LAUSR

Abstract Analyzing spatiotemporal dynamics of meteorological variables in evolving climate, especially in regions where rain-fed agribusiness is predominant is paramount. In this regard, this study examines the spatiotemporal variability, trend, and magnitude of observed rainfall from 33 synoptic stations during 1979–2017 over Kenya. Non-parametric techniques like Mann–Kendall test, Sequential Mann-Kendall (SMK), and Sen's slope estimator (SSE) were deployed in assessing trend, change points, and magnitude. Results indicate that significant net change of monthly total rainfall mean is observed in May and December. Boreal spring and annual rainfall are insignificantly diminishing at a rate of −1.58 and −0.93 mm/year respectively. During boreal autumn, rainfall is insignificantly increasing at a rate of 1.48 mm/year. High interannual variability was observed, but no abrupt change was recorded during the study period. Spatial distribution of trend reveals dynamic variations from station to station. Amidst boreal spring (autumn), 27 (29) of stations considered had negative (positive) trends wherein 4 (8) stations recorded significant trend at α ≤ 0.01, respectively. At the annual scale, 10 (23) of the stations reported positive (negative) with 1 (5) significant trends at α ≤ 0.01, respectively. At all-time scales, stations at lower eastern and northeastern parts of the country reported a negative trend. Spatial decadal analysis portrays anomalous dry event over the study domain with the first decade being the wettest during boreal spring. To date, anomalously wet conditions exist throughout the country except lower eastern parts during boreal autumn. Comprehensive knowledge about variability, trend, and change point in precipitation is of unprecedented interest to scientists since they are prerequisites for developing effective contingency strategies to curb the impacts of climate change and future preparedness.