Onset and performance of the March- May seasonal rainfall over the equatorial east African region

Abstract

The onset of the March-May seasonal rainfall season has been observed to be highly variable. There are years of early, normal and late onset. Prior knowledge of whether there is going to be early or late onsets can be gained if the atmospheric circulation patterns associated with the onsets are understood. In this study the onset of the March-May rainfall season was examined with the aim of identifying atmospheric circulation patterns, which could influence early and late onsets. Forty years of daily and monthly rainfall data for stations spread over the Equatorial East Africa (EEA) region were used in the study. The onset was determined through graphical analysis, use of pentad rainfall and monthly-standardized rainfall anomaly indices. In general the month of March was selected as the onset month of this season. Positive (negati~ve) standardized anomaly indices for the month of March were an indication of an early (late) onset. Analysis of pentad rainfall showed that years of early onset were those that had rainfall starting on or before the 14th pentad. While the years of late onset were those whose rainfall started in the is" pentad or later. Results from the analysis of the QBO phase showed that early (late) onset was favored by the changing phase from negative (positive) at about the months of July/August of the year prior to the rainfall season. It was also noted that onset occurred during the peak negative· phase of the SOL The results from the wind showed some unique circulation patterns during the early/late onset. A veering of the southeasterlies to southwesterlies and northeasterlies to III northwesterlies with a closed circulation near the equator at about 50S, 400E during the months of February which intensified in March at the 700mb level was a common feature in all the selected months of early onset. This mid-level circulation is accompanied by equatorial easterlies with no meridional component at the 200mb level. These circulations suggest low level convergence and upper level divergence within the region and with sufficient moisture could lead to convective activities thus early onset. The prominent features observed during early onset were either very weak or absent during years with late onset especially at 700mb level. Instead the northeasterlies in February were observed in some years to pass over the EEA region and to change to northwesterlies further south of the equator at about 100S. These NE winds were strong and almost parallel to the EEA coast. In other years (for example 1984, 1999) the winds were seen to converge further into the ocean at about 100S, 55°E. The overlying easterlies at 200mb level have a southerly component. One other salient feature in the wind circulation influencing late onset is an anticyclonic circulation in the Arabian region at 700mb level which is conspicuously absent during early onset. These circulations suggest that Arabian high pressure system is closer to the equator and stronger in February and March during late onset. This also explains the presence of the NE observed in the EEA region at this level during this time. The low level divergence occasioned by the presence of easterlies during the month of February could explain the delayed onset in the month of March.