The spatial and temporal patterns of the East African seasonal rainfall derived from principal component analysis

Abstract In this study, rotated principal component analysis (RPCA) was used to study the spatial and temporal characteristics of seasonal rainfall over East Africa during the period 1922–1983. The RPCA solutions were derived from both spatial and temporal correlation matrices. The spatial correlation matrices described intercorrelation between pairs of stations, while the temporal matrices gave correlations between pairs of map patterns. Results obtained with the spatial correlation matrices indicated seasonal shifts in the patterns of the dominant RPCA modes which closely resembled the seasonal migration patterns of the rainfall belts induced by the Inter Tropical Convergence Zone (ITCZ). The influence of the large water bodies, especially Lake Victoria and the Indian Ocean, were however, outstanding throughout the year. Twenty‐six homogeneous regional groups were delineated from the spatial characteristics of the dominant eigenvectors. Solutions based on the temporal correlation matrices clustered together some of the wet and dry episodes. Some of the map patterns clustered together could be associated significantly with the El‐Niño/Southern Oscillation events.