Spatial and temporal variability ofrainfall offers considerable challenges for assessing and understanding climatechange and variability over East Africa.
(Omondi,et al. 2014). East African rainfall ischaracterized by high spatial-temporal variability, and slight variation isassociated with huge socio-economic losses, (Ongoma et al, 2015).The ‘short rains’ (also referred toas SON), though the secondary season in the greater eastern equatorial Africa,exhibit high inter-annual variability. In contrast to the ‘long rains’ (alsoreferred to as MAM), the short rains have shown a rising trend in recent yearsin the equatorial Eastern Africa (Nicholson,2017). They also exhibit one of thestrongest associations ever demonstrated to global circulation (Nicholson,2014). Although, for most of the regionthe short rains are less intense than the long rains, they provide the maximumrainfall in some parts of Kenya (Nicholson,2015). The rainfall pattern over the regionis highly variable both in space and time because of the diverse topographicalfeatures, such as The Rift-valley, Ethiopian highlands, and the Kenyan highlandsalso influence local circulation patterns which in turn influence mesoscaleconvection, lakes such as Lake Victoria, Tanganyika and Kyoga as well as theIndian Ocean along the East African coast causes land-sea contrasts, (Mutai,et al.
,1998). Another feature is the TurkanaLow-Level Jet (LLJ) (Kinuthia& Asnani, 1982) which forms in the channel createdby the two highland ranges. It affects rainfall of the surrounding area byalerting moisture, temperature, and vertical velocity fields, thus modulatingmesoscale circulations (Indeje et al., 2001; Nicholson, 2015).
The ENSOphenomenon which is as a result of the atmospheric and Oceanic interaction inthe Pacific has been established as the dominant mode contributing to eastAfrican rainfall variability (Ogallo, 1988; Hastenrath et al., 1993). Withperiodicities ranging from seasonal to about 8 years (Halpert and Ropelewski,1992).(Nyakwada etal., 2009) used principal component analysis, compositeanalysis and correlation analysis to establish the sea surface temperaturemodes that might denote the collective effect of the Atlantic and Indian Oceanson the seasonal rainfall over East Africa.
Most of the first four modesrepresented sea surface temperature variability associated with the individualOceans, such as basin wide warming/cooling associated with El Nino_SouthernOscillation, inter-hemispheric SST variability over the Atlantic Ocean andIndian Ocean Dipole.In their studyof the relationship between global rainfall and the Southern Oscillation Index(SOI), (Ropelewski and Halpert 1987) concluded that, although the statisticalassociation between rainfall over East Africa and the SOI were weak, there was a high probability ofabnormally wet conditions in the region during El Nino years.The Indian OceanZonal Mode (IOZM), explained by a quasi sea-saw east–west SST gradient reversalacross the Indian Ocean, is also believed to affect the variability of EastAfrican rainfall whereby a positive IOZM indicates warmer and a negative IOZMindicates cooler than normal SST along equatorial East Africa, which alters thelarge-scale wind fields and evaporation, thus higher moisture supply .(Saji, et al. 1999)(Manatsa etal., 2014) explored the inter-annual variability ofEast African ‘short rains’ (EASR) and the influence of Mascarene High (MH)variation using Correlation and composite analyses for flood and droughtoccasions which revealed that the EASR variability is strongly associated tothe MH zonal movement, in specific, the zonal movement of the MH eastern ridge.(Oettli &Camberlin, 2005) compared monthly rainfall distribution withseveral topographic parameters, such as slope and the mean and standarddeviation of elevation.
The relationships varied seasonally. For example,east-facing stations were wetter during the boreal autumn but were drier duringthe boreal summer and winter seasons. Mean elevation appears to have littleeffect on rainfall amount, but it has strong control on the frequency ofrainfall occurrence (Camberlin et al., 2014).