Some aspects of the anomalous off-season rainfall over east Africa during the December February season
Mwangi, Charles N.
MetadataShow full item record
For effective planning, development and management of rainfall related social-economic activities there is a need to understand the characteristics of the rainfall variability during the rainy and dry seasons. East Africa, generally experiences two rainfall seasons; between March to May and September to November. These are referred to as the Long and Short rains respectively. Although these are the two major rainy periods over most parts of East Africa, there are some regions, especially over the western parts of East Africa, which receive substantial rainfall during the June-August season. Hence December- February (DJF) may be considered as the truly dry season over most parts of East Africa. However, in some years the supposedly dry season experience quite a lot of rainfall. This study attempts to understand some of the characteristics of these off-season rainfalls. Most of the agriculture activities are concentrated in the main rainfall season, while the off-season rainfall are generally under-utilised. The data used for this study includes daily rainfall data, monthly rainfall data, winds data (surface and upper level), Southern oscillation index (SOl) and sea surface temperatures (SST) over eastern tropical pacific (Nifio-J). Most of the data was obtained from Kenya Meteorological Department. However, some of the Ugandan data were obtained from Uganda Meteorological Department through correspondence. These data extended over the Period 1961-1990. Standard statistical methods, which included regression were used to estimate missing data, whereas inhomogeneity was tested and corrected using the cumulative mass curve technique. Various statistical methods were used to investigate the characteristics of the December-February rainfall events. Rainfall anomaly indices were used to delineate the anomalous wet years during this season. Rotated principal component analysis(RPCA) was used to investigate the space-time features of the DJF-seasonal rainfalls. Spectral analysis examined whether the rainfall anomalous events exhibited any periodic variations. Correlation analysis was used to examine whether there is teleconnection between the anomalous DJF rainfall events and the El-Nifio/Southern Oscillation (ENSO). Five-day (pentad) were used to study the rainfall distribution within the anomalous wet seasons. The results from this study indicated that a number of years had anomalous rainfall during the DJF dry season. The wettest years during this season were 1961/62, 1963/64, 1968/69, 1977178 and 1978179. Pentads analysis from November to the end of March indicated that the rainfall observed during the DJF season in some years was an extension of the short rains, while in others it could be from the early onset of the long rains. In Xl some years both scenanos were observed. However some seasons like 1978179 showed no linkage with either of the rainy seasons. The spatial characteristics revealed by the S-mode analysis of the RPCA agreed well with those observed from the anomaly indices. The wet areas are concentrated near the large water bodies and over southern Tanzania. Spectral analysis of the seasonal rainfall showed periodic fluctuations centred around 2.3-2.7, 3.3-4.3, 4.3-6.3, 6.2-9.5, 9.5-12.2 and 10.8-16.7 years. When daily rainfalls for anomalous rainy DIF-seasons were similarly subjected to spectral analysis, cyclic variations centred around 2.1-2.5, 3.2-4.1, 3.6-4.8, 9-12.9 and 22.5-25.2 days were noted. The DIF rainfall were negatively correlated with SOl and positively correlated with SST. This suggest a possible teleconnection between the anomalous rains during the DJF season and the ENSO. The results of this study intend to form base for the development of a model for the forecast of the off-season rainfall. This will assist in averting the adverse consequences of the heavy rainfall events and proper use of the rain water, in agriculture, industrial, domestic among others.