Spatial and temporal trends and variability of rainfall using long-term satellite product over the Upper Blue Nile Basin in Ethiopia

Abstract

The present study aimed to analyze the long-term trends and variability of rainfall in the Upper Blue Nile (UBN) basin in Ethiopia using Climate Hazards Group Infrared Precipitation with Stations (CHIRPS-v2) rainfall product from 1981 to 2018. A pixel-based rainfall variability and trend analysis was made at the annual, seasonal, and monthly temporal scales. The coefficient of variation (CV) was applied to compute the rainfall variability. The trends for each pixel were analyzed using Mann–Kendall (MK) trend test, while the Theil-Sen slope was used to estimate the magnitude of the trend. Results indicated that monthly and seasonal rainfall variabilities are high with the CV up to 140% and 70%, respectively. Annual rainfall variability was found to be low with CV ranging from 6 to 18%. The annual rainfall anomaly analysis, on the other hand, indicated that there are variations in the amount of rainfall received at different years and parts of the basin. It captured the spatial distributions of below- and above-average rainfall periods and its associated drought and flood events, respectively, observed in the basin. In general, the MK monotonic trend analysis for the annual, summer, spring, and autumn rainfall over the entire basin showed an average increasing trend at a rate of 2.48 mm year−1,1.16 mm year−1, 0.92 mm year−1, and 0.67 mm year−1, respectively. In contrast, the rainfall average trend during the winter season (− 0.15 mm year−1) indicated a decreasing trend. The consistency of per-pixel trends with previous studies, based on rain gauge observations, demonstrates the robustness of the trends. Furthermore, the completely raster-based analysis made in this study has provided essential information regarding the spatial variability and trends of rainfall in the basin, which was not satisfactorily captured by the conventional systems due to the larger inter-station distance between rain gauges. Thus, this could be a valuable addition to the existing knowledge of rainfall characteristics in the UBN basin.