Dry spells and probability of rainfall occurrence for Lake Kyoga Basin in Uganda, East Africa

Abstract

The economy of Uganda depends heavily on rainfed agriculture. In this study, daily observed rainfall datasets from 9 weather stations with length varying within 1955 and 2017 were used to generate the probability of rainfall and dry spells occurrence using a Markov chain approach. The length of the maximum dry spell was obtained using the direct method based on the definition of a dry day that rainfall is less than 0.85 mm (R < 0.85 mm) and the length of a dry spell is the sum of the number of dry days in a sequence. Mann–Kendall’s statistics (MK) was used to assess the trends in the length of maximum dry spells and Sen’s slope test to estimate the magnitude of change (Q2) in days/per month. MK test results show increasing trends in the length of the maximum dry spells in March at 5 stations, while an insignificant decrease in the length of maximum dry spells is revealed for remaining stations. For the month of April and May, the length of a maximum dry spell is observed to be decreasing across most stations although not statistically significant at the 5% significance level during their respective study periods. The probability of 8 days dry spell is high across all the stations (38–69%) in March, April, and August. This could strongly be related to the changing climate in the region. Negative impacts due to increased length of dry spells could be mitigated through well-timed planting of crops, use of irrigation, and growing of heat-/drought-tolerant crop varieties to match the changing weather and climate patterns.