potential impacts on water resources, agriculture, and ecosystems. Furthermore, this study also explores the spatiotemporal relationship between SPI and VCI, highlighting VCI's effectiveness as a drought indicator and had weak relationship. Additionally, the association between drought occurrence and El Nino Southern Oscillation (ENSO) highlighted the influence of climate drivers on rainfall patterns. Overall, this research contributes to improving drought monitoring capabilities in Ethiopia, which are essential for mitigating the adverse effects of drought on various sectors. By leveraging multi-satellite rainfall products and robust drought monitoring indices, policymakers and stakeholders can make informed decisions to enhance resilience and adaptation strategies in drought-prone regions like the GDRB.