Trends and drivers of historical and projected intra-seasonal rainfall characteristics of the MAM season in the cattle corridor of Uganda

Abstract

Uganda’s cattle corridor is largely characterized with semi-arid conditions exhibiting high rainfall variability, periodic late onset, and early cessation. The main objective of this study was to examine the trends and drivers of historical and projected intra-seasonal rainfall characteristics of the MAM season in the selected districts within the cattle corridor. The study used rainfall data based on (CHIRPS), (MERRA-2), and (CORDEX) data sets. This study used Climate Data Tool (CDT) to compute the onset, cessation of rainfall and duration of rain season, used the Man-Kendall test and Sen Slope estimator for trend detection and coefficient of variation to examine intra-seasonal variability. Standardized Anomaly Index was used to investigate the frequency of dry events within the MAM rainfall years. Two sampled student’s t-tests were used to assess the relationship between the drivers and MAM seasonal rainfall and applied a single student’s t-test to test the hypothesis. Study results indicated that historical MAM rainfall season (1991-2020) was less variable (i.e., average CV = 15.3%) and depicted insignificant (p-value >0.05) negative trends (-0.526, -1.375, - 0.010 and -0.840) in onset and cessation (-0.413, -1.043, -1.563, and -0.597) across all study districts, an indication that the rains were coming early and ceasing early in the historical period though not significant. In the future, the MAM seasonal rainfall is expected to be more variable under RCP8.5 (i.e., average CV = 28.0%) than under RCP4.5 (i.e., average CV = 18.7%) and, under these two scenarios, the trends of both onset and cessation dates are expected to be somewhat insignificantly (p-value>0.05) negative, an indication that across the study districts, the rains are expected to come early and cease early. For the drivers of MAM seasonal rainfall, the Nino index, an index of El Niño evolution, showed a statistically significant positive (i.e., t-value = 2.123, pvalue = 0.038) relationship with MAM rainfall and was thus regarded to have a greater influence on MAM seasonal rainfall over the selected districts within the Ugandan cattle corridor. Results from one sample t-test indicated that during the historical period (1991-2020) did change significantly and it is expected to change significantly in the future (2021-2050) under both RCP4.5 and RCP8.5 scenarios. Based on the findings, it is thus imperative to strengthen early warning systems as well as the development of sustainable rangeland, agriculture/agropastoral systems, water resources, and other rainfall-dependent activities within the Uganda’s Cattle Corridor for sustainable livelihoods and development.