Assessing Spatio-temporal Variations and Trends of Rainfall over Somalia from 1991 to 2020

Abstract

This study aims to examine the spatiotemporal variability and rainfall trends over Somalia using the Mann-Kendall (MK) test and Sen’s slope estimator to investigate the trend and magnitude of the rainfall changes and the arithmetic mean and coefficient of variation (CV) to examine the rainfall variability. To achieve this aim, the gridded monthly satellite rainfall datasets from Climate Hazards Group InfraRed Precipitation with Station (CHIRPS) for the period between 1991 and 2020 were used. Results show that annual rainfall generally exhibits bimodal rainfall distribution, March to May (Gu) and September to November (Deyr), though southern coastal regions receive extended Gu rains into July and northwestern regions experience early onset of Deyr rains. The study further showed that a low and moderate degree of inter-annual variability (CV<30%) is observed in most parts of the country, while on a seasonal basis, the short rainy season (Deyr) has a higher rainfall variability (CV>30%) compared to the primary rainy season (Gu) and most regions in the south receiving highest amounts of rainfall experience highest variability (CV>50%) during Deyr season. Furthermore, the trend analysis of the annual rainfall showed a non-significant increasing trend. On a seasonal basis, the trends of Gu rainfall decreased in some regions while trends for the Deyr season increased in most regions. However, rainfall trends observed were not statistically significant throughout the country at a significance level of 0.05.